bigcode/jupyter-parsed · Datasets at Hugging Face

ec7a1edd7210f519edca4467ffb09133f87f207d

46,132

ipynb

Jupyter Notebook

notebook/Text_Classification_RoBERTa.ipynb

bimhud/pytorch-transformers

e4eb875244e165d563d31fa9f127bdafa598145d

[ "Apache-2.0" ]

null

notebook/Text_Classification_RoBERTa.ipynb

bimhud/pytorch-transformers

e4eb875244e165d563d31fa9f127bdafa598145d

[ "Apache-2.0" ]

null

notebook/Text_Classification_RoBERTa.ipynb

bimhud/pytorch-transformers

e4eb875244e165d563d31fa9f127bdafa598145d

[ "Apache-2.0" ]

null

33.16463

262

0.480166

[ [ [ "<a href=\"https://colab.research.google.com/github/bimhud/pytorch-transformers/blob/master/notebook/Text_Classification_RoBERTa.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>", "_____no_output_____" ] ], [ [ "import pandas as pd\nimport numpy as np\nimport json, re\nfrom tqdm import tqdm_notebook\nfrom uuid import uuid4\n\n## Torch Modules\nimport torch\nimport torch.optim as optim\nimport torch.nn as nn\nimport torch.nn.functional as F\nfrom torch.autograd import Variable\nfrom torch.utils.data import Dataset, DataLoader\n", "_____no_output_____" ], [ "!pip install -U pytorch-transformers", "Collecting pytorch-transformers\n\u001b[?25l Downloading https://files.pythonhosted.org/packages/a3/b7/d3d18008a67e0b968d1ab93ad444fc05699403fa662f634b2f2c318a508b/pytorch_transformers-1.2.0-py3-none-any.whl (176kB)\n\u001b[K |████████████████████████████████| 184kB 2.8MB/s \n\u001b[?25hRequirement already satisfied, skipping upgrade: boto3 in /usr/local/lib/python3.6/dist-packages (from pytorch-transformers) (1.9.224)\nCollecting sacremoses (from pytorch-transformers)\n\u001b[?25l Downloading https://files.pythonhosted.org/packages/df/24/0b86f494d3a5c7531f6d0c77d39fd8f9d42e651244505d3d737e31db9a4d/sacremoses-0.0.33.tar.gz (802kB)\n\u001b[K |████████████████████████████████| 808kB 43.7MB/s \n\u001b[?25hRequirement already satisfied, skipping upgrade: tqdm in /usr/local/lib/python3.6/dist-packages (from pytorch-transformers) (4.28.1)\nRequirement already satisfied, skipping upgrade: torch>=1.0.0 in /usr/local/lib/python3.6/dist-packages (from pytorch-transformers) (1.1.0)\nCollecting sentencepiece (from pytorch-transformers)\n\u001b[?25l Downloading https://files.pythonhosted.org/packages/14/3d/efb655a670b98f62ec32d66954e1109f403db4d937c50d779a75b9763a29/sentencepiece-0.1.83-cp36-cp36m-manylinux1_x86_64.whl (1.0MB)\n\u001b[K |████████████████████████████████| 1.0MB 38.8MB/s \n\u001b[?25hRequirement already satisfied, skipping upgrade: requests in /usr/local/lib/python3.6/dist-packages (from pytorch-transformers) (2.21.0)\nRequirement already satisfied, skipping upgrade: numpy in /usr/local/lib/python3.6/dist-packages (from pytorch-transformers) (1.16.5)\nCollecting regex (from pytorch-transformers)\n\u001b[?25l Downloading https://files.pythonhosted.org/packages/6f/a6/99eeb5904ab763db87af4bd71d9b1dfdd9792681240657a4c0a599c10a81/regex-2019.08.19.tar.gz (654kB)\n\u001b[K |████████████████████████████████| 655kB 42.8MB/s \n\u001b[?25hRequirement already satisfied, skipping upgrade: s3transfer<0.3.0,>=0.2.0 in /usr/local/lib/python3.6/dist-packages (from boto3->pytorch-transformers) (0.2.1)\nRequirement already satisfied, skipping upgrade: botocore<1.13.0,>=1.12.224 in /usr/local/lib/python3.6/dist-packages (from boto3->pytorch-transformers) (1.12.224)\nRequirement already satisfied, skipping upgrade: jmespath<1.0.0,>=0.7.1 in /usr/local/lib/python3.6/dist-packages (from boto3->pytorch-transformers) (0.9.4)\nRequirement already satisfied, skipping upgrade: six in /usr/local/lib/python3.6/dist-packages (from sacremoses->pytorch-transformers) (1.12.0)\nRequirement already satisfied, skipping upgrade: click in /usr/local/lib/python3.6/dist-packages (from sacremoses->pytorch-transformers) (7.0)\nRequirement already satisfied, skipping upgrade: joblib in /usr/local/lib/python3.6/dist-packages (from sacremoses->pytorch-transformers) (0.13.2)\nRequirement already satisfied, skipping upgrade: urllib3<1.25,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from requests->pytorch-transformers) (1.24.3)\nRequirement already satisfied, skipping upgrade: idna<2.9,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests->pytorch-transformers) (2.8)\nRequirement already satisfied, skipping upgrade: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests->pytorch-transformers) (2019.6.16)\nRequirement already satisfied, skipping upgrade: chardet<3.1.0,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests->pytorch-transformers) (3.0.4)\nRequirement already satisfied, skipping upgrade: docutils<0.16,>=0.10 in /usr/local/lib/python3.6/dist-packages (from botocore<1.13.0,>=1.12.224->boto3->pytorch-transformers) (0.15.2)\nRequirement already satisfied, skipping upgrade: python-dateutil<3.0.0,>=2.1; python_version >= \"2.7\" in /usr/local/lib/python3.6/dist-packages (from botocore<1.13.0,>=1.12.224->boto3->pytorch-transformers) (2.5.3)\nBuilding wheels for collected packages: sacremoses, regex\n Building wheel for sacremoses (setup.py) ... \u001b[?25l\u001b[?25hdone\n Created wheel for sacremoses: filename=sacremoses-0.0.33-cp36-none-any.whl size=833106 sha256=245901e68cfa928189929216a6976dd8f8e5ef0a1555f92e54cb69f4827d73df\n Stored in directory: /root/.cache/pip/wheels/70/87/56/e40575cca30d12fee8875d523b8878b7aba866a9f03b2fd983\n Building wheel for regex (setup.py) ... \u001b[?25l\u001b[?25hdone\n Created wheel for regex: filename=regex-2019.8.19-cp36-cp36m-linux_x86_64.whl size=609261 sha256=23f0f950a31c077258d25b38d0e835dbaeba03d534f5f180a751a0db5e73e27e\n Stored in directory: /root/.cache/pip/wheels/90/04/07/b5010fb816721eb3d6dd64ed5cc8111ca23f97fdab8619b5be\nSuccessfully built sacremoses regex\nInstalling collected packages: sacremoses, sentencepiece, regex, pytorch-transformers\nSuccessfully installed pytorch-transformers-1.2.0 regex-2019.8.19 sacremoses-0.0.33 sentencepiece-0.1.83\n" ], [ "## PyTorch Transformer\nfrom pytorch_transformers import RobertaModel, RobertaTokenizer\nfrom pytorch_transformers import RobertaForSequenceClassification, RobertaConfig", "_____no_output_____" ], [ "", "_____no_output_____" ], [ "## Check if Cuda is Available\nprint(torch.cuda.is_available())", "True\n" ], [ "## Install PyTorch-Transformer", "_____no_output_____" ], [ "!git clone https://github.com/snipsco/nlu-benchmark.git", "Cloning into 'nlu-benchmark'...\nremote: Enumerating objects: 11, done.\u001b[K\nremote: Counting objects: 100% (11/11), done.\u001b[K\nremote: Compressing objects: 100% (9/9), done.\u001b[K\nremote: Total 389 (delta 2), reused 11 (delta 2), pack-reused 378\u001b[K\nReceiving objects: 100% (389/389), 1.24 MiB | 10.55 MiB/s, done.\nResolving deltas: 100% (242/242), done.\n" ], [ "## Importing Datasets", "_____no_output_____" ], [ "!ls nlu-benchmark/2017-06-custom-intent-engines", "AddToPlaylist\tGetWeather RateBook SearchCreativeWork\nBookRestaurant\tPlayMusic README.md SearchScreeningEvent\n" ], [ "dataset_path = \"nlu-benchmark/2017-06-custom-intent-engines\"", "_____no_output_____" ], [ "import os\ndataset = pd.DataFrame(columns = ['utterance', 'label'])\nfor intent in ['AddToPlaylist', 'BookRestaurant', 'GetWeather', 'PlayMusic', 'RateBook', 'SearchCreativeWork',\n 'SearchScreeningEvent']:\n with open(dataset_path + os.sep + intent + os.sep + \"train_\" + intent + \".json\",\n encoding='cp1251') as data_file:\n data = json.load(data_file)\n print(\"Class: {}, # utterances: {}\".format(intent,len(data[intent])))\n texts = []\n for i in range(len(data[intent])):\n text = ''\n for j in range(len(data[intent][i]['data'])):\n text += data[intent][i]['data'][j]['text']\n dataset = dataset.append({'utterance': text, 'label': intent}, ignore_index=True)", "Class: AddToPlaylist, # utterances: 300\nClass: BookRestaurant, # utterances: 300\nClass: GetWeather, # utterances: 300\nClass: PlayMusic, # utterances: 300\nClass: RateBook, # utterances: 300\nClass: SearchCreativeWork, # utterances: 300\nClass: SearchScreeningEvent, # utterances: 300\n" ], [ "#Get index from multiple labels\nlabel_to_ix = {}\nfor label in dataset.label:\n for word in label.split():\n if word not in label_to_ix:\n label_to_ix[word]=len(label_to_ix)\nlabel_to_ix", "_____no_output_____" ], [ "## Loading RoBERTa classes", "_____no_output_____" ], [ "config = RobertaConfig.from_pretrained('roberta-base')\nconfig.num_labels = len(list(label_to_ix.values()))\nconfig", "100%|██████████| 473/473 [00:00<00:00, 200010.67B/s]\n" ], [ "tokenizer = RobertaTokenizer.from_pretrained('roberta-base')\nmodel = RobertaForSequenceClassification(config)", "100%|██████████| 898823/898823 [00:00<00:00, 11059490.21B/s]\n100%|██████████| 456318/456318 [00:00<00:00, 7037203.60B/s]\n" ], [ "## Feature Preparation", "_____no_output_____" ], [ "def prepare_features(seq_1, max_seq_length = 300, \n zero_pad = False, include_CLS_token = True, include_SEP_token = True):\n ## Tokenzine Input\n tokens_a = tokenizer.tokenize(seq_1)\n\n ## Truncate\n if len(tokens_a) > max_seq_length - 2:\n tokens_a = tokens_a[0:(max_seq_length - 2)]\n ## Initialize Tokens\n tokens = []\n if include_CLS_token:\n tokens.append(tokenizer.cls_token)\n ## Add Tokens and separators\n for token in tokens_a:\n tokens.append(token)\n\n if include_SEP_token:\n tokens.append(tokenizer.sep_token)\n\n input_ids = tokenizer.convert_tokens_to_ids(tokens)\n ## Input Mask \n input_mask = [1] * len(input_ids)\n ## Zero-pad sequence lenght\n if zero_pad:\n while len(input_ids) < max_seq_length:\n input_ids.append(0)\n input_mask.append(0)\n return torch.tensor(input_ids).unsqueeze(0), input_mask", "_____no_output_____" ], [ "msg = \"My dog is cute!\"\nprepare_features(msg)", "_____no_output_____" ], [ "## Dataset Loader Classes", "_____no_output_____" ], [ "class Intents(Dataset):\n def __init__(self, dataframe):\n self.len = len(dataframe)\n self.data = dataframe\n \n def __getitem__(self, index):\n utterance = self.data.utterance[index]\n label = self.data.label[index]\n X, _ = prepare_features(utterance)\n y = label_to_ix[self.data.label[index]]\n return X, y\n \n def __len__(self):\n return self.len", "_____no_output_____" ], [ "train_size = 0.8\ntrain_dataset=dataset.sample(frac=train_size,random_state=200).reset_index(drop=True)\ntest_dataset=dataset.drop(train_dataset.index).reset_index(drop=True)", "_____no_output_____" ], [ "print(\"FULL Dataset: {}\".format(dataset.shape))\nprint(\"TRAIN Dataset: {}\".format(train_dataset.shape))\nprint(\"TEST Dataset: {}\".format(test_dataset.shape))", "FULL Dataset: (2100, 2)\nTRAIN Dataset: (1680, 2)\nTEST Dataset: (420, 2)\n" ], [ "training_set = Intents(train_dataset)\ntesting_set = Intents(test_dataset)", "_____no_output_____" ], [ "training_set.__getitem__(0)[0].shape", "_____no_output_____" ], [ "model(training_set.__getitem__(0)[0])", "_____no_output_____" ], [ "## Training Params", "_____no_output_____" ], [ "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\nmodel = model.cuda()", "_____no_output_____" ], [ "# Parameters\nparams = {'batch_size': 1,\n 'shuffle': True,\n 'num_workers': 1}", "_____no_output_____" ], [ "training_loader = DataLoader(training_set, **params)\ntesting_loader = DataLoader(testing_set, **params)", "_____no_output_____" ], [ "loss_function = nn.CrossEntropyLoss()\nlearning_rate = 1e-05\noptimizer = optim.Adam(params = model.parameters(), lr=learning_rate)", "_____no_output_____" ], [ "## Test Forward Pass\ninp = training_set.__getitem__(0)[0].cuda()\noutput = model(inp)[0]\ntorch.max(output.data, 1)", "_____no_output_____" ], [ "max_epochs = 3\nmodel = model.train()\nfor epoch in tqdm_notebook(range(max_epochs)):\n print(\"EPOCH -- {}\".format(epoch))\n for i, (sent, label) in enumerate(training_loader):\n optimizer.zero_grad()\n sent = sent.squeeze(0)\n if torch.cuda.is_available():\n sent = sent.cuda()\n label = label.cuda()\n output = model.forward(sent)[0]\n _, predicted = torch.max(output, 1)\n \n loss = loss_function(output, label)\n loss.backward()\n optimizer.step()\n \n if i%100 == 0:\n correct = 0\n total = 0\n for sent, label in testing_loader:\n sent = sent.squeeze(0)\n if torch.cuda.is_available():\n sent = sent.cuda()\n label = label.cuda()\n output = model.forward(sent)[0]\n _, predicted = torch.max(output.data, 1)\n total += label.size(0)\n correct += (predicted.cpu() == label.cpu()).sum()\n accuracy = 100.00 * correct.numpy() / total\n print('Iteration: {}. Loss: {}. Accuracy: {}%'.format(i, loss.item(), accuracy))", "_____no_output_____" ], [ "!mkdir nlu-benchmark/2017-06-custom-intent-engines/RoBerta_Model", "_____no_output_____" ], [ " torch.save(model.state_dict(), 'nlu-benchmark/2017-06-custom-intent-engines/RoBerta_Model/roberta_state_dict_'+ str(uuid4())+'.pth')", "_____no_output_____" ], [ "dataset.tail(5)", "_____no_output_____" ], [ "## Load model\n!ls nlu-benchmark/2017-06-custom-intent-engines/RoBerta_Model", "roberta_state_dict_4fd91890-1424-41fb-8a7a-059fc60bc379.pth\n" ], [ "model_path = 'nlu-benchmark/2017-06-custom-intent-engines/RoBerta_Model/roberta_state_dict_4fd91890-1424-41fb-8a7a-059fc60bc379.pth'", "_____no_output_____" ], [ "%%time\nmodel.load_state_dict(torch.load(model_path, map_location=device))", "CPU times: user 92.8 ms, sys: 240 ms, total: 333 ms\nWall time: 339 ms\n" ], [ "def get_reply(msg):\n model.eval()\n input_msg, _ = prepare_features(msg)\n if torch.cuda.is_available():\n input_msg = input_msg.cuda()\n output = model(input_msg)[0]\n pred_score, pred_label = torch.max(output.data, 1)\n \n prediction=list(label_to_ix.keys())[pred_label]\n return prediction,pred_score.cpu().numpy()[0]", "_____no_output_____" ], [ "label_to_ix.keys()", "_____no_output_____" ], [ "get_reply(\"play radiohead song\")", "_____no_output_____" ], [ "get_reply(\"it is rainy in Sao Paulo\")", "_____no_output_____" ], [ "get_reply(\"sun shinnes all day\")", "_____no_output_____" ], [ "get_reply(\"low humidity, high altitude\")", "_____no_output_____" ], [ "get_reply(\"Book tacos for me tonight\")", "_____no_output_____" ], [ "get_reply(\"Book a table for me tonight\")", "_____no_output_____" ], [ "get_reply(\"I want BBQ tonight under the rain\")", "_____no_output_____" ], [ "", "_____no_output_____" ] ] ]

[ "markdown", "code" ]

[ [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7a2203c3a54b1d5a1db6197e6c3b98f41d62b8

59,891

ipynb

Jupyter Notebook

notebooks/community/sdk/sdk_custom_image_classification_online.ipynb

nayaknishant/vertex-ai-samples

3ce120b953f1cdc2ec2c5a3f4509cfeab106b7d0

[ "Apache-2.0" ]

213

2021-06-10T20:05:20.000Z

2022-03-31T16:09:29.000Z

notebooks/community/sdk/sdk_custom_image_classification_online.ipynb

nayaknishant/vertex-ai-samples

3ce120b953f1cdc2ec2c5a3f4509cfeab106b7d0

[ "Apache-2.0" ]

343

2021-07-25T22:55:25.000Z

2022-03-31T23:58:47.000Z

notebooks/community/sdk/sdk_custom_image_classification_online.ipynb

nayaknishant/vertex-ai-samples

3ce120b953f1cdc2ec2c5a3f4509cfeab106b7d0

[ "Apache-2.0" ]

143

2021-07-21T17:27:47.000Z

2022-03-29T01:20:43.000Z

38.293478

425

0.543404

[ [ [ "# Copyright 2021 Google LLC\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not use this file except in compliance with the License.\n# You may obtain a copy of the License at\n#\n# https://www.apache.org/licenses/LICENSE-2.0\n#\n# Unless required by applicable law or agreed to in writing, software\n# distributed under the License is distributed on an \"AS IS\" BASIS,\n# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n# See the License for the specific language governing permissions and\n# limitations under the License.", "_____no_output_____" ] ], [ [ "# Vertex SDK: Custom training image classification model for online prediction\n\n<table align=\"left\">\n <td>\n <a href=\"https://colab.research.google.com/github/GoogleCloudPlatform/vertex-ai-samples/tree/master/notebooks/official/automl/sdk_custom_image_classification_online.ipynb\">\n <img src=\"https://cloud.google.com/ml-engine/images/colab-logo-32px.png\" alt=\"Colab logo\"> Run in Colab\n </a>\n </td>\n <td>\n <a href=\"https://github.com/GoogleCloudPlatform/vertex-ai-samples/tree/master/notebooks/official/automl/sdk_custom_image_classification_online.ipynb\">\n <img src=\"https://cloud.google.com/ml-engine/images/github-logo-32px.png\" alt=\"GitHub logo\">\n View on GitHub\n </a>\n </td>\n <td>\n <a href=\"https://console.cloud.google.com/ai/platform/notebooks/deploy-notebook?download_url=https://github.com/GoogleCloudPlatform/vertex-ai-samples/tree/master/notebooks/official/automl/sdk_custom_image_classification_online.ipynb\">\n Open in Google Cloud Notebooks\n </a>\n </td>\n</table>\n<br/><br/><br/>", "_____no_output_____" ], [ "## Overview\n\n\nThis tutorial demonstrates how to use the Vertex SDK to train and deploy a custom image classification model for online prediction.", "_____no_output_____" ], [ "### Dataset\n\nThe dataset used for this tutorial is the [CIFAR10 dataset](https://www.tensorflow.org/datasets/catalog/cifar10) from [TensorFlow Datasets](https://www.tensorflow.org/datasets/catalog/overview). The version of the dataset you will use is built into TensorFlow. The trained model predicts which type of class an image is from ten classes: airplane, automobile, bird, cat, deer, dog, frog, horse, ship, truck.", "_____no_output_____" ], [ "### Objective\n\nIn this tutorial, you create a custom model from a Python script in a Google prebuilt Docker container using the Vertex SDK, and then do a prediction on the deployed model by sending data. You can alternatively create custom models using `gcloud` command-line tool or online using Cloud Console.\n\nThe steps performed include:\n\n- Create a Vertex custom job for training a model.\n- Train a TensorFlow model.\n- Retrieve and load the model artifacts.\n- View the model evaluation.\n- Upload the model as a Vertex `Model` resource.\n- Deploy the `Model` resource to a serving `Endpoint` resource.\n- Make a prediction.\n- Undeploy the `Model` resource.", "_____no_output_____" ], [ "### Costs\n\nThis tutorial uses billable components of Google Cloud:\n\n* Vertex AI\n* Cloud Storage\n\nLearn about [Vertex AI\npricing](https://cloud.google.com/vertex-ai/pricing) and [Cloud Storage\npricing](https://cloud.google.com/storage/pricing), and use the [Pricing\nCalculator](https://cloud.google.com/products/calculator/)\nto generate a cost estimate based on your projected usage.", "_____no_output_____" ], [ "### Set up your local development environment\n\nIf you are using Colab or Google Cloud Notebooks, your environment already meets all the requirements to run this notebook. You can skip this step.\n\nOtherwise, make sure your environment meets this notebook's requirements. You need the following:\n\n- The Cloud Storage SDK\n- Git\n- Python 3\n- virtualenv\n- Jupyter notebook running in a virtual environment with Python 3\n\nThe Cloud Storage guide to [Setting up a Python development environment](https://cloud.google.com/python/setup) and the [Jupyter installation guide](https://jupyter.org/install) provide detailed instructions for meeting these requirements. The following steps provide a condensed set of instructions:\n\n1. [Install and initialize the SDK](https://cloud.google.com/sdk/docs/).\n\n2. [Install Python 3](https://cloud.google.com/python/setup#installing_python).\n\n3. [Install virtualenv](https://cloud.google.com/python/setup#installing_and_using_virtualenv) and create a virtual environment that uses Python 3. Activate the virtual environment.\n\n4. To install Jupyter, run `pip3 install jupyter` on the command-line in a terminal shell.\n\n5. To launch Jupyter, run `jupyter notebook` on the command-line in a terminal shell.\n\n6. Open this notebook in the Jupyter Notebook Dashboard.\n", "_____no_output_____" ], [ "## Installation\n\nInstall the latest version of Vertex SDK for Python.", "_____no_output_____" ] ], [ [ "import os\n\n# Google Cloud Notebook\nif os.path.exists(\"/opt/deeplearning/metadata/env_version\"):\n USER_FLAG = \"--user\"\nelse:\n USER_FLAG = \"\"\n\n! pip3 install --upgrade google-cloud-aiplatform $USER_FLAG", "_____no_output_____" ] ], [ [ "Install the latest GA version of *google-cloud-storage* library as well.", "_____no_output_____" ] ], [ [ "! pip3 install -U google-cloud-storage $USER_FLAG", "_____no_output_____" ], [ "if os.environ[\"IS_TESTING\"]:\n ! pip3 install --upgrade tensorflow $USER_FLAG", "_____no_output_____" ] ], [ [ "### Restart the kernel\n\nOnce you've installed the additional packages, you need to restart the notebook kernel so it can find the packages.", "_____no_output_____" ] ], [ [ "import os\n\nif not os.getenv(\"IS_TESTING\"):\n # Automatically restart kernel after installs\n import IPython\n\n app = IPython.Application.instance()\n app.kernel.do_shutdown(True)", "_____no_output_____" ] ], [ [ "## Before you begin\n\n### GPU runtime\n\nThis tutorial does not require a GPU runtime.\n\n### Set up your Google Cloud project\n\n**The following steps are required, regardless of your notebook environment.**\n\n1. [Select or create a Google Cloud project](https://console.cloud.google.com/cloud-resource-manager). When you first create an account, you get a $300 free credit towards your compute/storage costs.\n\n2. [Make sure that billing is enabled for your project.](https://cloud.google.com/billing/docs/how-to/modify-project)\n\n3. [Enable the following APIs: Vertex AI APIs, Compute Engine APIs, and Cloud Storage.](https://console.cloud.google.com/flows/enableapi?apiid=ml.googleapis.com,compute_component,storage-component.googleapis.com)\n\n4. If you are running this notebook locally, you will need to install the [Cloud SDK]((https://cloud.google.com/sdk)).\n\n5. Enter your project ID in the cell below. Then run the cell to make sure the\nCloud SDK uses the right project for all the commands in this notebook.\n\n**Note**: Jupyter runs lines prefixed with `!` as shell commands, and it interpolates Python variables prefixed with `$`.", "_____no_output_____" ] ], [ [ "PROJECT_ID = \"[your-project-id]\" # @param {type:\"string\"}", "_____no_output_____" ], [ "if PROJECT_ID == \"\" or PROJECT_ID is None or PROJECT_ID == \"[your-project-id]\":\n # Get your GCP project id from gcloud\n shell_output = ! gcloud config list --format 'value(core.project)' 2>/dev/null\n PROJECT_ID = shell_output[0]\n print(\"Project ID:\", PROJECT_ID)", "_____no_output_____" ], [ "! gcloud config set project $PROJECT_ID", "_____no_output_____" ] ], [ [ "#### Region\n\nYou can also change the `REGION` variable, which is used for operations\nthroughout the rest of this notebook. Below are regions supported for Vertex AI. We recommend that you choose the region closest to you.\n\n- Americas: `us-central1`\n- Europe: `europe-west4`\n- Asia Pacific: `asia-east1`\n\nYou may not use a multi-regional bucket for training with Vertex AI. Not all regions provide support for all Vertex AI services.\n\nLearn more about [Vertex AI regions](https://cloud.google.com/vertex-ai/docs/general/locations)", "_____no_output_____" ] ], [ [ "REGION = \"us-central1\" # @param {type: \"string\"}", "_____no_output_____" ] ], [ [ "#### Timestamp\n\nIf you are in a live tutorial session, you might be using a shared test account or project. To avoid name collisions between users on resources created, you create a timestamp for each instance session, and append the timestamp onto the name of resources you create in this tutorial.", "_____no_output_____" ] ], [ [ "from datetime import datetime\n\nTIMESTAMP = datetime.now().strftime(\"%Y%m%d%H%M%S\")", "_____no_output_____" ] ], [ [ "### Authenticate your Google Cloud account\n\n**If you are using Google Cloud Notebooks**, your environment is already authenticated. Skip this step.\n\n**If you are using Colab**, run the cell below and follow the instructions when prompted to authenticate your account via oAuth.\n\n**Otherwise**, follow these steps:\n\nIn the Cloud Console, go to the [Create service account key](https://console.cloud.google.com/apis/credentials/serviceaccountkey) page.\n\n**Click Create service account**.\n\nIn the **Service account name** field, enter a name, and click **Create**.\n\nIn the **Grant this service account access to project** section, click the Role drop-down list. Type \"Vertex\" into the filter box, and select **Vertex Administrator**. Type \"Storage Object Admin\" into the filter box, and select **Storage Object Admin**.\n\nClick Create. A JSON file that contains your key downloads to your local environment.\n\nEnter the path to your service account key as the GOOGLE_APPLICATION_CREDENTIALS variable in the cell below and run the cell.", "_____no_output_____" ] ], [ [ "# If you are running this notebook in Colab, run this cell and follow the\n# instructions to authenticate your GCP account. This provides access to your\n# Cloud Storage bucket and lets you submit training jobs and prediction\n# requests.\n\nimport os\nimport sys\n\n# If on Google Cloud Notebook, then don't execute this code\nif not os.path.exists(\"/opt/deeplearning/metadata/env_version\"):\n if \"google.colab\" in sys.modules:\n from google.colab import auth as google_auth\n\n google_auth.authenticate_user()\n\n # If you are running this notebook locally, replace the string below with the\n # path to your service account key and run this cell to authenticate your GCP\n # account.\n elif not os.getenv(\"IS_TESTING\"):\n %env GOOGLE_APPLICATION_CREDENTIALS ''", "_____no_output_____" ] ], [ [ "### Create a Cloud Storage bucket\n\n**The following steps are required, regardless of your notebook environment.**\n\nWhen you initialize the Vertex SDK for Python, you specify a Cloud Storage staging bucket. The staging bucket is where all the data associated with your dataset and model resources are retained across sessions.\n\nSet the name of your Cloud Storage bucket below. Bucket names must be globally unique across all Google Cloud projects, including those outside of your organization.", "_____no_output_____" ] ], [ [ "BUCKET_NAME = \"gs://[your-bucket-name]\" # @param {type:\"string\"}", "_____no_output_____" ], [ "if BUCKET_NAME == \"\" or BUCKET_NAME is None or BUCKET_NAME == \"gs://[your-bucket-name]\":\n BUCKET_NAME = \"gs://\" + PROJECT_ID + \"aip-\" + TIMESTAMP", "_____no_output_____" ] ], [ [ "**Only if your bucket doesn't already exist**: Run the following cell to create your Cloud Storage bucket.", "_____no_output_____" ] ], [ [ "! gsutil mb -l $REGION $BUCKET_NAME", "_____no_output_____" ] ], [ [ "Finally, validate access to your Cloud Storage bucket by examining its contents:", "_____no_output_____" ] ], [ [ "! gsutil ls -al $BUCKET_NAME", "_____no_output_____" ] ], [ [ "### Set up variables\n\nNext, set up some variables used throughout the tutorial.\n### Import libraries and define constants", "_____no_output_____" ] ], [ [ "import google.cloud.aiplatform as aip", "_____no_output_____" ] ], [ [ "## Initialize Vertex SDK for Python\n\nInitialize the Vertex SDK for Python for your project and corresponding bucket.", "_____no_output_____" ] ], [ [ "aip.init(project=PROJECT_ID, staging_bucket=BUCKET_NAME)", "_____no_output_____" ] ], [ [ "#### Set hardware accelerators\n\nYou can set hardware accelerators for training and prediction.\n\nSet the variables `TRAIN_GPU/TRAIN_NGPU` and `DEPLOY_GPU/DEPLOY_NGPU` to use a container image supporting a GPU and the number of GPUs allocated to the virtual machine (VM) instance. For example, to use a GPU container image with 4 Nvidia Telsa K80 GPUs allocated to each VM, you would specify:\n\n (aip.AcceleratorType.NVIDIA_TESLA_K80, 4)\n\n\nOtherwise specify `(None, None)` to use a container image to run on a CPU.\n\nLearn more [here](https://cloud.google.com/vertex-ai/docs/general/locations#accelerators) hardware accelerator support for your region\n\n*Note*: TF releases before 2.3 for GPU support will fail to load the custom model in this tutorial. It is a known issue and fixed in TF 2.3 -- which is caused by static graph ops that are generated in the serving function. If you encounter this issue on your own custom models, use a container image for TF 2.3 with GPU support.", "_____no_output_____" ] ], [ [ "if os.getenv(\"IS_TESTING_TRAIN_GPU\"):\n TRAIN_GPU, TRAIN_NGPU = (\n aip.gapic.AcceleratorType.NVIDIA_TESLA_K80,\n int(os.getenv(\"IS_TESTING_TRAIN_GPU\")),\n )\nelse:\n TRAIN_GPU, TRAIN_NGPU = (None, None)\n\nif os.getenv(\"IS_TESTING_DEPLOY_GPU\"):\n DEPLOY_GPU, DEPLOY_NGPU = (\n aip.gapic.AcceleratorType.NVIDIA_TESLA_K80,\n int(os.getenv(\"IS_TESTING_DEPLOY_GPU\")),\n )\nelse:\n DEPLOY_GPU, DEPLOY_NGPU = (None, None)", "_____no_output_____" ] ], [ [ "#### Set pre-built containers\n\nSet the pre-built Docker container image for training and prediction.\n\n\nFor the latest list, see [Pre-built containers for training](https://cloud.google.com/ai-platform-unified/docs/training/pre-built-containers).\n\n\nFor the latest list, see [Pre-built containers for prediction](https://cloud.google.com/ai-platform-unified/docs/predictions/pre-built-containers).", "_____no_output_____" ] ], [ [ "if os.getenv(\"IS_TESTING_TF\"):\n TF = os.getenv(\"IS_TESTING_TF\")\nelse:\n TF = \"2-1\"\n\nif TF[0] == \"2\":\n if TRAIN_GPU:\n TRAIN_VERSION = \"tf-gpu.{}\".format(TF)\n else:\n TRAIN_VERSION = \"tf-cpu.{}\".format(TF)\n if DEPLOY_GPU:\n DEPLOY_VERSION = \"tf2-gpu.{}\".format(TF)\n else:\n DEPLOY_VERSION = \"tf2-cpu.{}\".format(TF)\nelse:\n if TRAIN_GPU:\n TRAIN_VERSION = \"tf-gpu.{}\".format(TF)\n else:\n TRAIN_VERSION = \"tf-cpu.{}\".format(TF)\n if DEPLOY_GPU:\n DEPLOY_VERSION = \"tf-gpu.{}\".format(TF)\n else:\n DEPLOY_VERSION = \"tf-cpu.{}\".format(TF)\n\nTRAIN_IMAGE = \"gcr.io/cloud-aiplatform/training/{}:latest\".format(TRAIN_VERSION)\nDEPLOY_IMAGE = \"gcr.io/cloud-aiplatform/prediction/{}:latest\".format(DEPLOY_VERSION)\n\nprint(\"Training:\", TRAIN_IMAGE, TRAIN_GPU, TRAIN_NGPU)\nprint(\"Deployment:\", DEPLOY_IMAGE, DEPLOY_GPU, DEPLOY_NGPU)", "_____no_output_____" ] ], [ [ "#### Set machine type\n\nNext, set the machine type to use for training and prediction.\n\n- Set the variables `TRAIN_COMPUTE` and `DEPLOY_COMPUTE` to configure the compute resources for the VMs you will use for for training and prediction.\n - `machine type`\n - `n1-standard`: 3.75GB of memory per vCPU.\n - `n1-highmem`: 6.5GB of memory per vCPU\n - `n1-highcpu`: 0.9 GB of memory per vCPU\n - `vCPUs`: number of \\[2, 4, 8, 16, 32, 64, 96 \\]\n\n*Note: The following is not supported for training:*\n\n - `standard`: 2 vCPUs\n - `highcpu`: 2, 4 and 8 vCPUs\n\n*Note: You may also use n2 and e2 machine types for training and deployment, but they do not support GPUs*.", "_____no_output_____" ] ], [ [ "if os.getenv(\"IS_TESTING_TRAIN_MACHINE\"):\n MACHINE_TYPE = os.getenv(\"IS_TESTING_TRAIN_MACHINE\")\nelse:\n MACHINE_TYPE = \"n1-standard\"\n\nVCPU = \"4\"\nTRAIN_COMPUTE = MACHINE_TYPE + \"-\" + VCPU\nprint(\"Train machine type\", TRAIN_COMPUTE)\n\nif os.getenv(\"IS_TESTING_DEPLOY_MACHINE\"):\n MACHINE_TYPE = os.getenv(\"IS_TESTING_DEPLOY_MACHINE\")\nelse:\n MACHINE_TYPE = \"n1-standard\"\n\nVCPU = \"4\"\nDEPLOY_COMPUTE = MACHINE_TYPE + \"-\" + VCPU\nprint(\"Deploy machine type\", DEPLOY_COMPUTE)", "_____no_output_____" ] ], [ [ "# Tutorial\n\nNow you are ready to start creating your own custom model and training for CIFAR10.", "_____no_output_____" ], [ "### Examine the training package\n\n#### Package layout\n\nBefore you start the training, you will look at how a Python package is assembled for a custom training job. When unarchived, the package contains the following directory/file layout.\n\n- PKG-INFO\n- README.md\n- setup.cfg\n- setup.py\n- trainer\n - \\_\\_init\\_\\_.py\n - task.py\n\nThe files `setup.cfg` and `setup.py` are the instructions for installing the package into the operating environment of the Docker image.\n\nThe file `trainer/task.py` is the Python script for executing the custom training job. *Note*, when we referred to it in the worker pool specification, we replace the directory slash with a dot (`trainer.task`) and dropped the file suffix (`.py`).\n\n#### Package Assembly\n\nIn the following cells, you will assemble the training package.", "_____no_output_____" ] ], [ [ "# Make folder for Python training script\n! rm -rf custom\n! mkdir custom\n\n# Add package information\n! touch custom/README.md\n\nsetup_cfg = \"[egg_info]\\n\\ntag_build =\\n\\ntag_date = 0\"\n! echo \"$setup_cfg\" > custom/setup.cfg\n\nsetup_py = \"import setuptools\\n\\nsetuptools.setup(\\n\\n install_requires=[\\n\\n 'tensorflow_datasets==1.3.0',\\n\\n ],\\n\\n packages=setuptools.find_packages())\"\n! echo \"$setup_py\" > custom/setup.py\n\npkg_info = \"Metadata-Version: 1.0\\n\\nName: CIFAR10 image classification\\n\\nVersion: 0.0.0\\n\\nSummary: Demostration training script\\n\\nHome-page: www.google.com\\n\\nAuthor: Google\\n\\nAuthor-email: [email protected]\\n\\nLicense: Public\\n\\nDescription: Demo\\n\\nPlatform: Vertex\"\n! echo \"$pkg_info\" > custom/PKG-INFO\n\n# Make the training subfolder\n! mkdir custom/trainer\n! touch custom/trainer/__init__.py", "_____no_output_____" ] ], [ [ "#### Task.py contents\n\nIn the next cell, you write the contents of the training script task.py. We won't go into detail, it's just there for you to browse. In summary:\n\n- Get the directory where to save the model artifacts from the command line (`--model_dir`), and if not specified, then from the environment variable `AIP_MODEL_DIR`.\n- Loads CIFAR10 dataset from TF Datasets (tfds).\n- Builds a model using TF.Keras model API.\n- Compiles the model (`compile()`).\n- Sets a training distribution strategy according to the argument `args.distribute`.\n- Trains the model (`fit()`) with epochs and steps according to the arguments `args.epochs` and `args.steps`\n- Saves the trained model (`save(args.model_dir)`) to the specified model directory.", "_____no_output_____" ] ], [ [ "%%writefile custom/trainer/task.py\n# Single, Mirror and Multi-Machine Distributed Training for CIFAR-10\n\nimport tensorflow_datasets as tfds\nimport tensorflow as tf\nfrom tensorflow.python.client import device_lib\nimport argparse\nimport os\nimport sys\ntfds.disable_progress_bar()\n\nparser = argparse.ArgumentParser()\nparser.add_argument('--model-dir', dest='model_dir',\n default=os.getenv(\"AIP_MODEL_DIR\"), type=str, help='Model dir.')\nparser.add_argument('--lr', dest='lr',\n default=0.01, type=float,\n help='Learning rate.')\nparser.add_argument('--epochs', dest='epochs',\n default=10, type=int,\n help='Number of epochs.')\nparser.add_argument('--steps', dest='steps',\n default=200, type=int,\n help='Number of steps per epoch.')\nparser.add_argument('--distribute', dest='distribute', type=str, default='single',\n help='distributed training strategy')\nargs = parser.parse_args()\n\nprint('Python Version = {}'.format(sys.version))\nprint('TensorFlow Version = {}'.format(tf.__version__))\nprint('TF_CONFIG = {}'.format(os.environ.get('TF_CONFIG', 'Not found')))\nprint('DEVICES', device_lib.list_local_devices())\n\n# Single Machine, single compute device\nif args.distribute == 'single':\n if tf.test.is_gpu_available():\n strategy = tf.distribute.OneDeviceStrategy(device=\"/gpu:0\")\n else:\n strategy = tf.distribute.OneDeviceStrategy(device=\"/cpu:0\")\n# Single Machine, multiple compute device\nelif args.distribute == 'mirror':\n strategy = tf.distribute.MirroredStrategy()\n# Multiple Machine, multiple compute device\nelif args.distribute == 'multi':\n strategy = tf.distribute.experimental.MultiWorkerMirroredStrategy()\n\n# Multi-worker configuration\nprint('num_replicas_in_sync = {}'.format(strategy.num_replicas_in_sync))\n\n# Preparing dataset\nBUFFER_SIZE = 10000\nBATCH_SIZE = 64\n\n\ndef make_datasets_unbatched():\n\n # Scaling CIFAR10 data from (0, 255] to (0., 1.]\n def scale(image, label):\n image = tf.cast(image, tf.float32)\n image /= 255.0\n return image, label\n\n\n datasets, info = tfds.load(name='cifar10',\n with_info=True,\n as_supervised=True)\n return datasets['train'].map(scale).cache().shuffle(BUFFER_SIZE).repeat()\n\n\n# Build the Keras model\ndef build_and_compile_cnn_model():\n model = tf.keras.Sequential([\n tf.keras.layers.Conv2D(32, 3, activation='relu', input_shape=(32, 32, 3)),\n tf.keras.layers.MaxPooling2D(),\n tf.keras.layers.Conv2D(32, 3, activation='relu'),\n tf.keras.layers.MaxPooling2D(),\n tf.keras.layers.Flatten(),\n tf.keras.layers.Dense(10, activation='softmax')\n ])\n model.compile(\n loss=tf.keras.losses.sparse_categorical_crossentropy,\n optimizer=tf.keras.optimizers.SGD(learning_rate=args.lr),\n metrics=['accuracy'])\n return model\n\n\n# Train the model\nNUM_WORKERS = strategy.num_replicas_in_sync\n# Here the batch size scales up by number of workers since\n# `tf.data.Dataset.batch` expects the global batch size.\nGLOBAL_BATCH_SIZE = BATCH_SIZE * NUM_WORKERS\ntrain_dataset = make_datasets_unbatched().batch(GLOBAL_BATCH_SIZE)\n\nwith strategy.scope():\n # Creation of dataset, and model building/compiling need to be within\n # `strategy.scope()`.\n model = build_and_compile_cnn_model()\n\nmodel.fit(x=train_dataset, epochs=args.epochs, steps_per_epoch=args.steps)\nmodel.save(args.model_dir)", "_____no_output_____" ] ], [ [ "#### Store training script on your Cloud Storage bucket\n\nNext, you package the training folder into a compressed tar ball, and then store it in your Cloud Storage bucket.", "_____no_output_____" ] ], [ [ "! rm -f custom.tar custom.tar.gz\n! tar cvf custom.tar custom\n! gzip custom.tar\n! gsutil cp custom.tar.gz $BUCKET_NAME/trainer_cifar10.tar.gz", "_____no_output_____" ] ], [ [ "### Create and run custom training job\n\n\nTo train a custom model, you perform two steps: 1) create a custom training job, and 2) run the job.\n\n#### Create custom training job\n\nA custom training job is created with the `CustomTrainingJob` class, with the following parameters:\n\n- `display_name`: The human readable name for the custom training job.\n- `container_uri`: The training container image.\n- `requirements`: Package requirements for the training container image (e.g., pandas).\n- `script_path`: The relative path to the training script.", "_____no_output_____" ] ], [ [ "job = aip.CustomTrainingJob(\n display_name=\"cifar10_\" + TIMESTAMP,\n script_path=\"custom/trainer/task.py\",\n container_uri=TRAIN_IMAGE,\n requirements=[\"gcsfs==0.7.1\", \"tensorflow-datasets==4.4\"],\n)\n\nprint(job)", "_____no_output_____" ] ], [ [ "### Prepare your command-line arguments\n\nNow define the command-line arguments for your custom training container:\n\n- `args`: The command-line arguments to pass to the executable that is set as the entry point into the container.\n - `--model-dir` : For our demonstrations, we use this command-line argument to specify where to store the model artifacts.\n - direct: You pass the Cloud Storage location as a command line argument to your training script (set variable `DIRECT = True`), or\n - indirect: The service passes the Cloud Storage location as the environment variable `AIP_MODEL_DIR` to your training script (set variable `DIRECT = False`). In this case, you tell the service the model artifact location in the job specification.\n - `\"--epochs=\" + EPOCHS`: The number of epochs for training.\n - `\"--steps=\" + STEPS`: The number of steps per epoch.", "_____no_output_____" ] ], [ [ "MODEL_DIR = \"{}/{}\".format(BUCKET_NAME, TIMESTAMP)\n\nEPOCHS = 20\nSTEPS = 100\n\nDIRECT = True\nif DIRECT:\n CMDARGS = [\n \"--model-dir=\" + MODEL_DIR,\n \"--epochs=\" + str(EPOCHS),\n \"--steps=\" + str(STEPS),\n ]\nelse:\n CMDARGS = [\n \"--epochs=\" + str(EPOCHS),\n \"--steps=\" + str(STEPS),\n ]", "_____no_output_____" ] ], [ [ "#### Run the custom training job\n\nNext, you run the custom job to start the training job by invoking the method `run`, with the following parameters:\n\n- `args`: The command-line arguments to pass to the training script.\n- `replica_count`: The number of compute instances for training (replica_count = 1 is single node training).\n- `machine_type`: The machine type for the compute instances.\n- `accelerator_type`: The hardware accelerator type.\n- `accelerator_count`: The number of accelerators to attach to a worker replica.\n- `base_output_dir`: The Cloud Storage location to write the model artifacts to.\n- `sync`: Whether to block until completion of the job.", "_____no_output_____" ] ], [ [ "if TRAIN_GPU:\n job.run(\n args=CMDARGS,\n replica_count=1,\n machine_type=TRAIN_COMPUTE,\n accelerator_type=TRAIN_GPU.name,\n accelerator_count=TRAIN_NGPU,\n base_output_dir=MODEL_DIR,\n sync=True,\n )\nelse:\n job.run(\n args=CMDARGS,\n replica_count=1,\n machine_type=TRAIN_COMPUTE,\n base_output_dir=MODEL_DIR,\n sync=True,\n )\n\nmodel_path_to_deploy = MODEL_DIR", "_____no_output_____" ] ], [ [ "## Load the saved model\n\nYour model is stored in a TensorFlow SavedModel format in a Cloud Storage bucket. Now load it from the Cloud Storage bucket, and then you can do some things, like evaluate the model, and do a prediction.\n\nTo load, you use the TF.Keras `model.load_model()` method passing it the Cloud Storage path where the model is saved -- specified by `MODEL_DIR`.", "_____no_output_____" ] ], [ [ "import tensorflow as tf\n\nlocal_model = tf.keras.models.load_model(MODEL_DIR)", "_____no_output_____" ] ], [ [ "## Evaluate the model\n\nNow find out how good the model is.\n\n### Load evaluation data\n\nYou will load the CIFAR10 test (holdout) data from `tf.keras.datasets`, using the method `load_data()`. This returns the dataset as a tuple of two elements. The first element is the training data and the second is the test data. Each element is also a tuple of two elements: the image data, and the corresponding labels.\n\nYou don't need the training data, and hence why we loaded it as `(_, _)`.\n\nBefore you can run the data through evaluation, you need to preprocess it:\n\n`x_test`:\n1. Normalize (rescale) the pixel data by dividing each pixel by 255. This replaces each single byte integer pixel with a 32-bit floating point number between 0 and 1.\n\n`y_test`:<br/>\n2. The labels are currently scalar (sparse). If you look back at the `compile()` step in the `trainer/task.py` script, you will find that it was compiled for sparse labels. So we don't need to do anything more.", "_____no_output_____" ] ], [ [ "import numpy as np\nfrom tensorflow.keras.datasets import cifar10\n\n(_, _), (x_test, y_test) = cifar10.load_data()\nx_test = (x_test / 255.0).astype(np.float32)\n\nprint(x_test.shape, y_test.shape)", "_____no_output_____" ] ], [ [ "### Perform the model evaluation\n\nNow evaluate how well the model in the custom job did.", "_____no_output_____" ] ], [ [ "local_model.evaluate(x_test, y_test)", "_____no_output_____" ] ], [ [ "### Serving function for image data\n\nTo pass images to the prediction service, you encode the compressed (e.g., JPEG) image bytes into base 64 -- which makes the content safe from modification while transmitting binary data over the network. Since this deployed model expects input data as raw (uncompressed) bytes, you need to ensure that the base 64 encoded data gets converted back to raw bytes before it is passed as input to the deployed model.\n\nTo resolve this, define a serving function (`serving_fn`) and attach it to the model as a preprocessing step. Add a `@tf.function` decorator so the serving function is fused to the underlying model (instead of upstream on a CPU).\n\nWhen you send a prediction or explanation request, the content of the request is base 64 decoded into a Tensorflow string (`tf.string`), which is passed to the serving function (`serving_fn`). The serving function preprocesses the `tf.string` into raw (uncompressed) numpy bytes (`preprocess_fn`) to match the input requirements of the model:\n- `io.decode_jpeg`- Decompresses the JPG image which is returned as a Tensorflow tensor with three channels (RGB).\n- `image.convert_image_dtype` - Changes integer pixel values to float 32.\n- `image.resize` - Resizes the image to match the input shape for the model.\n- `resized / 255.0` - Rescales (normalization) the pixel data between 0 and 1.\n\nAt this point, the data can be passed to the model (`m_call`).", "_____no_output_____" ] ], [ [ "CONCRETE_INPUT = \"numpy_inputs\"\n\n\ndef _preprocess(bytes_input):\n decoded = tf.io.decode_jpeg(bytes_input, channels=3)\n decoded = tf.image.convert_image_dtype(decoded, tf.float32)\n resized = tf.image.resize(decoded, size=(32, 32))\n rescale = tf.cast(resized / 255.0, tf.float32)\n return rescale\n\n\[email protected](input_signature=[tf.TensorSpec([None], tf.string)])\ndef preprocess_fn(bytes_inputs):\n decoded_images = tf.map_fn(\n _preprocess, bytes_inputs, dtype=tf.float32, back_prop=False\n )\n return {\n CONCRETE_INPUT: decoded_images\n } # User needs to make sure the key matches model's input\n\n\[email protected](input_signature=[tf.TensorSpec([None], tf.string)])\ndef serving_fn(bytes_inputs):\n images = preprocess_fn(bytes_inputs)\n prob = m_call(**images)\n return prob\n\n\nm_call = tf.function(local_model.call).get_concrete_function(\n [tf.TensorSpec(shape=[None, 32, 32, 3], dtype=tf.float32, name=CONCRETE_INPUT)]\n)\n\ntf.saved_model.save(\n local_model, model_path_to_deploy, signatures={\"serving_default\": serving_fn}\n)", "_____no_output_____" ] ], [ [ "## Get the serving function signature\n\nYou can get the signatures of your model's input and output layers by reloading the model into memory, and querying it for the signatures corresponding to each layer.\n\nFor your purpose, you need the signature of the serving function. Why? Well, when we send our data for prediction as a HTTP request packet, the image data is base64 encoded, and our TF.Keras model takes numpy input. Your serving function will do the conversion from base64 to a numpy array.\n\nWhen making a prediction request, you need to route the request to the serving function instead of the model, so you need to know the input layer name of the serving function -- which you will use later when you make a prediction request.", "_____no_output_____" ] ], [ [ "loaded = tf.saved_model.load(model_path_to_deploy)\n\nserving_input = list(\n loaded.signatures[\"serving_default\"].structured_input_signature[1].keys()\n)[0]\nprint(\"Serving function input:\", serving_input)", "_____no_output_____" ] ], [ [ "## Upload the model\n\nNext, upload your model to a `Model` resource using `Model.upload()` method, with the following parameters:\n\n- `display_name`: The human readable name for the `Model` resource.\n- `artifact`: The Cloud Storage location of the trained model artifacts.\n- `serving_container_image_uri`: The serving container image.\n- `sync`: Whether to execute the upload asynchronously or synchronously.\n\nIf the `upload()` method is run asynchronously, you can subsequently block until completion with the `wait()` method.", "_____no_output_____" ] ], [ [ "model = aip.Model.upload(\n display_name=\"cifar10_\" + TIMESTAMP,\n artifact_uri=MODEL_DIR,\n serving_container_image_uri=DEPLOY_IMAGE,\n sync=False,\n)\n\nmodel.wait()", "_____no_output_____" ] ], [ [ "## Deploy the model\n\nNext, deploy your model for online prediction. To deploy the model, you invoke the `deploy` method, with the following parameters:\n\n- `deployed_model_display_name`: A human readable name for the deployed model.\n- `traffic_split`: Percent of traffic at the endpoint that goes to this model, which is specified as a dictionary of one or more key/value pairs.\nIf only one model, then specify as { \"0\": 100 }, where \"0\" refers to this model being uploaded and 100 means 100% of the traffic.\nIf there are existing models on the endpoint, for which the traffic will be split, then use model_id to specify as { \"0\": percent, model_id: percent, ... }, where model_id is the model id of an existing model to the deployed endpoint. The percents must add up to 100.\n- `machine_type`: The type of machine to use for training.\n- `accelerator_type`: The hardware accelerator type.\n- `accelerator_count`: The number of accelerators to attach to a worker replica.\n- `starting_replica_count`: The number of compute instances to initially provision.\n- `max_replica_count`: The maximum number of compute instances to scale to. In this tutorial, only one instance is provisioned.", "_____no_output_____" ] ], [ [ "DEPLOYED_NAME = \"cifar10-\" + TIMESTAMP\n\nTRAFFIC_SPLIT = {\"0\": 100}\n\nMIN_NODES = 1\nMAX_NODES = 1\n\nif DEPLOY_GPU:\n endpoint = model.deploy(\n deployed_model_display_name=DEPLOYED_NAME,\n traffic_split=TRAFFIC_SPLIT,\n machine_type=DEPLOY_COMPUTE,\n accelerator_type=DEPLOY_GPU,\n accelerator_count=DEPLOY_NGPU,\n min_replica_count=MIN_NODES,\n max_replica_count=MAX_NODES,\n )\nelse:\n endpoint = model.deploy(\n deployed_model_display_name=DEPLOYED_NAME,\n traffic_split=TRAFFIC_SPLIT,\n machine_type=DEPLOY_COMPUTE,\n accelerator_type=DEPLOY_GPU,\n accelerator_count=0,\n min_replica_count=MIN_NODES,\n max_replica_count=MAX_NODES,\n )", "_____no_output_____" ] ], [ [ "### Get test item\n\nYou will use an example out of the test (holdout) portion of the dataset as a test item.", "_____no_output_____" ] ], [ [ "test_image = x_test[0]\ntest_label = y_test[0]\nprint(test_image.shape)", "_____no_output_____" ] ], [ [ "### Prepare the request content\nYou are going to send the CIFAR10 image as compressed JPG image, instead of the raw uncompressed bytes:\n\n- `cv2.imwrite`: Use openCV to write the uncompressed image to disk as a compressed JPEG image.\n - Denormalize the image data from \\[0,1) range back to [0,255).\n - Convert the 32-bit floating point values to 8-bit unsigned integers.\n- `tf.io.read_file`: Read the compressed JPG images back into memory as raw bytes.\n- `base64.b64encode`: Encode the raw bytes into a base 64 encoded string.", "_____no_output_____" ] ], [ [ "import base64\n\nimport cv2\n\ncv2.imwrite(\"tmp.jpg\", (test_image * 255).astype(np.uint8))\n\nbytes = tf.io.read_file(\"tmp.jpg\")\nb64str = base64.b64encode(bytes.numpy()).decode(\"utf-8\")", "_____no_output_____" ] ], [ [ "### Make the prediction\n\nNow that your `Model` resource is deployed to an `Endpoint` resource, you can do online predictions by sending prediction requests to the Endpoint resource.\n\n#### Request\n\nSince in this example your test item is in a Cloud Storage bucket, you open and read the contents of the image using `tf.io.gfile.Gfile()`. To pass the test data to the prediction service, you encode the bytes into base64 -- which makes the content safe from modification while transmitting binary data over the network.\n\nThe format of each instance is:\n\n { serving_input: { 'b64': base64_encoded_bytes } }\n\nSince the `predict()` method can take multiple items (instances), send your single test item as a list of one test item.\n\n#### Response\n\nThe response from the `predict()` call is a Python dictionary with the following entries:\n\n- `ids`: The internal assigned unique identifiers for each prediction request.\n- `predictions`: The predicted confidence, between 0 and 1, per class label.\n- `deployed_model_id`: The Vertex AI identifier for the deployed `Model` resource which did the predictions.", "_____no_output_____" ] ], [ [ "# The format of each instance should conform to the deployed model's prediction input schema.\ninstances = [{serving_input: {\"b64\": b64str}}]\n\nprediction = endpoint.predict(instances=instances)\n\nprint(prediction)", "_____no_output_____" ] ], [ [ "## Undeploy the model\n\nWhen you are done doing predictions, you undeploy the model from the `Endpoint` resouce. This deprovisions all compute resources and ends billing for the deployed model.", "_____no_output_____" ] ], [ [ "endpoint.undeploy_all()", "_____no_output_____" ] ], [ [ "# Cleaning up\n\nTo clean up all Google Cloud resources used in this project, you can [delete the Google Cloud\nproject](https://cloud.google.com/resource-manager/docs/creating-managing-projects#shutting_down_projects) you used for the tutorial.\n\nOtherwise, you can delete the individual resources you created in this tutorial:\n\n- Dataset\n- Pipeline\n- Model\n- Endpoint\n- AutoML Training Job\n- Batch Job\n- Custom Job\n- Hyperparameter Tuning Job\n- Cloud Storage Bucket", "_____no_output_____" ] ], [ [ "delete_all = True\n\nif delete_all:\n # Delete the dataset using the Vertex dataset object\n try:\n if \"dataset\" in globals():\n dataset.delete()\n except Exception as e:\n print(e)\n\n # Delete the model using the Vertex model object\n try:\n if \"model\" in globals():\n model.delete()\n except Exception as e:\n print(e)\n\n # Delete the endpoint using the Vertex endpoint object\n try:\n if \"endpoint\" in globals():\n endpoint.delete()\n except Exception as e:\n print(e)\n\n # Delete the AutoML or Pipeline trainig job\n try:\n if \"dag\" in globals():\n dag.delete()\n except Exception as e:\n print(e)\n\n # Delete the custom trainig job\n try:\n if \"job\" in globals():\n job.delete()\n except Exception as e:\n print(e)\n\n # Delete the batch prediction job using the Vertex batch prediction object\n try:\n if \"batch_predict_job\" in globals():\n batch_predict_job.delete()\n except Exception as e:\n print(e)\n\n # Delete the hyperparameter tuning job using the Vertex hyperparameter tuning object\n try:\n if \"hpt_job\" in globals():\n hpt_job.delete()\n except Exception as e:\n print(e)\n\n if \"BUCKET_NAME\" in globals():\n ! gsutil rm -r $BUCKET_NAME", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

ec7a23ba73e7d9b63ffb500f52ae35a61fe15fad

20,151

ipynb

Jupyter Notebook

content/dod_sql_21.ipynb

vass1138/SQLForExcelUsersADS

28e687008a00666839ccc7c9d3d1062890d656d1

[ "MIT" ]

null

content/dod_sql_21.ipynb

vass1138/SQLForExcelUsersADS

28e687008a00666839ccc7c9d3d1062890d656d1

[ "MIT" ]

null

content/dod_sql_21.ipynb

vass1138/SQLForExcelUsersADS

28e687008a00666839ccc7c9d3d1062890d656d1

[ "MIT" ]

1

2020-11-27T05:38:03.000Z

45.283146

205

0.461218

[ [ [ "empty" ] ] ]

[ "empty" ]

[ [ "empty" ] ]

ec7a2f723fe00b1976f673cdc73de9b963053866

252,195

ipynb

Jupyter Notebook

source/lesson02/script.ipynb

psteinb/deeplearning540.github.io

1186266f187e7529423c0a1ecdaf2d9949074b55

[ "MIT" ]

6

2021-03-01T07:00:32.000Z

2022-03-23T15:33:53.000Z

source/lesson02/script.ipynb

psteinb/deeplearning540.github.io

1186266f187e7529423c0a1ecdaf2d9949074b55

[ "MIT" ]

17

2021-02-22T13:16:37.000Z

2022-03-23T16:42:47.000Z

source/lesson02/script.ipynb

psteinb/deeplearning540.github.io

1186266f187e7529423c0a1ecdaf2d9949074b55

[ "MIT" ]

14

2021-02-22T11:38:28.000Z

2022-01-12T15:08:48.000Z

354.704641

126,296

0.927033

[ [ [ "# Lesson 02, Part 1: Machine Learning and our first look into data\n\nMachine Learning is divided into three main categories:\n\n- supervised learning\n- unsupervised learning\n- reinforcement learning\n\n", "_____no_output_____" ], [ "## Supervised Learning\n\n- all data used is labelled (with ground truth information)\n- the algorithm is provided direct feedback\n- the algorithm is meant to predict outcome\n", "_____no_output_____" ], [ "### Classification\n\n<p><a href=\"https://commons.wikimedia.org/wiki/File:Svm_separating_hyperplanes.png#/media/File:Svm_separating_hyperplanes.png\"><img src=\"https://upload.wikimedia.org/wikipedia/commons/2/20/Svm_separating_hyperplanes.png\" alt=\"Svm separating hyperplanes.png\" width=\"503\" height=\"480\"></a><br>By <a href=\"//commons.wikimedia.org/w/index.php?title=User:Cyc&amp;action=edit&amp;redlink=1\" class=\"new\" title=\"User:Cyc (page does not exist)\">Cyc</a> - <span class=\"int-own-work\" lang=\"en\">Own work</span>, Public Domain, <a href=\"https://commons.wikimedia.org/w/index.php?curid=3566969\">Link</a></p>\n", "_____no_output_____" ], [ "### Regression\n\n<p><a href=\"https://commons.wikimedia.org/wiki/File:Linear_regression.svg#/media/File:Linear_regression.svg\"><img src=\"https://upload.wikimedia.org/wikipedia/commons/thumb/3/3a/Linear_regression.svg/640px-Linear_regression.svg.png\" alt=\"Linear regression.svg\"></a><br>By <a href=\"//commons.wikimedia.org/w/index.php?title=User:Sewaqu&amp;action=edit&amp;redlink=1\" class=\"new\" title=\"User:Sewaqu (page does not exist)\">Sewaqu</a> - <span class=\"int-own-work\" lang=\"en\">Own work</span>, Public Domain, <a href=\"https://commons.wikimedia.org/w/index.php?curid=11967659\">Link</a></p>\n\n", "_____no_output_____" ], [ "## Unsupervised Learning\n\n- there are **NO** labels (with ground truth information)\n- no feedback is provided to the algorithm\n- goal: find hidden structure in data\n\n<p><a href=\"https://commons.wikimedia.org/wiki/File:KMeans-Gaussian-data.svg#/media/File:KMeans-Gaussian-data.svg\"><img src=\"https://upload.wikimedia.org/wikipedia/commons/thumb/e/e5/KMeans-Gaussian-data.svg/1200px-KMeans-Gaussian-data.svg.png\" alt=\"KMeans-Gaussian-data.svg\"></a><br>By <a href=\"//commons.wikimedia.org/wiki/User:Chire\" title=\"User:Chire\">Chire</a> - <span class=\"int-own-work\" lang=\"en\">Own work</span>, <a href=\"https://creativecommons.org/publicdomain/zero/1.0/deed.en\" title=\"Creative Commons CC0 1.0 Universal Public Domain Dedication\">CC0 1.0</a>, <a href=\"\">Link</a></p>", "_____no_output_____" ], [ "## Reinforcement Learning\n\n- model a decision process\n- reward system\n- learn series of actions based on reward\n\n<p><a href=\"https://upload.wikimedia.org/wikipedia/commons/1/1b/Reinforcement_learning_diagram.svg\"><img src=\"https://upload.wikimedia.org/wikipedia/commons/thumb/1/1b/Reinforcement_learning_diagram.svg/794px-Reinforcement_learning_diagram.svg.png\" alt=\"Reinforcement learning diagram.svg\" width=\"800\"></a><br>By <a href=\"//commons.wikimedia.org/wiki/User:Megajuice\" title=\"User:Megajuice\">Megajuice</a> - <span class=\"int-own-work\" lang=\"en\">Own work</span>, <a href=\"https://creativecommons.org/publicdomain/zero/1.0/deed.en\" title=\"\">CC0 1.0</a></p>", "_____no_output_____" ], [ "# Important Notation \n\n- we are given a dataset of size $n$ as \n$\\mathcal{D} = \\{ \\langle \\vec{x}, y \\rangle_{i}, i = 1, \\dots, n \\} $\n\n- the data represents a mapping: \n$f(\\vec{x}) = y$\n\n- machine learning produces a hypothesis (or a prediction): \n$h(\\vec{x}) = \\hat{y}$\n\n## classification versus regression\n\n- classification: \n$h : \\mathcal{R}^n \\rightarrow \\mathcal{Z} $ \n(e.g. for 3 categories $\\{0,1,2\\}$)\n\n- regression: \n$h : \\mathcal{R}^n \\rightarrow \\mathcal{R} $ (regression can also produce $\\mathcal{R}^{n}$) \n", "_____no_output_____" ], [ "# Data\n\nFor the following, I will rely on the Palmer penguin dataset obtained from [this repo](https://github.com/allisonhorst/palmerpenguins). To quote the repo:\n\n> Data were collected and made available by [Dr. Kristen Gorman](https://www.uaf.edu/cfos/people/faculty/detail/kristen-gorman.php)\n> and the [Palmer Station, Antarctica LTER](https://pal.lternet.edu/), a member of the [Long Term Ecological Research Network](https://lternet.edu/).\n", "_____no_output_____" ] ], [ [ "import pandas as pd\nprint(\"pandas version:\", pd.__version__)\n", "pandas version: 1.2.4\n" ], [ "df = pd.read_csv(\"https://raw.githubusercontent.com/allisonhorst/palmerpenguins/master/inst/extdata/penguins.csv\")\nprint(df.head())\nprint(df.tail())", " species island bill_length_mm bill_depth_mm flipper_length_mm \\\n0 Adelie Torgersen 39.1 18.7 181.0 \n1 Adelie Torgersen 39.5 17.4 186.0 \n2 Adelie Torgersen 40.3 18.0 195.0 \n3 Adelie Torgersen NaN NaN NaN \n4 Adelie Torgersen 36.7 19.3 193.0 \n\n body_mass_g sex year \n0 3750.0 male 2007 \n1 3800.0 female 2007 \n2 3250.0 female 2007 \n3 NaN NaN 2007 \n4 3450.0 female 2007 \n species island bill_length_mm bill_depth_mm flipper_length_mm \\\n339 Chinstrap Dream 55.8 19.8 207.0 \n340 Chinstrap Dream 43.5 18.1 202.0 \n341 Chinstrap Dream 49.6 18.2 193.0 \n342 Chinstrap Dream 50.8 19.0 210.0 \n343 Chinstrap Dream 50.2 18.7 198.0 \n\n body_mass_g sex year \n339 4000.0 male 2009 \n340 3400.0 female 2009 \n341 3775.0 male 2009 \n342 4100.0 male 2009 \n343 3775.0 female 2009 \n" ], [ "#let's remove the rows with NaN values\ndf = df.dropna()\n\nprint(\"after cleaning:\")\nprint(df.head())\n", "after cleaning:\n species island bill_length_mm bill_depth_mm flipper_length_mm \\\n0 Adelie Torgersen 39.1 18.7 181.0 \n1 Adelie Torgersen 39.5 17.4 186.0 \n2 Adelie Torgersen 40.3 18.0 195.0 \n4 Adelie Torgersen 36.7 19.3 193.0 \n5 Adelie Torgersen 39.3 20.6 190.0 \n\n body_mass_g sex year \n0 3750.0 male 2007 \n1 3800.0 female 2007 \n2 3250.0 female 2007 \n4 3450.0 female 2007 \n5 3650.0 male 2007 \n" ], [ "print(df.shape)\nprint(df.dtypes)\nprint(\">> columns species, island and sex are encoded as strings in the data (therefor we see them as object class)!\")", "(333, 8)\nspecies object\nisland object\nbill_length_mm float64\nbill_depth_mm float64\nflipper_length_mm float64\nbody_mass_g float64\nsex object\nyear int64\ndtype: object\n>> columns species, island and sex are encoded as strings in the data (therefor we see them as object class)!\n" ], [ "df[[\"species_\"]] = df[[\"species\"]].astype(\"category\")\nprint(\"our categories\\n\", df[[\"species_\"]].head(),\"\\n\")\nprint(\"encoded as numbers\\n\", df.species_.cat.codes.head())", "our categories\n species_\n0 Adelie\n1 Adelie\n2 Adelie\n4 Adelie\n5 Adelie \n\nencoded as numbers\n 0 0\n1 0\n2 0\n4 0\n5 0\ndtype: int8\n" ], [ "print(\"make sure the categories are correctly encoded:\\n\")\nprint(pd.unique(df.species_))\n", "make sure the categories are correctly encoded:\n\n['Adelie', 'Gentoo', 'Chinstrap']\nCategories (3, object): ['Adelie', 'Gentoo', 'Chinstrap']\n" ], [ "print(pd.unique(df.species_.cat.codes))", "[0 2 1]\n" ], [ "print(df.species_.value_counts())", "Adelie 146\nGentoo 119\nChinstrap 68\nName: species_, dtype: int64\n" ] ], [ [ "There are 3 types of penguins in this dataset: \n\n![Artwork by @allison_horst](https://raw.githubusercontent.com/allisonhorst/palmerpenguins/master/man/figures/lter_penguins.png)\n\n", "_____no_output_____" ], [ "Important feature columns are:\n\n![Artwork by @allison_horst](https://raw.githubusercontent.com/allisonhorst/palmerpenguins/master/man/figures/culmen_depth.png)", "_____no_output_____" ] ], [ [ "import matplotlib.pyplot as plt\n\n#only use for dark theme\nplt.style.use('dark_background')", "_____no_output_____" ], [ "#to make out lives a bit easier, let's use a wrapper library for matplotlib\nimport seaborn as sns\nprint(f'seaborn version: {sns.__version__}')", "seaborn version: 0.11.1\n" ], [ "#we want to cluster the penguins and work towards a classification using this\nsns.scatterplot(x=df.bill_length_mm, y=df.flipper_length_mm, palette=\"husl\")\nplt.show()", "_____no_output_____" ] ], [ [ "\n# Conclusions\n\n- machine learning can be divided into 3 main categories: supervised learning, unsupervised learning, reinforcement learning\n- supervised learning: classification or regression\n- unsupervised learning: uncovering hidden structures\n- reinforcement learning: training a sequence of actions (given a reward)\n\n- all driven by data:\n + open datasets for the win\n + check your data before you train (`dropna`, `shape`, `dtypes`, ...)", "_____no_output_____" ], [ "# lesson 02, part 2: Enter Nearest Neighbors Clustering\n\n## Problem\n\n- we already know that our dataset consists of 3 classes: `Adelie, Gentoo, Chinstrap`\n- let's assume with **DO NOT** have the correct class label for each row\n", "_____no_output_____" ], [ "## Task \n\n- given this dataset \n$\\mathcal{D} = \\{ \\langle \\vec{x} \\rangle_{i}, i = 1, \\dots, n \\} $\n- find the $k=3$ clusters to which any of the known points belong to!\n\n", "_____no_output_____" ], [ "## Analysis\n\n- basis algorithm: finding the nearest point for a query `x_q` given a reference `dataset`\n \n```\ndef bruteforce_nearest_neighbor( x_q, dataset):\n closest_point = None\n closest_distance = infinity\n \n for i in range(n):\n x_i = dataset[i]\n current_distance = distance(x_i, x_q)\n if current_distance < closest_distance:\n closest_distance = current_distance\n closest_point = x_i\n \n return closest_point\n```\n\n- most common distance metric: Euclidean Distance $d(\\vec{x}_a, \\vec{x}_b) = \\sqrt{ \\sum_{j=1}^{m} (x_{j,a} - x_{j,b})^2 }$\n- price: \n + keep entire \"training set\" in memory\n + for each query point, go through entire dataset again (`bruteforce`)\n ", "_____no_output_____" ], [ "### Naive Clustering / Llyod's algorithm\n\ngoal: create $k$ sets $S$ such as $argmin_{S} \\sum^{k}_{i=1} \\sum_{\\vec{x} \\in S_i} || \\vec{x} - \\vec{\\mu}_i ||^2$\n\n\nalgorithm:\n\n1. select $k$ points at random and assign them a cluster_id\n (consider these points to be the __mean__ of the cluster)\n \n![by Weston.pace, from commons.wikimedia.org under CC-BY-SA 3.0](https://upload.wikimedia.org/wikipedia/commons/thumb/5/5e/K_Means_Example_Step_1.svg/249px-K_Means_Example_Step_1.svg.png)\n \n2. assign samples closest to a given cluster mean/centroid so that the variance of the cluster remains minimal\n\n![by Weston.pace, from commons.wikimedia.org under CC-BY-SA 3.0](https://upload.wikimedia.org/wikipedia/commons/thumb/a/a5/K_Means_Example_Step_2.svg/278px-K_Means_Example_Step_2.svg.png)\n\n3. calculate the distance of all points to those cluster means (also called centroids) and update the mean cluster centroid for each cluster\n\n![by Weston.pace, from commons.wikimedia.org under CC-BY-SA 3.0](https://upload.wikimedia.org/wikipedia/commons/thumb/3/3e/K_Means_Example_Step_3.svg/278px-K_Means_Example_Step_3.svg.png)\n\n4. Steps 2 and 3 are repeated until convergence has been reached, i.e. the cluster association does not change anymore.\n\n![by Weston.pace, from commons.wikimedia.org under CC-BY-SA 3.0](https://upload.wikimedia.org/wikipedia/commons/thumb/d/d2/K_Means_Example_Step_4.svg/278px-K_Means_Example_Step_4.svg.png)", "_____no_output_____" ] ], [ [ "import sklearn.cluster as skl_cluster\nimport numpy as np\n\nkmeans = skl_cluster.KMeans(n_clusters=3, init='random')\n\ndata = np.stack((df.bill_length_mm, df.flipper_length_mm), axis=-1)\n\nkmeans = kmeans.fit(data)\n", "_____no_output_____" ], [ "#let's see which clusters have been found\nfig, ax = plt.subplots(1, figsize=(12,8))\nsns.scatterplot(ax=ax,\n x=df.bill_length_mm,\n y=df.flipper_length_mm,\n hue=kmeans.labels_,\n palette=\"Set2\")\n\nkmeans.cluster_centers_", "_____no_output_____" ], [ "#let's compare to the real labels\nfig, (left, right) = plt.subplots(1,2,figsize=(16,6), sharey = True, sharex=False)\n\nsns.scatterplot(ax=left,\n x=df.bill_length_mm,\n y=df.flipper_length_mm,\n hue=kmeans.labels_,\n palette=\"Set2\")\nleft = left.set_title(\"kmeans prediction\")\n\nsns.scatterplot(ax=right,\n x=df.bill_length_mm,\n y=df.flipper_length_mm,\n hue=df.species_,\n palette=\"Set2\")\nright = right.set_title(\"ground truth\")\n\n", "_____no_output_____" ] ], [ [ "# Conclusions\n\n- unsupervised learning should be used to uncover hidden structure in data sets\n - it is useful if true labels/annotations are not available\n - kmeans clustering is one of the most widespread unsupervised algorithms available\n \n- kmeans clustering works by creating cluster associations that minimize the cluster variance (similar to least squares regression)\n - if separation of clusters is not clear (they overlap), kmeans struggles\n - intelligent search algorithms are put to use to accelerate kmeans (among other aspects)", "_____no_output_____" ], [ "# Further Reading\n\n- some parts of this material were inspired by the ever awesome [Sebastian Raschka](https://sebastianraschka.com)\n + general overview of machine learning [lesson 01](https://sebastianraschka.com/resources/ml-lectures-1.html#l01-what-is-machine-learning)\n + nearest neighbor methods [lesson 02](https://sebastianraschka.com/resources/ml-lectures-1.html#l02-nearest-neighbor-methods)\n \n- the [wikipedia page on kmeans clustering](https://en.wikipedia.org/wiki/K-means_clustering) is well written too- plot the [decision boundary of the clustering](https://scikit-learn.org/stable/auto_examples/ensemble/plot_voting_decision_regions.html#sphx-glr-auto-examples-ensemble-plot-voting-decision-regions-py)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown", "markdown" ], [ "code", "code", "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code", "code", "code" ], [ "markdown", "markdown" ] ]

ec7a3626625406a7a8814fb97d4cb35fba4bf04c

43,560

ipynb

Jupyter Notebook

merging_fractionated.ipynb

robertbuecker/serialed-examples

418094a8da456804aa4b296ee55eed96c127bdef

[ "MIT" ]

2

2021-01-28T02:30:23.000Z

2021-02-12T18:45:31.000Z

merging_fractionated.ipynb

robertbuecker/serialed-examples

418094a8da456804aa4b296ee55eed96c127bdef

[ "MIT" ]

null

merging_fractionated.ipynb

robertbuecker/serialed-examples

418094a8da456804aa4b296ee55eed96c127bdef

[ "MIT" ]

2

2021-01-28T02:31:57.000Z

2021-04-26T16:04:37.000Z

40.333333

202

0.348623

[ [ [ "# only for development\n%load_ext autoreload\n%autoreload 2", "_____no_output_____" ], [ "import warnings\nwarnings.filterwarnings('ignore')\nimport matplotlib.pyplot as plt\nfrom diffractem import io, tools\nfrom diffractem.stream_parser import StreamParser, make_substream\nimport numpy as np\nimport pandas as pd\nimport os\nimport matplotlib\nimport seaborn as sns\nbin_path = '/opts/crystfel_latest/bin/' # might be different than standard\nfrom glob import glob\nfrom concurrent.futures import ProcessPoolExecutor, ThreadPoolExecutor", "_____no_output_____" ] ], [ [ "# Merging and anaysis of dose-fractionated sets\n...works essentially just as in `merging.ipynb`, just additionally making use of `partialator`s `custom-split` capability to make sub-sets with the individual frame numbers.\n\nTo make this work, we first need to get a `hits_allframe_split.txt` file which contains labels for the custom split, that is, the frame number per crystal, to which each Event in the file belongs.", "_____no_output_____" ] ], [ [ "stream = StreamParser('streams/hits_allframe.stream')\nprint(f'Stream file contains {stream.num_crystals} crystals in {stream.num_shots} shots')\nsplit_data = stream.shots[['file', 'Event', 'header/int//%/shots/frame']]\nsplit_data.to_csv(stream.filename.rsplit('.', 1)[0] + '_split.txt', index=False, header=False, sep=' ')\n\n!head -n 11 streams/hits_allframe_split.txt # have a look", "Stream file contains 12750 crystals in 13220 shots\nproc_data/LysoS1_001_00000_allframe_hit.h5 entry//0 0\nproc_data/LysoS1_001_00000_allframe_hit.h5 entry//1 1\nproc_data/LysoS1_001_00000_allframe_hit.h5 entry//2 2\nproc_data/LysoS1_001_00000_allframe_hit.h5 entry//3 3\nproc_data/LysoS1_001_00000_allframe_hit.h5 entry//4 4\nproc_data/LysoS1_001_00000_allframe_hit.h5 entry//5 5\nproc_data/LysoS1_001_00000_allframe_hit.h5 entry//6 6\nproc_data/LysoS1_001_00000_allframe_hit.h5 entry//7 7\nproc_data/LysoS1_001_00000_allframe_hit.h5 entry//8 8\nproc_data/LysoS1_001_00000_allframe_hit.h5 entry//9 9\nproc_data/LysoS1_001_00000_allframe_hit.h5 entry//10 0\n" ] ], [ [ "## Preparation of a partialator script\n...just as in `merging.ipynb`, just that we do not scan the `stop-after`, but instead set `split=True`.", "_____no_output_____" ] ], [ [ "# get a list of stream files\nstream_list = glob('streams/hits_allframe.stream')\n\npopts = {'no-polarisation': True, 'no-Bscale': False, 'no-scale': False, \n 'force-bandwidth': 2e-5, 'force-radius': False, 'force-lambda': 0.0251,\n 'push-res': 1.4, 'min-measurements': [3, ], 'model': ['unity', 'xsphere'],\n 'symmetry': '422', 'no-logs': False, 'iterations': 3, 'j': 10}\n\n# you need to set those if you want to use slurm to submit merging runs\nslurm_opts = {'C': 'scratch', \n 'partition': 'medium', \n 'time': '\"04:00:00\"',\n 'nodes': 1}\n\nsettings = tools.call_partialator(stream_list, popts, par_runs=4, \n split=True, out_dir='merged_frac',\n slurm=False, cache_streams=False, \n slurm_opts=slurm_opts)\n\n!chmod +x partialator_run.sh", "Please run partialator_run.sh to start merging.\n" ], [ "# example how to send data to a cluster\n# !scp -r streams [email protected]:~/SHARED/EDIFF/temp\n!scp partialator_run.sh [email protected]:~/SHARED/EDIFF/temp", "_____no_output_____" ], [ "# example how to get it back from a cluster\n# %mkdir merged\n# !scp '[email protected]:~/SHARED/EDIFF/temp/merged/*.hkl*' merged/", "_____no_output_____" ] ], [ [ "## Analyze and validate results\n...again, the only difference to `merging.ipynb` being that `custom_split=True` in the first command.", "_____no_output_____" ] ], [ [ "# check what hkls we have available....\nsettings = tools.get_hkl_settings('merged_frac/hits_allframe*-*.hkl', unique_only=True, custom_split=True)\n\nif 'input' in settings.columns:\n settings['input'] = settings['input'].str.rsplit('/', 1, expand=True).iloc[:,-1]\n\ndefault_symmetry = '422'\nhighres = 1.75 # highest shell, in A\nnshells = 10\n\n# tools.analyze_hkl() #...is used using ProcessPoolExecutor\n\nftrs = {}\nwith ProcessPoolExecutor() as exc:\n for _, s in settings.iterrows():\n ftrs[s.hklfile] = exc.submit(tools.analyze_hkl, fn=s.hklfile, cell='refined.cell', \n point_group=s.symmetry if 'symmetry' in s else default_symmetry, \n highres=highres, nshells=nshells, bin_path='/opts/crystfel_master/bin') \n\nerr = {lbl: v.exception() for lbl, v in ftrs.items() if v.exception()}\nif err:\n print('Analysis gave errors!', str(err))\nout = {lbl: v.result() for lbl, v in ftrs.items() if not v.exception()}\n\nsd = pd.concat([v.result()[0].assign(hklfile=lbl) \n for lbl, v in ftrs.items() \n if not v.exception()], axis=0).merge(\n settings, on='hklfile')\n\noverall = pd.concat([pd.DataFrame(v.result()[1], index=[lbl])\n for lbl, v in ftrs.items() \n if not v.exception()], axis=0).merge(\n settings, right_on='hklfile', left_index=True).rename(\n columns={'<snr>': 'SNR', 'redundancy': 'Red', 'completeness': 'Compl', 'CC*': 'CCstar'})\n\n# write out results\n%rm -f shell/*\nfor ident, grp in sd.groupby(['hklfile']):\n grp.sort_values('Center 1/nm')[['Center 1/nm', 'nref', 'Possible', 'Compl', 'Meas', 'Red', 'SNR',\n 'Mean', 'd/A', 'Min 1/nm', 'Max 1/nm', 'CC', 'CCstar',\n 'Rsplit']].to_csv(f'shell/{ident.rsplit(\"/\",1)[-1]}.csv', index=False, float_format='%.2f')", "_____no_output_____" ] ], [ [ "#### Example to show results\n...using the result DataFrame's `pivot` function.", "_____no_output_____" ] ], [ [ "# convenient function to get FOMs. Set the one you want as 'value'\nsd.pivot(index='d/A', columns='hklfile', values=['CC']).sort_index(ascending=False)", "_____no_output_____" ] ], [ [ "## Analysis Plot\n...as Fig. 1 in `merging.ipynb`.", "_____no_output_____" ] ], [ [ "# %matplotlib inline\n%matplotlib widget\n\nmodel = 'unity' # 'unity' or 'xsphere'\n\n# SETTINGS ---\n\nfh, axs = plt.subplots(2, 2, figsize=(18/2.54,15/2.54), dpi=120, sharex=True)\nlsp, lrow = 0.85, 3 # space near top left for legend, and # of legend columns\n\n# pick your FOMs and their y ranges\nFOMs = [('CC', 0, 1), ('Mean', 0, 40), ('Compl', 0, 100), ('Red', 0, 100)]\nsdsel = sd.query(f'model == \"{model}\"')\nangstrom = False # if True, show x axis in A, instead of 1/nm\n\n# ------\n\ntry:\n import seaborn as sns\n sns.set('notebook','whitegrid') # optional. Set a style...\nexcept:\n print('Seaborn not installed, it seems.')\n\naxs = axs.ravel()\n\n# ids = get_id_table(sdsel['identifier'])\nidcols = [cn for cn, col in sdsel[settings.columns].iteritems() \n if len(col.unique()) > 1 and (cn != 'hklfile')]\nprint('Legend is', ' '.join(idcols))\n\nfor ident, grp in sdsel.groupby(['hklfile']):\n \n ls = '-'\n \n lbl = tuple(grp[idcols].drop_duplicates().values.astype(str).ravel())\n \n for ax, (fom, ymin, ymax) in zip(axs, FOMs):\n ax.plot(grp['d/A'] if angstrom else grp['Center 1/nm'], grp[fom], \n label=' '.join(lbl), ls=ls)\n ax.set_title(fom)\n ax.set_ylim((ymin, ymax))\n if angstrom:\n ax.set_xlim(sorted(ax.get_xlim(), reverse=True))\n ax.grid(True)\n if fom in ['CC', 'CCstar']:\n ax.axhline(0.143 if fom == 'CC' else 0.5,ls=':')\n \nlg = fh.legend(*ax.get_legend_handles_labels(), ncol=lrow, \n fontsize='xx-small', loc='lower center', \n bbox_to_anchor=(0.5, lsp), frameon=True)\naxs[-1].set_xlabel(r'Resolution shell/Å' if angstrom else r'Resolution shell/nm$^{-1}$')\nplt.draw()\n# lpos = lg.get_window_extent()\n\nfh.subplots_adjust(wspace=0.3, top=lsp-0.05)", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

ec7a3ed76462fa4cf7c00d56927a67f0cd965058

24,880

ipynb

Jupyter Notebook

book/tools-for-mathematics/02-algebra/solutions/.main.md.bcp.ipynb

daffidwilde/pfm

dcf38faccee3c212c8394c36f4c093a2916d283e

[ "MIT" ]

null

book/tools-for-mathematics/02-algebra/solutions/.main.md.bcp.ipynb

daffidwilde/pfm

dcf38faccee3c212c8394c36f4c093a2916d283e

[ "MIT" ]

null

book/tools-for-mathematics/02-algebra/solutions/.main.md.bcp.ipynb

daffidwilde/pfm

dcf38faccee3c212c8394c36f4c093a2916d283e

[ "MIT" ]

null

20.561983

786

0.447629

[ [ [ "# Solutions\n\n## Question 1\n\n> `1`. Simplify the following expressions:\n\n> $\\frac{3}{\\sqrt{3}}$:", "_____no_output_____" ] ], [ [ "import sympy as sym\n\nexpression = sym.S(3) / sym.sqrt(3)\nsym.simplify(expression)", "_____no_output_____" ] ], [ [ "> $\\frac{2 ^ {78}}{2 ^ {12}2^{-32}}$:", "_____no_output_____" ] ], [ [ "sym.S(2) ** 78 / (sym.S(2) ** 12 * sym.S(2) ** (-32))", "_____no_output_____" ] ], [ [ "> $8^0$:", "_____no_output_____" ] ], [ [ "sym.S(8) ** 0", "_____no_output_____" ] ], [ [ "> $a^4b^{-2}+a^{3}b^{2}+a^{4}b^0$:", "_____no_output_____" ] ], [ [ "a = sym.Symbol(\"a\")\nb = sym.Symbol(\"b\")\nsym.factor(a ** 4 * b ** (-2) + a ** 3 * b ** 2 + a ** 4 * b ** 0)", "_____no_output_____" ] ], [ [ "## Question 2\n\n> `2`. Solve the following equations:\n\n> $x + 3 = -1$:", "_____no_output_____" ] ], [ [ "x = sym.Symbol(\"x\")\nequation = sym.Eq(x + 3, -1)\nsym.solveset(equation, x)", "_____no_output_____" ] ], [ [ "> $3 x ^ 2 - 2 x = 5$:", "_____no_output_____" ] ], [ [ "equation = sym.Eq(3 * x ** 2 - 2 * x, 5)\nsym.solveset(equation, x)", "_____no_output_____" ] ], [ [ "> $x (x - 1) (x + 3) = 0$:", "_____no_output_____" ] ], [ [ "equation = sym.Eq(x * (x - 1) * (x + 3), 0)\nsym.solveset(equation, x)", "_____no_output_____" ] ], [ [ "> $4 x ^3 + 7x - 24 = 1$:", "_____no_output_____" ] ], [ [ "equation = sym.Eq(4 * x ** 3 + 7 * x - 24, 1)\nsym.solveset(equation, x)", "_____no_output_____" ] ], [ [ "## Question 3\n\n> `3`. Consider the equation: $x ^ 2 + 4 - y = \\frac{1}{y}$:\n\n> Find the solution to this equation for $x$.", "_____no_output_____" ] ], [ [ "y = sym.Symbol(\"y\")\nequation = sym.Eq(x ** 2 + 4 - y, 1 / y)\nsolution = sym.solveset(equation, x)\nsolution", "_____no_output_____" ] ], [ [ "> Obtain the specific solution when $y = 5$. Do this in two ways:\n> substitute the value in to your equation and substitute the value in to\n> your solution.", "_____no_output_____" ] ], [ [ "solution.subs({y: 5})", "_____no_output_____" ], [ "solution = sym.solveset(equation.subs({y: 5}), x)\nsolution", "_____no_output_____" ] ], [ [ "## Question 4\n\n> `4`. Consider the quadratic: $f(x)=4x ^ 2 + 16x + 25$:\n\n> Calculate the discriminant of the quadratic equation $4x ^ 2 + 16x + 25 =\n> 0$. What does this tell us about the solutions to the equation? What\n> does this tell us about the graph of $f(x)$?", "_____no_output_____" ] ], [ [ "quadratic = 4 * x ** 2 + 16 * x + 25\nsym.discriminant(quadratic)", "_____no_output_____" ] ], [ [ " \nThis is negative so we know that the equation does not have any real solutions and\nhence the graph does not cross the x-axis. \nSince the coefficient of $x^2$ is positive it means that the graph is above \nthe $y=0$ line.\n\n\n> By completing the square, show that the minimum point of $f(x)$ is\n> $\\left(-2, 9\\right)$", "_____no_output_____" ] ], [ [ "a, b, c = sym.Symbol(\"a\"), sym.Symbol(\"b\"), sym.Symbol(\"c\")\ncompleted_square = a * (x - b) ** 2 + c\nsym.expand(completed_square)", "_____no_output_____" ] ], [ [ "This gives $a=4$.", "_____no_output_____" ] ], [ [ "completed_square = completed_square.subs({a: 4})\nsym.expand(completed_square)", "_____no_output_____" ] ], [ [ "Comparing the coefficients of $x$ we have the equation:\n\n$$\n - 8 b = 16\n$$", "_____no_output_____" ] ], [ [ "equation = sym.Eq(-8 * b, 16)\nsym.solveset(equation, b)", "_____no_output_____" ] ], [ [ "Substituting:", "_____no_output_____" ] ], [ [ "completed_square = completed_square.subs({b: -2})\nsym.expand(completed_square)", "_____no_output_____" ] ], [ [ "Comparing the coefficients of $x^0$ this gives:\n\n$$c+16=25$$", "_____no_output_____" ] ], [ [ "equation = sym.Eq(c + 16, 25)\nsym.solveset(equation, c)", "_____no_output_____" ], [ "completed_square = completed_square.subs({c: 9})\ncompleted_square", "_____no_output_____" ] ], [ [ "The lowest value of $f(x)$ is for $x=-2$ which gives: $f(-2)=9$ as expected.\n\n## Question 5\n\n> `5`. Consider the quadratic: $f(x)=-3x ^ 2 + 24x - 97$:\n\n> Calculate the discriminant of the quadratic equation $-3x ^ 2 + 24x - 97 =\n> 0$. What does this tell us about the solutions to the equation? What\n> does this tell us about the graph of $f(x)$?", "_____no_output_____" ] ], [ [ "quadratic = -3 * x ** 2 + 24 * x - 97\nsym.discriminant(quadratic)", "_____no_output_____" ] ], [ [ "This is negative so we know that the equation does not have any real solutions and\nhence the graph does not cross the x-axis. \nSince the coefficient of $x^2$ is negative it means that the graph is below \nthe $y=0$ line.\n\n\n> By completing the square, show that the maximum point of $f(x)$ is\n> $\\left(4, -49\\right)$", "_____no_output_____" ] ], [ [ "a, b, c = sym.Symbol(\"a\"), sym.Symbol(\"b\"), sym.Symbol(\"c\")\ncompleted_square = a * (x - b) ** 2 + c\nsym.expand(completed_square)", "_____no_output_____" ] ], [ [ "This gives $a=-3$.", "_____no_output_____" ] ], [ [ "completed_square = completed_square.subs({a: -3})\nsym.expand(completed_square)", "_____no_output_____" ] ], [ [ "Comparing the coefficients of $x$ we have the equation:\n\n$$\n 6 b = 24\n$$", "_____no_output_____" ] ], [ [ "equation = sym.Eq(6 * b, 24)\nsym.solveset(equation, b)", "_____no_output_____" ] ], [ [ "Substituting:", "_____no_output_____" ] ], [ [ "completed_square = completed_square.subs({b: 4})\nsym.expand(completed_square)", "_____no_output_____" ] ], [ [ "Comparing the coefficients of $x^0$ this gives:\n\n$$c-48=-97$$", "_____no_output_____" ] ], [ [ "equation = sym.Eq(c - 48, -97)\nsym.solveset(equation, c)", "_____no_output_____" ], [ "completed_square = completed_square.subs({c: -49})\ncompleted_square", "_____no_output_____" ] ], [ [ "The highest value of $f(x)$ is for $x=4$ which gives: $f(4)=-49$ as expected.\n\n`6`. Consider the function $f(x) = x^ 2 + a x + b$.\n\n> Given that $f(0) = 0$ and $f(3) = 0$ obtain the values of $a$ and $b$.\n\nSubstituting 0 in to $f$ gives:", "_____no_output_____" ] ], [ [ "expression = x ** 2 + a * x + b\nexpression.subs({x: 0})", "_____no_output_____" ] ], [ [ "This implies that $b=0$. Substituting back in to the expression:", "_____no_output_____" ] ], [ [ "expression = expression.subs({b: 0})\nexpression", "_____no_output_____" ] ], [ [ "Substituting $x=3$ in to this expression gives:", "_____no_output_____" ] ], [ [ "expression.subs({x: 3})", "_____no_output_____" ] ], [ [ "This gives the equation:\n\n$$\n 3 a + 9 = 0\n$$", "_____no_output_____" ] ], [ [ "sym.solveset(expression.subs({x: 3}), a)", "_____no_output_____" ] ], [ [ "Our expression is thus:", "_____no_output_____" ] ], [ [ "expression = expression.subs({a: -3})\nexpression", "_____no_output_____" ] ], [ [ "> By completing the square confirm that graph of $f(x)$ has a line of symmetry\n> at $x=\\frac{3}{2}$", "_____no_output_____" ] ], [ [ "completed_square = a * (x - b) ** 2 + c\nsym.expand(completed_square)", "_____no_output_____" ] ], [ [ "We see that $a=1$ and. Substituting:", "_____no_output_____" ] ], [ [ "completed_square = completed_square.subs({a: 1})\nsym.expand(completed_square)", "_____no_output_____" ] ], [ [ "This gives:\n\n$$\n -2b=-3\n$$", "_____no_output_____" ] ], [ [ "equation = sym.Eq(-2 * b, -3)\nsym.solveset(equation, b)", "_____no_output_____" ] ], [ [ "Substituting:", "_____no_output_____" ] ], [ [ "completed_square = completed_square.subs({b: sym.S(3) / 2})\nsym.expand(completed_square)", "_____no_output_____" ] ], [ [ "Which gives:\n\n$$\n c + 9 / 4 = 0\n$$", "_____no_output_____" ] ], [ [ "equation = sym.Eq(c + sym.S(9) / 4, 0)\nsym.solveset(equation, c)", "_____no_output_____" ] ], [ [ "Substituting:", "_____no_output_____" ] ], [ [ "completed_square = completed_square.subs({c: -sym.S(9) / 4})\ncompleted_square", "_____no_output_____" ] ], [ [ "Thus $x=3/2$ is a line of symmetry.", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ] ]

ec7a46de854c36a3f4ed26f2258ec7c2585a7f00

531,271

ipynb

Jupyter Notebook

SomeAstroNBs/rad_no.ecsv.ipynb

iskren-y-g/HowToAstroNBs

a8b079c0bc6eb5a2958ac1fe9254b9ed22764cca

[ "BSD-3-Clause" ]

1

2021-04-21T15:00:35.000Z

SomeAstroNBs/rad_no.ecsv.ipynb

iskren-y-g/SomeAstroNBs

a8b079c0bc6eb5a2958ac1fe9254b9ed22764cca

[ "BSD-3-Clause" ]

null

SomeAstroNBs/rad_no.ecsv.ipynb

iskren-y-g/SomeAstroNBs

a8b079c0bc6eb5a2958ac1fe9254b9ed22764cca

[ "BSD-3-Clause" ]

null

526.010891

231,712

0.937704

[ [ [ "# Perform a Maximum likelihood and MCMC analysis on a set of data\n\nThe model is a Sersic function.", "_____no_output_____" ] ], [ [ "# Importing modules\n\nimport sys,os,gzip, pickle\nimport numpy as np\nfrom matplotlib import pyplot as plt\n\nplt.rcParams['font.family'] = 'serif'\nplt.rcParams['font.size'] = 12\nfor axticks in ['xtick','ytick']:\n plt.rcParams['{:}.direction'.format(axticks)] = 'in'\n plt.rcParams['{:}.minor.visible'.format(axticks)] = True\n\nfrom astropy import table\nfrom astropy.table import Table\nfrom astropy.modeling import models, fitting\nfrom sklearn.neighbors import KernelDensity\n", "_____no_output_____" ], [ "# Read data \n\nrad_no_tbl = Table.read('rad_no.ecsv', format ='ascii.ecsv')\nprint(rad_no_tbl.colnames)", "['GCC_group', 'Rproj', 'Nobs', 'Ncorr', 'area', 'rho', 'rho_upp_unc', 'rho_low_unc']\n" ], [ "rad_no_tbl", "_____no_output_____" ] ], [ [ "## Define the input data\nSelect a subset of the data to work on", "_____no_output_____" ] ], [ [ "[print(grp[:], end = ' ') for grp in rad_no_tbl['GCC_group']][0]", "All All All All All All All Unresolved Unresolved Unresolved Unresolved Unresolved Unresolved Unresolved Partially resolved Partially resolved Partially resolved Partially resolved Partially resolved Partially resolved Partially resolved Extended Extended Extended Extended Extended Extended Extended Blue Blue Blue Blue Blue Blue Blue Red Red Red Red Red Red Red OMP OMP OMP OMP OMP OMP OMP OMR OMR OMR OMR OMR OMR OMR OM(P+R) OM(P+R) OM(P+R) OM(P+R) OM(P+R) OM(P+R) OM(P+R) IMR IMR IMR IMR IMR IMR IMR " ], [ "select = (rad_no_tbl['GCC_group'] == 'OMP')\nselect = (rad_no_tbl['GCC_group'] == 'OMR')\nselect = (rad_no_tbl['GCC_group'] == 'Blue')\nselect = (rad_no_tbl['GCC_group'] == 'Red')\nselect = (rad_no_tbl['GCC_group'] == 'IMR') & (rad_no_tbl['rho']>0.0)\nselect = (rad_no_tbl['GCC_group'] == 'All')\nselect = (rad_no_tbl['GCC_group'] == 'OM(P+R)')\n\nx_ml = rad_no_tbl['Rproj'][select]\n\ny_ml = rad_no_tbl['rho'][select]\ny_err_ml = np.sqrt(rad_no_tbl['rho_upp_unc'][select]**2 + rad_no_tbl['rho_low_unc'][select]**2)\n\ny_err_ml_upp = rad_no_tbl['rho_upp_unc']\ny_err_ml_low = rad_no_tbl['rho_low_unc']\n", "_____no_output_____" ], [ "# Visualize the data\n\nfig,ax = plt.subplots(figsize=(10,10))\n\nfor tx,ty,terr in zip(x_ml,y_ml,y_err_ml):\n ax.errorbar(tx, ty, yerr=terr, marker='o', mfc='white', ms=8, color='C0',\n label='Data', alpha=.5\n )\n \nax.set_xscale('log')\nax.set_yscale('log')\n#ax.set_xlim(7,1.5e2)\nax.set_ylim(10**(-4),10**(-.12))\n", "_____no_output_____" ] ], [ [ "# Maximum Likelihood fitting and MCMC", "_____no_output_____" ] ], [ [ "import scipy\nimport scipy.optimize as op\n", "_____no_output_____" ] ], [ [ "## Define the model", "_____no_output_____" ] ], [ [ "def b_n_exact( n ):\n \"\"\"Exact calculation of the Sersic derived parameter b_n, via solution\n of the function\n Gamma(2n) = 2 gamma_inc(2n, b_n)\n where Gamma = Gamma function and gamma_inc = lower incomplete gamma function.\n\n If n is a list or Numpy array, the return value is a 1-d Numpy array\n \"\"\"\n import scipy\n from scipy.special import gamma as Gamma\n \n def gammainc_lower_scipy( z, b ):\n return scipy.special.gamma(z) * scipy.special.gammainc(z, b)\n GammaInc = gammainc_lower_scipy\n \n def myfunc(bn, n):\n return abs(float(2*GammaInc(2*n, bn) - Gamma(2*n)))\n if np.iterable(n):\n b = [scipy.optimize.brent(myfunc, (nn,)) for nn in n]\n b = np.array(b)\n else:\n b = scipy.optimize.brent(myfunc, (n,))\n return b", "_____no_output_____" ], [ "def Sersic(x, I_e, r_e, n):\n return I_e * np.exp( -b_n_exact(n)*(pow(x/r_e, 1.0/n) - 1.0) )", "_____no_output_____" ] ], [ [ "## Define likelihood function", "_____no_output_____" ] ], [ [ "def lnlike(theta, x, y, yerr):\n I_e, r_e, n, lnf = theta\n \n model = Sersic(x, I_e, r_e, n)\n \n inv_sigma2 = 1.0/(yerr**2 + model**2*np.exp(2*lnf))\n return -0.5*(np.sum((y-model)**2*inv_sigma2 \n + np.log(2*np.pi*inv_sigma2))\n )", "_____no_output_____" ], [ "# Make an initial guesses for the model parameters\n# For this particular case, we set no additional error and rely fully on the \n# measurement uncertainties. But you can test by setting lnf_init>-3\n\nI_e_init, r_e_init, n_init, lnf_init = 5e-2, 50.0, 1.5, -10\n\nparams_0 = [I_e_init, r_e_init, n_init, lnf_init]\n\nnll = lambda *args: -lnlike(*args)\nresult = op.minimize(nll,\n params_0,\n args=(x_ml, y_ml, y_err_ml), method = 'L-BFGS-B' #SLSQP' #BFGS' #\n )\nif not(result['success']):\n print(\"Max likelihood failed.\")\n print(result['message'])\n\nml_I_e, ml_r_e, ml_n, ml_lnf = result['x']\nprint('ml_I_e={:.2e}, ml_r_e={:.2f}, ml_n={:.2f}, lnf={:.2f}'.format(ml_I_e, ml_r_e, ml_n, ml_lnf))\n", "ml_I_e=5.79e-02, ml_r_e=40.81, ml_n=1.76, lnf=-10.00\n" ], [ "# Store the result in a list to pass\nparams_init = [val for val in [I_e_init, r_e_init, n_init]]\nparams_ml = result['x'][:-1]\n", "_____no_output_____" ], [ "# Have a look at the ML fit\n\nfig,ax = plt.subplots(figsize=(10,10))\n\nfor i,(tx,ty,terr) in enumerate(zip(x_ml,y_ml,y_err_ml)):\n if i==0:\n label = 'Data'\n else:\n label = ''\n ax.errorbar(tx, ty, yerr=terr, marker='o', mfc='white', ms=8, color='C0',\n label=label, alpha=0.5\n )\n \n#for i,n in enumerate(y_ml):\n# ax.text(x_ml[i],np.log10(y_ml[i]*1.05), '{:}'.format(round(n)))\n\nx_samples = np.arange(x_ml.min(),x_ml.max(),(x_ml.max()-x_ml.min())/100)\nax.plot(x_samples, Sersic(x_samples, *params_ml), '-', label='Fitted')\nax.plot(x_samples, Sersic(x_samples, *params_init), '--', label='Initial')\n\nax.set_xscale('log')\nax.set_yscale('log')\n\nax.legend()", "_____no_output_____" ], [ "# Set priors for the fitted parameters\n\ndef lnprior(theta):\n I_e, r_e, n, lnf = theta\n if 1e-4 <= I_e <= .1 and 10. <= r_e <= 100 and 0.2<=n<=5. and -12.0 < lnf < -8:\n lnprior_val = 0.0\n return lnprior_val\n\n return -np.inf", "_____no_output_____" ], [ "def lnprob(theta, x, y, yerr):\n \n lp = lnprior(theta)\n if not np.isfinite(lp):\n return -np.inf\n \n try:\n lnprobval = lp\n if not np.any(np.isnan(lnlike(theta, x, y, yerr))):\n lnprobval = lp + lnlike(theta, x, y, yerr)\n except:\n lnprobval = lp + -np.inf\n \n return lnprobval\n", "_____no_output_____" ], [ "# Set up the MCMC chains\nndim = 4\nnwalkers = 50\nsteps = 10000\nnthreads = 4", "_____no_output_____" ], [ "# Initialize the walkers to the vicinity of the parameters. There are a few options.\n# You can either start from the best fit ML, or from the initial values.\n\n#pos = [result[\"x\"] + 1e-1*np.random.randn(ndim) for i in range(nwalkers)]\npos = [np.array([I_e_init, r_e_init, n_init, lnf_init]) + 1e-3*np.random.randn(ndim) for i in range(nwalkers)]\n#pos = [np.array([ml_I_e, ml_r_e, ml_n, ml_lnf]) + 1e-3*np.random.randn(ndim) for i in range(nwalkers)]\n", "_____no_output_____" ], [ "import time \nimport emcee\nfrom multiprocessing import Pool\n\nwith Pool() as pool:\n sampler = emcee.EnsembleSampler(nwalkers, ndim, lnprob, pool=pool,\n args=(x_ml, y_ml, y_err_ml))\n start = time.time()\n sampler.run_mcmc(pos, steps, progress=True)\n end = time.time()\n multi_time = end - start\n print(\"Multiprocessing took {0:.1f} seconds\".format(multi_time))\n", "100%|██████████| 10000/10000 [03:47<00:00, 43.93it/s]" ], [ "tau = sampler.get_autocorr_time()\nprint(tau)", "[135.97144663 150.19419933 95.20913964 88.92618216]\n" ], [ "# The acceptance rate should be between 0.3-0.5.\n\nprint(\"Mean acceptance rate is: {0:1.2f}\".format(np.mean(sampler.acceptance_fraction)))", "Mean acceptance rate is: 0.44\n" ], [ "# (Optional) Save the sampler as a pickle\n\nfilename = 'rad_no_OMP_OMR_sampler_pickle.gz'\n\nwith gzip.open(filename, 'wb') as f:\n print('Compressing sampler to pickle file ... {:s}'.format(filename))\n pickle.dump(sampler, f, pickle.HIGHEST_PROTOCOL)", "_____no_output_____" ], [ "# (OPTIONAL) Load the sampler from a saved pickle.\n\nfilename = 'rad_no_IMR_sampler_pickle.gz'\n\nwith gzip.open(filename, 'rb') as f:\n print('Decompressing the pickle file ... {:s}'.format(filename))\n sampler = pickle.load(f)\n ", "_____no_output_____" ], [ "# Let's extract the chains\n\nsamples = sampler.chain", "_____no_output_____" ], [ "# Visualize the chains.\n\nfig = plt.figure()\ndim_name = [r'I_e', r'r_e', r'n', r'$\\ln f$']\n\nfor dim in range(ndim):\n ax = fig.add_subplot(ndim, 1, dim+1)\n for i in range(nwalkers):\n ax.plot(np.arange(steps),\n samples[i, :, dim],\n ls='-',\n color='black',\n alpha=10./nwalkers,\n rasterized=True\n )\n ax.set_ylabel(dim_name[dim], fontsize='large')\n ax.set_xscale('log')\nax.set_xlabel('step', fontsize='large')", "_____no_output_____" ], [ "filename = 'rho_OMP_OMR_Sersic_mcmc_steps_test.pdf'\n\nfig.savefig('{:}'.format(filename), format='pdf', transparent=True,\n bbox_inches='tight', dpi=100\n )\nos.system('pdfcrop {:} {:}'.format(filename,filename))", "_____no_output_____" ], [ "# Do a bit of cleaning\nimport gc, time\nstart = time.time()\ngc.collect()\nprint('Cleaning duration {:.1f} seconds'.format(time.time()-start))", "Cleaning duration 0.1 seconds\n" ], [ "#It looks like the walkers have \"burned in\" by ~50-100 steps, so keep only those samples.\n\nsamples = sampler.chain[:, 50:, :].reshape((-1, ndim))\nsamples = sampler.chain[:, 50:, :ndim-1].reshape((-1, ndim-1)) # Use this if you want to avoid showing the lnf\n", "_____no_output_____" ], [ "# Limit by patrameter value\nsample_limits_bl = False\n\nIe_low = samples[:,0].min() ; Ie_upp = .1 #samples[:,0].max() #\nre_low = samples[:,1].min() ; re_upp = 150 #samples[:,1].max() #90\nn_low = samples[:,2].min() ; n_upp = 3 #samples[:,2].max()\n\nsample_lims = (samples[:,0]>Ie_low) & (samples[:,0]<Ie_upp) & \\\n (samples[:,1]>re_low) & (samples[:,1]<re_upp) & \\\n (samples[:,2]>n_low) & (samples[:,2]<n_upp)\n\n\nif sample_limits_bl ==True:\n samples = samples[:,:][ sample_lims ]", "_____no_output_____" ], [ "# Show kde over the corner histograms?\nshow_kde = False", "_____no_output_____" ], [ "import corner\nfrom sklearn.neighbors import KernelDensity\n\nfig = plt.figure()\nfig = corner.corner(samples,\n labels=[r'I$_e$', r'r$_e$', r'$n$'], #r'$\\ln f$'],\n quantiles=[0.16, 0.5, 0.84], rasterized=True,\n bins=30, #hist_bin_factor=2, \n show_titles=True)\n\nif show_kde:\n #t_samples = sampler.chain[:, 50:, :ndim].reshape((-1, ndim)) # Use this if you want to avoid showing the lnf\n #if sample_limits_bl: t_samples = t_samples[ sample_lims ]\n #t_samples[:, ndim-1] = np.exp(t_samples[:, ndim-1])\n if samples.shape[1] == ndim: # If the lnf is not removed from the samples\n samples[:, ndim-1] = np.exp(samples[:, ndim-1])\n\n # Extract the axes\n axes = np.array(fig.axes).reshape((ndim-1, ndim-1))\n labels=[r'I$_e$', r'r$_e$', r'$n$']\n\n for alabel,dim in zip(labels,range(ndim-1)):\n #dim_samples = t_samples[:, dim]\n dim_samples = samples[:, dim]\n kde_bin = (np.nanmax(dim_samples)-np.nanmin(dim_samples))/30.\n\n X = dim_samples[:,None]\n\n print('KDE estimation with kde bin = {:.3f}'.format(kde_bin))\n kde = KernelDensity(kernel='gaussian', bandwidth=kde_bin).fit(X)\n\n print('KDE scores')\n samples_kde = np.linspace(X.min(),X.max(),100)\n log_dens_kde = kde.score_samples(samples_kde[:,None])\n\n samp_d = (X.max()-X.min())/100\n\n kde_xy = np.array([samples_kde, np.exp(log_dens_kde) * len(X) * samp_d * (100/30)])\n\n print('KDE peak = ', end ='')\n kde_y_max = np.nanmax(kde_xy[1])\n kde_peak = kde_xy[0][kde_xy[1] == kde_xy[1].max()][0]\n fwhm_low = 2*kde_peak/(kde_peak*.93)\n fwhm_upp = 2*kde_peak/(kde_peak*1.07)\n fwhm_peak_kde = kde_xy[0][(kde_xy[1] > kde_y_max/fwhm_low) & (kde_xy[1] < kde_y_max/fwhm_upp)]\n hwhm_peak_kde_low = kde_peak-fwhm_peak_kde.min()\n hwhm_peak_kde_upp = fwhm_peak_kde.max()-kde_peak\n print('{:.3f} +{:.3f}, -{:.3f}'.format(kde_peak,hwhm_peak_kde_upp,hwhm_peak_kde_low))\n dim_par_lab = r'{:} = {:.3f} $^{{+{:.3f}}}_{{-{:.3f}}}$'.format(alabel,kde_peak,hwhm_peak_kde_upp,abs(hwhm_peak_kde_low))\n\n # Loop over the diagonal\n ax = axes[dim, dim]\n ax.axvline(kde_peak, color='C0')\n ax.axvline(kde_peak-abs(hwhm_peak_kde_low), color='C0')\n ax.axvline(hwhm_peak_kde_upp+kde_peak, color='C0')\n ax.plot(kde_xy[0], kde_xy[1], '-', color='C0')\n ax.text((ax.get_xlim()[1]-ax.get_xlim()[0])*.5*.35,\n ax.get_ylim()[1]*1.22, '{:s}'.format(dim_par_lab), color='C0')", "_____no_output_____" ], [ "filename = 'rho_OMP_OMR_mcmc_corner.pdf'\n\nfig.savefig('{:}'.format(filename), format='pdf', transparent=True,\n bbox_inches='tight', dpi=100\n )\n#os.system('pdfcrop {:} {:}'.format(filename,filename))", "_____no_output_____" ], [ "lb_use_params_mcmc_kde = False\nparams_mcmc_kde = [0.021, 45.356, 0.958]", "_____no_output_____" ], [ "# Print the MCMC parameters\n\nsamples = sampler.chain[:, 20:, :].reshape((-1, ndim))\nsamples[:, ndim-1] = np.exp(samples[:, ndim-1])\nif sample_limits_bl ==True:\n samples = samples[:,:][ sample_lims ]\namplitude_mcmc, sigma_mcmc, mu_mcmc, f_mcmc = map(lambda v: (v[1], v[2]-v[1], v[1]-v[0]),\n zip(*np.percentile(samples, [16, 50, 84],\n axis=0)))\n#samples[:, 3] = np.exp(samples[:, 3])\nI_e_mcmc, r_e_mcmc, n_mcmc, f_mcmc = map(lambda v: (v[1], v[2]-v[1], v[1]-v[0]),\n zip(*np.percentile(samples, [16, 50, 84],\n axis=0)))\nprint(\"MCMC Parameter estimates:\")\nprint(\"I_e:\\t{:.2e}\\t(+{:.2e}, -{:.2e}) [N/arcsec^2]\".format(I_e_mcmc[0],\n I_e_mcmc[1],\n I_e_mcmc[2]))\nprint(\"r_e:\\t{:.2f}\\t(+{:.2f}, -{:.2f}) [arcsec]\".format(r_e_mcmc[0],\n r_e_mcmc[1],\n r_e_mcmc[2]))\nprint(\"n:\\t{:.2f}\\t(+{:.2f}, -{:.2f})\".format(n_mcmc[0],\n n_mcmc[1],\n n_mcmc[2]))", "MCMC Parameter estimates:\nI_e:\t4.58e-02\t(+3.13e-02, -2.42e-02) [N/arcsec^2]\nr_e:\t43.57\t(+19.47, -10.78) [arcsec]\nn:\t2.83\t(+1.43, -1.36)\n" ], [ "print('I_0 = {:.2e}'.format(Sersic(0,I_e_mcmc[0],r_e_mcmc[0],n_mcmc[0])))", "I_0 = 9.54e+00\n" ], [ "# Visualize the results\nlb_show_init = False # Show the Initial expextation and ML fit?\n\nn_fits = 500\nplt_samples = samples\n\nx_samples = np.arange(x_ml.min(),x_ml.max(),(x_ml.max()-x_ml.min())/100)\ny_plt_ml = Sersic(x_samples, ml_I_e, ml_r_e, ml_n)\ny_plt_mcmc = Sersic(x_samples, I_e_mcmc[0],r_e_mcmc[0],n_mcmc[0])\ny_plt_mcmc_kde = Sersic(x_samples, *params_mcmc_kde)\n\nfig,ax = plt.subplots(figsize=(10,10))\n\nax.errorbar(x_ml, y_ml, yerr=y_err_ml,\n marker='o', mfc='white', ms=8, lw=2,\n label='Data', zorder=n_fits-2, alpha=1\n )\nfor i,n in enumerate(rad_no_tbl['Nobs'][select]):\n ax.text(x_ml[i]*1.01,y_ml[i]*.98, '{:}'.format(round(n)))\n\nax.plot(x_samples, y_plt_ml, '--', color='darkorange', lw=4, zorder=n_fits-1,\n label='Maximum Likelihood fit')\nax.plot(x_samples, y_plt_mcmc, color='black', alpha=1, zorder=n_fits,\n lw=2, label='MCMC posteriror resul')\nif lb_use_params_mcmc_kde:\n ax.plot(x_samples, y_plt_mcmc_kde, color='navy', alpha=1, zorder=n_fits,\n lw=2, label='MCMC posteriror KDE')\n\nfor i,(tI_e, tr_e, tn, tf) in enumerate(plt_samples[np.random.randint(len(plt_samples), size=n_fits)]):\n if i==0:\n ax.plot(x_ml, Sersic(x_ml, tI_e, tr_e, tn), color='royalblue', alpha=0.2, zorder=0,\n label='MCMC {:} randomly selected fits from {:} in total'.format(n_fits,round(samples.size/ndim))\n )\n else:\n ax.plot(x_ml, Sersic(x_ml, tI_e, tr_e, tn), color='royalblue', alpha=0.1, zorder=0)\n\nax.set_ylim(y_plt_ml.min()*.5,y_ml.max()*10**(1.01))\nax.set_xscale('log')\nax.set_yscale('log')\nax.legend()", "_____no_output_____" ], [ "filename = 'rho_OMP_OMR_Sersic_mcmc.pdf'\n\nfig.savefig(filename, format='pdf', transparent=True,\n bbox_inches='tight', dpi=100\n )", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7a53a98fe6059963285125cc7db07054f1efb6

151,231

ipynb

Jupyter Notebook

Python_Stock/Candlestick_Patterns/Candlestick_Takuri.ipynb

chunsj/Stock_Analysis_For_Quant

5f28ef9537885a695245d26f3010592a29d45a34

[ "MIT" ]

962

2019-07-17T09:57:41.000Z

2022-03-29T01:55:20.000Z

Python_Stock/Candlestick_Patterns/Candlestick_Takuri.ipynb

chunsj/Stock_Analysis_For_Quant

5f28ef9537885a695245d26f3010592a29d45a34

[ "MIT" ]

5

2020-04-29T16:54:30.000Z

2022-02-10T02:57:30.000Z

Python_Stock/Candlestick_Patterns/Candlestick_Takuri.ipynb

chunsj/Stock_Analysis_For_Quant

5f28ef9537885a695245d26f3010592a29d45a34

[ "MIT" ]

286

2019-08-04T10:37:58.000Z

2022-03-28T06:31:56.000Z

204.090418

46,045

0.841011

[ [ [ "# Candlestick Takuri (Dragonfly Doji with very long lower shadow)", "_____no_output_____" ], [ "https://patternswizard.com/takuri-candlestick-pattern/", "_____no_output_____" ] ], [ [ "import numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport talib\n\nimport warnings\nwarnings.filterwarnings(\"ignore\")\n\n# yahoo finance is used to fetch data \nimport yfinance as yf\nyf.pdr_override()", "_____no_output_____" ], [ "# input\nsymbol = 'AMD'\nstart = '2020-01-01'\nend = '2021-10-22'\n\n# Read data \ndf = yf.download(symbol,start,end)\n\n# View Columns\ndf.head()", "[*********************100%***********************] 1 of 1 completed\n" ] ], [ [ "## Candlestick with Takuri (Dragonfly Doji with very long lower shadow)", "_____no_output_____" ] ], [ [ "from matplotlib import dates as mdates\nimport datetime as dt\n\ndfc = df.copy()\ndfc['VolumePositive'] = dfc['Open'] < dfc['Adj Close']\n#dfc = dfc.dropna()\ndfc = dfc.reset_index()\ndfc['Date'] = pd.to_datetime(dfc['Date'])\ndfc['Date'] = dfc['Date'].apply(mdates.date2num)\ndfc.head()", "_____no_output_____" ], [ "from mplfinance.original_flavor import candlestick_ohlc\n\nfig = plt.figure(figsize=(14,10))\nax = plt.subplot(2, 1, 1)\ncandlestick_ohlc(ax,dfc.values, width=0.5, colorup='g', colordown='r', alpha=1.0)\nax.xaxis_date()\nax.xaxis.set_major_formatter(mdates.DateFormatter('%d-%m-%Y'))\nax.grid(True, which='both')\nax.minorticks_on()\naxv = ax.twinx()\ncolors = dfc.VolumePositive.map({True: 'g', False: 'r'})\naxv.bar(dfc.Date, dfc['Volume'], color=colors, alpha=0.4)\naxv.axes.yaxis.set_ticklabels([])\naxv.set_ylim(0, 3*df.Volume.max())\nax.set_title('Stock '+ symbol +' Closing Price')\nax.set_ylabel('Price')", "_____no_output_____" ], [ "takuri = talib.CDLTAKURI(df['Open'], df['High'], df['Low'], df['Close'])\n\ntakuri = takuri[takuri != 0]", "_____no_output_____" ], [ "df['takuri'] = talib.CDLTAKURI(df['Open'], df['High'], df['Low'], df['Close'])", "_____no_output_____" ], [ "df.loc[df['takuri'] !=0]", "_____no_output_____" ], [ "df['Adj Close'].loc[df['takuri'] !=0]", "_____no_output_____" ], [ "df['takuri'].loc[df['takuri'] !=0].index", "_____no_output_____" ], [ "takuri", "_____no_output_____" ], [ "takuri.index", "_____no_output_____" ], [ "df", "_____no_output_____" ], [ "fig = plt.figure(figsize=(20,16))\nax = plt.subplot(2, 1, 1)\ncandlestick_ohlc(ax,dfc.values, width=0.5, colorup='g', colordown='r', alpha=1.0)\nax.xaxis_date()\nax.xaxis.set_major_formatter(mdates.DateFormatter('%d-%m-%Y'))\nax.grid(True, which='both')\nax.minorticks_on()\naxv = ax.twinx()\nax.plot_date(df['Adj Close'].loc[df['takuri'] !=0].index, df['Adj Close'].loc[df['takuri'] !=0],\n 'Dc', # marker style 'o', color 'g'\n fillstyle='none', # circle is not filled (with color)\n ms=10.0) \ncolors = dfc.VolumePositive.map({True: 'g', False: 'r'})\naxv.bar(dfc.Date, dfc['Volume'], color=colors, alpha=0.4)\naxv.axes.yaxis.set_ticklabels([])\naxv.set_ylim(0, 3*df.Volume.max())\nax.set_title('Stock '+ symbol +' Closing Price')\nax.set_ylabel('Price')", "_____no_output_____" ] ], [ [ "## Plot Certain dates", "_____no_output_____" ] ], [ [ "df = df['2021-06-01':'2021-07-01']\ndfc = df.copy()\ndfc['VolumePositive'] = dfc['Open'] < dfc['Adj Close']\n#dfc = dfc.dropna()\ndfc = dfc.reset_index()\ndfc['Date'] = pd.to_datetime(dfc['Date'])\ndfc['Date'] = dfc['Date'].apply(mdates.date2num)\ndfc.head()", "_____no_output_____" ], [ "fig = plt.figure(figsize=(20,16))\nax = plt.subplot(2, 1, 1)\nax.set_facecolor('white')\ncandlestick_ohlc(ax,dfc.values, width=0.5, colorup='grey', colordown='black', alpha=1.0)\nax.xaxis_date()\nax.xaxis.set_major_formatter(mdates.DateFormatter('%d-%m-%Y'))\n#ax.grid(True, which='both')\n#ax.minorticks_on()\naxv = ax.twinx()\nax.plot_date(df['Adj Close'].loc[df['takuri'] !=0].index, df['Adj Close'].loc[df['takuri'] !=0],\n 'pb', # marker style 'o', color 'g'\n fillstyle='none', # circle is not filled (with color)\n ms=25.0) \ncolors = dfc.VolumePositive.map({True: 'grey', False: 'black'})\naxv.bar(dfc.Date, dfc['Volume'], color=colors, alpha=0.4)\naxv.axes.yaxis.set_ticklabels([])\naxv.set_ylim(0, 3*df.Volume.max())\nax.set_title('Stock '+ symbol +' Closing Price')\nax.set_ylabel('Price')", "_____no_output_____" ] ], [ [ "# Highlight Candlestick", "_____no_output_____" ] ], [ [ "from matplotlib.dates import date2num\nfrom datetime import datetime\n\nfig = plt.figure(figsize=(20,16))\nax = plt.subplot(2, 1, 1)\ncandlestick_ohlc(ax,dfc.values, width=0.5, colorup='g', colordown='r', alpha=1.0)\nax.xaxis_date()\nax.xaxis.set_major_formatter(mdates.DateFormatter('%d-%m-%Y'))\n#ax.grid(True, which='both')\n#ax.minorticks_on()\naxv = ax.twinx()\nax.axvspan(date2num(datetime(2021,6,13)), date2num(datetime(2021,6,15)), \n label=\"Takuri Bullish\",color=\"green\", alpha=0.3)\nax.legend()\ncolors = dfc.VolumePositive.map({True: 'g', False: 'r'})\naxv.bar(dfc.Date, dfc['Volume'], color=colors, alpha=0.4)\naxv.axes.yaxis.set_ticklabels([])\naxv.set_ylim(0, 3*df.Volume.max())\nax.set_title('Stock '+ symbol +' Closing Price')\nax.set_ylabel('Price')", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ] ]

ec7a63e74fdc7cda362c79dd29d1a22c19af9372

18,466

ipynb

Jupyter Notebook

course/Data Visualization/exercise-scatter-plots.ipynb

furyhawk/kaggle_practice

04bf045ae179db6a849fd2c2e833acc2e869f0f8

[ "MIT" ]

2

2021-11-22T09:21:25.000Z

2021-12-18T13:12:06.000Z

course/Data Visualization/exercise-scatter-plots.ipynb

furyhawk/kaggle_practice

04bf045ae179db6a849fd2c2e833acc2e869f0f8

[ "MIT" ]

null

course/Data Visualization/exercise-scatter-plots.ipynb

furyhawk/kaggle_practice

04bf045ae179db6a849fd2c2e833acc2e869f0f8

[ "MIT" ]

null

18,466

0.736814

[ [ [ "**This notebook is an exercise in the [Data Visualization](https://www.kaggle.com/learn/data-visualization) course. You can reference the tutorial at [this link](https://www.kaggle.com/alexisbcook/scatter-plots).**\n\n---\n", "_____no_output_____" ], [ "In this exercise, you will use your new knowledge to propose a solution to a real-world scenario. To succeed, you will need to import data into Python, answer questions using the data, and generate **scatter plots** to understand patterns in the data.\n\n## Scenario\n\nYou work for a major candy producer, and your goal is to write a report that your company can use to guide the design of its next product. Soon after starting your research, you stumble across this [very interesting dataset](https://fivethirtyeight.com/features/the-ultimate-halloween-candy-power-ranking/) containing results from a fun survey to crowdsource favorite candies.\n\n## Setup\n\nRun the next cell to import and configure the Python libraries that you need to complete the exercise.", "_____no_output_____" ] ], [ [ "import pandas as pd\npd.plotting.register_matplotlib_converters()\nimport matplotlib.pyplot as plt\n%matplotlib inline\nimport seaborn as sns\nprint(\"Setup Complete\")", "_____no_output_____" ] ], [ [ "The questions below will give you feedback on your work. Run the following cell to set up our feedback system.", "_____no_output_____" ] ], [ [ "# Set up code checking\nimport os\nif not os.path.exists(\"../input/candy.csv\"):\n os.symlink(\"../input/data-for-datavis/candy.csv\", \"../input/candy.csv\") \nfrom learntools.core import binder\nbinder.bind(globals())\nfrom learntools.data_viz_to_coder.ex4 import *\nprint(\"Setup Complete\")", "_____no_output_____" ] ], [ [ "## Step 1: Load the Data\n\nRead the candy data file into `candy_data`. Use the `\"id\"` column to label the rows.", "_____no_output_____" ] ], [ [ "# Path of the file to read\ncandy_filepath = \"../input/candy.csv\"\n\n# Fill in the line below to read the file into a variable candy_data\ncandy_data = pd.read_csv(candy_filepath, index_col='id')\n\n# Run the line below with no changes to check that you've loaded the data correctly\nstep_1.check()", "_____no_output_____" ], [ "# Lines below will give you a hint or solution code\n#step_1.hint()\n#step_1.solution()", "_____no_output_____" ] ], [ [ "## Step 2: Review the data\n\nUse a Python command to print the first five rows of the data.", "_____no_output_____" ] ], [ [ "# Print the first five rows of the data\ncandy_data.head() # Your code here", "_____no_output_____" ] ], [ [ "The dataset contains 83 rows, where each corresponds to a different candy bar. There are 13 columns:\n- `'competitorname'` contains the name of the candy bar. \n- the next **9** columns (from `'chocolate'` to `'pluribus'`) describe the candy. For instance, rows with chocolate candies have `\"Yes\"` in the `'chocolate'` column (and candies without chocolate have `\"No\"` in the same column).\n- `'sugarpercent'` provides some indication of the amount of sugar, where higher values signify higher sugar content.\n- `'pricepercent'` shows the price per unit, relative to the other candies in the dataset.\n- `'winpercent'` is calculated from the survey results; higher values indicate that the candy was more popular with survey respondents.\n\nUse the first five rows of the data to answer the questions below.", "_____no_output_____" ] ], [ [ "# Fill in the line below: Which candy was more popular with survey respondents:\n# '3 Musketeers' or 'Almond Joy'? (Please enclose your answer in single quotes.)\nmore_popular = '3 Musketeers'\n\n# Fill in the line below: Which candy has higher sugar content: 'Air Heads'\n# or 'Baby Ruth'? (Please enclose your answer in single quotes.)\nmore_sugar = 'Air Heads'\n\n# Check your answers\nstep_2.check()", "_____no_output_____" ], [ "# Lines below will give you a hint or solution code\n#step_2.hint()\n#step_2.solution()", "_____no_output_____" ] ], [ [ "## Step 3: The role of sugar\n\nDo people tend to prefer candies with higher sugar content? \n\n#### Part A\n\nCreate a scatter plot that shows the relationship between `'sugarpercent'` (on the horizontal x-axis) and `'winpercent'` (on the vertical y-axis). _Don't add a regression line just yet -- you'll do that in the next step!_", "_____no_output_____" ] ], [ [ "# Scatter plot showing the relationship between 'sugarpercent' and 'winpercent'\nsns.scatterplot(x=candy_data['sugarpercent'], y=candy_data['winpercent']) # Your code here\n\n# Check your answer\nstep_3.a.check()", "_____no_output_____" ], [ "# Lines below will give you a hint or solution code\n#step_3.a.hint()\nstep_3.a.solution_plot()", "_____no_output_____" ] ], [ [ "#### Part B\n\nDoes the scatter plot show a **strong** correlation between the two variables? If so, are candies with more sugar relatively more or less popular with the survey respondents?", "_____no_output_____" ] ], [ [ "#step_3.b.hint()", "_____no_output_____" ], [ "# Check your answer (Run this code cell to receive credit!)\nstep_3.b.solution()", "_____no_output_____" ] ], [ [ "## Step 4: Take a closer look\n\n#### Part A\n\nCreate the same scatter plot you created in **Step 3**, but now with a regression line!", "_____no_output_____" ] ], [ [ "# Scatter plot w/ regression line showing the relationship between 'sugarpercent' and 'winpercent'\nsns.regplot(x=candy_data['sugarpercent'], y=candy_data['winpercent']) # Your code here\n\n# Check your answer\nstep_4.a.check()", "_____no_output_____" ], [ "# Lines below will give you a hint or solution code\n#step_4.a.hint()\nstep_4.a.solution_plot()", "_____no_output_____" ] ], [ [ "#### Part B\n\nAccording to the plot above, is there a **slight** correlation between `'winpercent'` and `'sugarpercent'`? What does this tell you about the candy that people tend to prefer?", "_____no_output_____" ] ], [ [ "#step_4.b.hint()", "_____no_output_____" ], [ "# Check your answer (Run this code cell to receive credit!)\nstep_4.b.solution()", "_____no_output_____" ] ], [ [ "## Step 5: Chocolate!\n\nIn the code cell below, create a scatter plot to show the relationship between `'pricepercent'` (on the horizontal x-axis) and `'winpercent'` (on the vertical y-axis). Use the `'chocolate'` column to color-code the points. _Don't add any regression lines just yet -- you'll do that in the next step!_", "_____no_output_____" ] ], [ [ "# Scatter plot showing the relationship between 'pricepercent', 'winpercent', and 'chocolate'\nsns.scatterplot(x=candy_data['pricepercent'], y=candy_data['winpercent'], hue=candy_data['chocolate']) # Your code here\n\n# Check your answer\nstep_5.check()", "_____no_output_____" ], [ "# Lines below will give you a hint or solution code\n#step_5.hint()\nstep_5.solution_plot()", "_____no_output_____" ] ], [ [ "Can you see any interesting patterns in the scatter plot? We'll investigate this plot further by adding regression lines in the next step!\n\n## Step 6: Investigate chocolate\n\n#### Part A\n\nCreate the same scatter plot you created in **Step 5**, but now with two regression lines, corresponding to (1) chocolate candies and (2) candies without chocolate.", "_____no_output_____" ] ], [ [ "# Color-coded scatter plot w/ regression lines\nsns.lmplot(x=\"pricepercent\", y=\"winpercent\", hue=\"chocolate\", data=candy_data) # Your code here\n\n# Check your answer\nstep_6.a.check()", "_____no_output_____" ], [ "# Lines below will give you a hint or solution code\n#step_6.a.hint()\nstep_6.a.solution_plot()", "_____no_output_____" ] ], [ [ "#### Part B\n\nUsing the regression lines, what conclusions can you draw about the effects of chocolate and price on candy popularity?", "_____no_output_____" ] ], [ [ "#step_6.b.hint()", "_____no_output_____" ], [ "# Check your answer (Run this code cell to receive credit!)\nstep_6.b.solution()", "_____no_output_____" ] ], [ [ "## Step 7: Everybody loves chocolate.\n\n#### Part A\n\nCreate a categorical scatter plot to highlight the relationship between `'chocolate'` and `'winpercent'`. Put `'chocolate'` on the (horizontal) x-axis, and `'winpercent'` on the (vertical) y-axis.", "_____no_output_____" ] ], [ [ "# Scatter plot showing the relationship between 'chocolate' and 'winpercent'\nsns.swarmplot(x=candy_data['chocolate'], y=candy_data['winpercent']) # Your code here\n\n# Check your answer\nstep_7.a.check()", "_____no_output_____" ], [ "# Lines below will give you a hint or solution code\n#step_7.a.hint()\nstep_7.a.solution_plot()", "_____no_output_____" ] ], [ [ "#### Part B\n\nYou decide to dedicate a section of your report to the fact that chocolate candies tend to be more popular than candies without chocolate. Which plot is more appropriate to tell this story: the plot from **Step 6**, or the plot from **Step 7**?", "_____no_output_____" ] ], [ [ "#step_7.b.hint()", "_____no_output_____" ], [ "# Check your answer (Run this code cell to receive credit!)\nstep_7.b.solution()", "_____no_output_____" ] ], [ [ "## Keep going\n\nExplore **[histograms and density plots](https://www.kaggle.com/alexisbcook/distributions)**.", "_____no_output_____" ], [ "---\n\n\n\n\n*Have questions or comments? Visit the [course discussion forum](https://www.kaggle.com/learn/data-visualization/discussion) to chat with other learners.*", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ] ]

ec7a6cb408dc3ca938c5225e85a2e21b2ef738a5

9,859

ipynb

Jupyter Notebook

components/gcp/dataproc/submit_sparksql_job/sample.ipynb

ccrrvvaa/pipelines

935a9b5ba5057bc9801fee87ed17c03c2907ec85

[ "Apache-2.0" ]

null

components/gcp/dataproc/submit_sparksql_job/sample.ipynb

ccrrvvaa/pipelines

935a9b5ba5057bc9801fee87ed17c03c2907ec85

[ "Apache-2.0" ]

21

2021-03-02T01:44:01.000Z

2022-03-21T14:34:03.000Z

components/gcp/dataproc/submit_sparksql_job/sample.ipynb

ccrrvvaa/pipelines

935a9b5ba5057bc9801fee87ed17c03c2907ec85

[ "Apache-2.0" ]

null

37.773946

329

0.622071

[ [ [ "# Name\nData preparation using SparkSQL on YARN with Cloud Dataproc\n\n# Label\nCloud Dataproc, GCP, Cloud Storage, YARN, SparkSQL, Kubeflow, pipelines, components \n\n# Summary\nA Kubeflow Pipeline component to prepare data by submitting a SparkSql job on YARN to Cloud Dataproc.\n\n# Details\n\n## Intended use\nUse the component to run an Apache SparkSql job as one preprocessing step in a Kubeflow Pipeline.\n\n## Runtime arguments\nArgument| Description | Optional | Data type| Accepted values| Default |\n:--- | :---------- | :--- | :------- | :------ | :------\nproject_id | The ID of the Google Cloud Platform (GCP) project that the cluster belongs to. | No| GCPProjectID | | |\nregion | The Cloud Dataproc region to handle the request. | No | GCPRegion|\ncluster_name | The name of the cluster to run the job. | No | String| | |\nqueries | The queries to execute the SparkSQL job. Specify multiple queries in one string by separating them with semicolons. You do not need to terminate queries with semicolons. | Yes | List | | None | \nquery_file_uri | The HCFS URI of the script that contains the SparkSQL queries.| Yes | GCSPath | | None |\nscript_variables | Mapping of the query’s variable names to their values (equivalent to the SparkSQL command: SET name=\"value\";).| Yes| Dict | | None |\nsparksql_job | The payload of a [SparkSqlJob](https://cloud.google.com/dataproc/docs/reference/rest/v1/SparkSqlJob). | Yes | Dict | | None |\njob | The payload of a [Dataproc job](https://cloud.google.com/dataproc/docs/reference/rest/v1/projects.regions.jobs). | Yes | Dict | | None |\nwait_interval | The number of seconds to pause between polling the operation. | Yes |Integer | | 30 |\n\n## Output\nName | Description | Type\n:--- | :---------- | :---\njob_id | The ID of the created job. | String\n\n## Cautions & requirements\nTo use the component, you must:\n* Set up a GCP project by following this [guide](https://cloud.google.com/dataproc/docs/guides/setup-project).\n* [Create a new cluster](https://cloud.google.com/dataproc/docs/guides/create-cluster).\n* The component can authenticate to GCP. Refer to [Authenticating Pipelines to GCP](https://www.kubeflow.org/docs/gke/authentication-pipelines/) for details.\n* Grant the Kubeflow user service account the role `roles/dataproc.editor` on the project.\n\n## Detailed Description\nThis component creates a Pig job from [Dataproc submit job REST API](https://cloud.google.com/dataproc/docs/reference/rest/v1/projects.regions.jobs/submit).\n\nFollow these steps to use the component in a pipeline:\n1. Install the Kubeflow Pipeline SDK:", "_____no_output_____" ] ], [ [ "%%capture --no-stderr\n\n!pip3 install kfp --upgrade", "_____no_output_____" ] ], [ [ "2. Load the component using KFP SDK", "_____no_output_____" ] ], [ [ "import kfp.components as comp\n\ndataproc_submit_sparksql_job_op = comp.load_component_from_url(\n 'https://raw.githubusercontent.com/kubeflow/pipelines/1.1.0-alpha.1/components/gcp/dataproc/submit_sparksql_job/component.yaml')\nhelp(dataproc_submit_sparksql_job_op)", "_____no_output_____" ] ], [ [ "### Sample\n\nNote: The following sample code works in an IPython notebook or directly in Python code. See the sample code below to learn how to execute the template.\n\n#### Setup a Dataproc cluster\n[Create a new Dataproc cluster](https://cloud.google.com/dataproc/docs/guides/create-cluster) (or reuse an existing one) before running the sample code.\n\n#### Prepare a SparkSQL job\nEither put your SparkSQL queries in the `queires` list, or upload your SparkSQL queries into a file to a Cloud Storage bucket and then enter the Cloud Storage bucket’s path in `query_file_uri`. In this sample, we will use a hard coded query in the `queries` list to select data from a public CSV file from Cloud Storage.\n\nFor more details about Spark SQL, see [Spark SQL, DataFrames and Datasets Guide](https://spark.apache.org/docs/latest/sql-programming-guide.html)\n\n#### Set sample parameters", "_____no_output_____" ] ], [ [ "PROJECT_ID = '<Please put your project ID here>'\nCLUSTER_NAME = '<Please put your existing cluster name here>'\nREGION = 'us-central1'\nQUERY = '''\nDROP TABLE IF EXISTS natality_csv;\nCREATE EXTERNAL TABLE natality_csv (\n source_year BIGINT, year BIGINT, month BIGINT, day BIGINT, wday BIGINT,\n state STRING, is_male BOOLEAN, child_race BIGINT, weight_pounds FLOAT,\n plurality BIGINT, apgar_1min BIGINT, apgar_5min BIGINT,\n mother_residence_state STRING, mother_race BIGINT, mother_age BIGINT,\n gestation_weeks BIGINT, lmp STRING, mother_married BOOLEAN,\n mother_birth_state STRING, cigarette_use BOOLEAN, cigarettes_per_day BIGINT,\n alcohol_use BOOLEAN, drinks_per_week BIGINT, weight_gain_pounds BIGINT,\n born_alive_alive BIGINT, born_alive_dead BIGINT, born_dead BIGINT,\n ever_born BIGINT, father_race BIGINT, father_age BIGINT,\n record_weight BIGINT\n)\nROW FORMAT DELIMITED FIELDS TERMINATED BY ','\nLOCATION 'gs://public-datasets/natality/csv';\n\nSELECT * FROM natality_csv LIMIT 10;'''\nEXPERIMENT_NAME = 'Dataproc - Submit SparkSQL Job'", "_____no_output_____" ] ], [ [ "#### Example pipeline that uses the component", "_____no_output_____" ] ], [ [ "import kfp.dsl as dsl\nimport json\[email protected](\n name='Dataproc submit SparkSQL job pipeline',\n description='Dataproc submit SparkSQL job pipeline'\n)\ndef dataproc_submit_sparksql_job_pipeline(\n project_id = PROJECT_ID, \n region = REGION,\n cluster_name = CLUSTER_NAME,\n queries = json.dumps([QUERY]),\n query_file_uri = '',\n script_variables = '', \n sparksql_job='', \n job='', \n wait_interval='30'\n):\n dataproc_submit_sparksql_job_op(\n project_id=project_id, \n region=region, \n cluster_name=cluster_name, \n queries=queries, \n query_file_uri=query_file_uri,\n script_variables=script_variables, \n sparksql_job=sparksql_job, \n job=job, \n wait_interval=wait_interval)\n ", "_____no_output_____" ] ], [ [ "#### Compile the pipeline", "_____no_output_____" ] ], [ [ "pipeline_func = dataproc_submit_sparksql_job_pipeline\npipeline_filename = pipeline_func.__name__ + '.zip'\nimport kfp.compiler as compiler\ncompiler.Compiler().compile(pipeline_func, pipeline_filename)", "_____no_output_____" ] ], [ [ "#### Submit the pipeline for execution", "_____no_output_____" ] ], [ [ "#Specify pipeline argument values\narguments = {}\n\n#Get or create an experiment and submit a pipeline run\nimport kfp\nclient = kfp.Client()\nexperiment = client.create_experiment(EXPERIMENT_NAME)\n\n#Submit a pipeline run\nrun_name = pipeline_func.__name__ + ' run'\nrun_result = client.run_pipeline(experiment.id, run_name, pipeline_filename, arguments)", "_____no_output_____" ] ], [ [ "## References\n* [Spark SQL, DataFrames and Datasets Guide](https://spark.apache.org/docs/latest/sql-programming-guide.html)\n* [SparkSqlJob](https://cloud.google.com/dataproc/docs/reference/rest/v1/SparkSqlJob)\n* [Cloud Dataproc job](https://cloud.google.com/dataproc/docs/reference/rest/v1/projects.regions.jobs)\n\n\n## License\nBy deploying or using this software you agree to comply with the [AI Hub Terms of Service](https://aihub.cloud.google.com/u/0/aihub-tos) and the [Google APIs Terms of Service](https://developers.google.com/terms/). To the extent of a direct conflict of terms, the AI Hub Terms of Service will control.", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ] ]

ec7a6d017a1ae73284e0525cb758c9184a1d7589

188,038

ipynb

Jupyter Notebook

notebooks/ICM_rubiks_simple.ipynb

tik0/2019_QUT_DAAD_DEMO

035634fa27280e36354c72b453e214aa35fdb530

[ "BSD-3-Clause" ]

null

notebooks/ICM_rubiks_simple.ipynb

tik0/2019_QUT_DAAD_DEMO

035634fa27280e36354c72b453e214aa35fdb530

[ "BSD-3-Clause" ]

null

notebooks/ICM_rubiks_simple.ipynb

tik0/2019_QUT_DAAD_DEMO

035634fa27280e36354c72b453e214aa35fdb530

[ "BSD-3-Clause" ]

null

318.708475

84,972

0.916309

[ [ [ "'''This script demonstrates ICM on image data\n\n #Reference\n Curiosity-driven Exploration by Self-supervised Prediction\n https://arxiv.org/abs/1705.05363\n'''\nimport warnings\nimport sys, os\nfrom glob import glob\nimport numpy as np\nnp.random.seed(0)\nimport matplotlib.pyplot as plt\nfrom mpl_toolkits.mplot3d import Axes3D\nimport matplotlib.mlab as mlab\nimport matplotlib.image as mpimg\n#plt.rcParams[\"figure.figsize\"] = [20,20]\nfrom scipy.stats import norm\nfrom keras.utils.vis_utils import model_to_dot\nfrom keras.utils import plot_model\nfrom keras.layers import Input, Dense, Lambda, Flatten, Reshape, Layer, Dropout\nfrom keras.layers import Conv2D, Conv2DTranspose\nfrom keras.models import Model\nfrom keras import backend as K\nfrom keras import metrics\nimport keras\nfrom skimage.transform import resize\nfrom PIL import Image\nimport csv\nfrom sklearn import preprocessing\n\ndata_dir = 'img_0'\nnum_data = 1710", "Using TensorFlow backend.\n" ], [ "# Read the poses and observations\nwith open(data_dir + \"/lookup_pitch_yaw.csv\", \"r\") as f:\n reader = csv.reader(f, delimiter=',')\n pose_pitch_yaw = np.asarray([[float(x[0]), float(x[1])] for x in list(reader)[1:]])\n f.close()\n\ndef get_iterator(batch_size):\n from keras.preprocessing.image import ImageDataGenerator\n train_datagen = ImageDataGenerator(\n rescale=1./255,\n shear_range=0,\n zoom_range=0,\n horizontal_flip=False,\n width_shift_range=0.0, # randomly shift images horizontally (fraction of total width)\n height_shift_range=0.0) # randomly shift images vertically (fraction of total height))\n\n train_generator = train_datagen.flow_from_directory(data_dir, interpolation='nearest',\n color_mode='rgb', shuffle=False, seed=None,\n target_size=(60, 80),\n batch_size=batch_size,\n #save_to_dir='img_0_augmented',\n class_mode=None)\n return train_generator\nX_train = get_iterator(num_data).next()", "Found 1710 images belonging to 1 classes.\n" ], [ "# Show some scaled image\ngenerator = get_iterator(1)\nidx = 1000\nfor _ in range(idx+1):\n x = generator.next()\nprint(x[0].shape)\nplt.imshow(x[0])\nplt.title(str(pose_pitch_yaw[idx]))\nplt.show()", "Found 1710 images belonging to 1 classes.\n(60, 80, 3)\n" ], [ "# Create the action/perception tuples\nnum_tuples = int(10 * num_data)\n\n# Creates actions for given start and end angles\ndef angle_diff(s, e):\n return (1 + np.concatenate([np.sin(e - s), np.cos(e - s)], axis=1)) / 2.0\n #return np.concatenate([e - s, e - s], axis=1)\n\n# example for angle_diff\n#s = np.asarray([[0.0, 0.0], [0.0, 0.0], [np.pi/2, 0.0]])\n#e = np.asarray([[np.pi/2, -np.pi/2], [np.pi/2, -np.pi/2], [np.pi/2, -np.pi/2]])\n#print(angle_diff(s, e))\n\nnp.random.seed(0)\ns1_idx = np.random.randint(0, high=num_data, size=num_tuples)\ns2_idx = np.random.randint(0, high=num_data, size=num_tuples)\n\nX_train_actions = angle_diff(pose_pitch_yaw[s1_idx], pose_pitch_yaw[s2_idx])\nX_train_s1 = X_train[s1_idx]\nX_train_s2 = X_train[s2_idx]\n\n# Show a tuple\nf, ax = plt.subplots(1,2,figsize=(7.5,7.5), dpi=96)\nax[0].imshow(X_train_s1[0])\nax[1].imshow(X_train_s2[0])\nax[0].set_title('s_1')\nax[1].set_title('s_2')\nplt.show()\nprint(\"Pose s_1: \", pose_pitch_yaw[s1_idx[0]])\nprint(\"Pose s_2: \", pose_pitch_yaw[s2_idx[0]])\nprint(\"Action s_1 -> s_2: \", X_train_actions[0])", "_____no_output_____" ], [ "# input image dimensions and config\nbatch_size = 128\nepochs = 100\nimage_rows_cols_chns = (60, 80, 3)\noriginal_dim = np.prod(image_rows_cols_chns)\naction_dim = 4\nfeat_dim = 2\nintermediate_dim = 32\nintermediate_dim_2 = int(intermediate_dim / 2.)\nbeta = 0.1 # in the paper beta := 0.2\n\n# Feature extractor\nfeat_trainable = True\nfeat_inp_s1 = Input(shape=(original_dim,), name=\"feat_inp_s1\")\nfeat_inp_s2 = Input(shape=(original_dim,), name=\"feat_inp_s2\")\nfeat_h1 = Dense(intermediate_dim, activation='elu', name=\"feat_h1\", trainable=feat_trainable)\nfeat_d1 = Dropout(rate=0.4, name=\"feat_d1\")\nfeat_h2 = Dense(intermediate_dim_2, activation='elu', name=\"feat_h2\", trainable=feat_trainable)\nfeat_d2 = Dropout(rate=0.4, name=\"feat_d2\")\nfeat_out = Dense(feat_dim, name=\"feat_out\", trainable=feat_trainable)\nfeat_out_bn = keras.layers.BatchNormalization()\n\n# w/o dropout\n#feat_model_s1 = Model(feat_inp_s1, feat_out(feat_h2(feat_h1(feat_inp_s1))), name=\"feat_model_s1\")\n#feat_model_s2 = Model(feat_inp_s2, feat_out(feat_h2(feat_h1(feat_inp_s2))), name=\"feat_model_s2\")\n# w/ dropout\n#feat_model_s1 = Model(feat_inp_s1, feat_out(feat_d1(feat_h2(feat_d1(feat_h1(feat_inp_s1))))), name=\"feat_model_s1\")\n#feat_model_s2 = Model(feat_inp_s2, feat_out(feat_d1(feat_h2(feat_d1(feat_h1(feat_inp_s2))))), name=\"feat_model_s2\")\n# w/ dropout and batchnorm\nfeat_model_s1 = Model(feat_inp_s1, feat_out_bn(feat_out(feat_d1(feat_h2(feat_d1(feat_h1(feat_inp_s1)))))), name=\"feat_model_s1\")\nfeat_model_s2 = Model(feat_inp_s2, feat_out_bn(feat_out(feat_d1(feat_h2(feat_d1(feat_h1(feat_inp_s2)))))), name=\"feat_model_s2\")\n\n# Inverse model (action predictor)\nfwd_feat_inp_s1 = Input(shape=(feat_dim,), name=\"fwd_feat_inp_s1\")\nfwd_feat_inp_s2 = Input(shape=(feat_dim,), name=\"fwd_feat_inp_s2\")\ninv_concat = keras.layers.Concatenate(axis=-1, name=\"inv_concat\")\ninv_h1 = Dense(intermediate_dim_2, activation='elu', name=\"inv_h1\")\ninv_d1 = Dropout(rate=0.4, name=\"inv_d1\")\ninv_h2 = Dense(intermediate_dim_2, activation='elu', name=\"inv_h2\")\ninv_d2 = Dropout(rate=0.4, name=\"inv_d2\")\ninv_out = Dense(action_dim, name=\"inv_out\")\n\n# w/o dropout\n#inv_model = Model([fwd_feat_inp_s1, fwd_feat_inp_s2], inv_out(inv_h2(inv_h1(inv_concat([fwd_feat_inp_s1, fwd_feat_inp_s2])))), name=\"inv_model\")\n# w/ dropout\ninv_model = Model([fwd_feat_inp_s1, fwd_feat_inp_s2],\n inv_out(inv_d2(inv_h2(inv_d1(inv_h1(inv_concat([fwd_feat_inp_s1, fwd_feat_inp_s2])))))), name=\"inv_model\")\n\n# Forward model (feature predictor)\nfwd_trainable = True\nfwd_inp = Input(shape=(action_dim,), name=\"a\")\nfwd_concat = keras.layers.Concatenate(axis=-1, name=\"fwd_concat\")\nfwd_h1 = Dense(intermediate_dim_2, activation='elu', name=\"fwd_h1\", trainable=fwd_trainable)\nfwd_d1 = Dropout(rate=0.4, name=\"fwd_d1\")\nfwd_h2 = Dense(intermediate_dim_2, activation='elu', name=\"fwd_h2\", trainable=fwd_trainable)\nfwd_d2 = Dropout(rate=0.4, name=\"fwd_d2\")\nfwd_out = Dense(feat_dim, name=\"fwd_out\", trainable=fwd_trainable)\n\n# w/o dropout\nfwd_model = Model([fwd_inp, fwd_feat_inp_s1], fwd_out(fwd_h2(fwd_h1(fwd_concat([fwd_inp, fwd_feat_inp_s1])))), name=\"fwd_model\")\n# w/ dropout\n#fwd_model = Model([fwd_inp, fwd_feat_inp_s1], fwd_out(fwd_d2(fwd_h2(fwd_d1(fwd_h1(fwd_concat([fwd_inp, fwd_feat_inp_s1])))))), name=\"fwd_model\")\n\n# Define the losses\n\n# loss layer\nclass InverseLossLayer(Layer):\n def __init__(self, **kwargs):\n self.is_placeholder = True\n super(InverseLossLayer, self).__init__(**kwargs)\n\n def call(self, inputs):\n fwd_inp = inputs[0]\n inv_out = inputs[1]\n loss = (1. - beta) * K.mean(keras.losses.mse(fwd_inp, inv_out))\n self.add_loss(loss, inputs=inputs)\n return loss\nclass ForwardLossLayer(Layer):\n def __init__(self, **kwargs):\n self.is_placeholder = True\n super(ForwardLossLayer, self).__init__(**kwargs)\n\n def call(self, inputs):\n feat_out_s2 = inputs[0]\n fwd_out = inputs[1]\n loss = beta * 0.5 * K.mean(keras.losses.mse(feat_out_s2, fwd_out))\n self.add_loss(loss, inputs=inputs)\n return loss\n\n# Build the complete model\nfwd_loss = ForwardLossLayer(name=\"fwd_loss\")([feat_model_s2(feat_inp_s2), fwd_model([fwd_inp, feat_model_s1(feat_inp_s1)])])\ninv_loss = InverseLossLayer(name=\"inv_loss\")([fwd_inp, inv_model([feat_model_s1(feat_inp_s1), feat_model_s2(feat_inp_s2)])])\nicm_model = Model([feat_inp_s1, feat_inp_s2, fwd_inp], [fwd_loss, inv_loss], name=\"icm_model\")\n\n# Create a history class which stores the ouputs of the loss layer after every epoch\nclass History(keras.callbacks.Callback):\n def on_train_begin(self, logs={}):\n self.history = {}\n self.epoch = []\n\n def on_epoch_end(self, epoch, logs={}):\n self.epoch.append(epoch)\n # Get the output layer names (we only have loss layers as output)\n output_layer_names = [output_layer.name for output_layer in self.model.output]\n # Get the mean error over all batches in this epoch\n num_eval = num_data\n output_layer_values = np.mean(self.model.predict([X_train_s1_flatten[:num_eval], X_train_s2_flatten[:num_eval], X_train_actions[:num_eval]]), axis = 1)\n #output_layer_values = self.model.predict([X_train_s1_flatten, X_train_s2_flatten, X_train_actions], batch_size = X_train_s1_flatten.shape[0])\n # Store it to the history\n for k, v in zip(output_layer_names, output_layer_values):\n self.history.setdefault(k, []).append(v)\n return\n\n# Compile and show it\nicm_model.compile(keras.optimizers.Adam(lr=0.0001), loss=None)\nicm_model.summary()\nplot_model(icm_model, to_file='/tmp/ICM.png', show_shapes=True, show_layer_names=True)\nf, ax = plt.subplots(1,1,figsize=(10, 10), dpi=150)\nax.imshow(mpimg.imread('/tmp/ICM.png'))\nplt.show()", "WARNING:tensorflow:From /usr/local/lib/python3.5/dist-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\nInstructions for updating:\nColocations handled automatically by placer.\nWARNING:tensorflow:From /usr/local/lib/python3.5/dist-packages/keras/backend/tensorflow_backend.py:3445: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.\nInstructions for updating:\nPlease use `rate` instead of `keep_prob`. Rate should be set to `rate = 1 - keep_prob`.\n__________________________________________________________________________________________________\nLayer (type) Output Shape Param # Connected to \n==================================================================================================\nfeat_inp_s2 (InputLayer) (None, 14400) 0 \n__________________________________________________________________________________________________\nfeat_inp_s1 (InputLayer) (None, 14400) 0 \n__________________________________________________________________________________________________\nfeat_model_s2 (Model) (None, 2) 461402 feat_inp_s2[0][0] \n feat_inp_s2[0][0] \n__________________________________________________________________________________________________\na (InputLayer) (None, 4) 0 \n__________________________________________________________________________________________________\nfeat_model_s1 (Model) (None, 2) 461402 feat_inp_s1[0][0] \n feat_inp_s1[0][0] \n__________________________________________________________________________________________________\nfwd_model (Model) (None, 2) 418 a[0][0] \n feat_model_s1[1][0] \n__________________________________________________________________________________________________\ninv_model (Model) (None, 4) 420 feat_model_s1[2][0] \n feat_model_s2[2][0] \n__________________________________________________________________________________________________\nfwd_loss (ForwardLossLayer) [(None, 2), (None, 2 0 feat_model_s2[1][0] \n fwd_model[1][0] \n__________________________________________________________________________________________________\ninv_loss (InverseLossLayer) [(None, 4), (None, 4 0 a[0][0] \n inv_model[1][0] \n==================================================================================================\nTotal params: 462,240\nTrainable params: 462,236\nNon-trainable params: 4\n__________________________________________________________________________________________________\n" ], [ "# Train\noutput_history = History()\nX_train_s1_flatten = X_train_s1.reshape((X_train_s1.shape[0], np.prod(X_train_s1.shape[1:])))\nX_train_s2_flatten = X_train_s2.reshape((X_train_s2.shape[0], np.prod(X_train_s2.shape[1:])))\nh = icm_model.fit([X_train_s1_flatten,\n X_train_s2_flatten,\n X_train_actions],\n shuffle=True,\n epochs=epochs,\n batch_size=batch_size,\n verbose = 0,\n callbacks=[output_history])\n#print(h.history)\n#print(output_history.history)", "WARNING:tensorflow:From /usr/local/lib/python3.5/dist-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\nInstructions for updating:\nUse tf.cast instead.\n" ], [ "plt_key = [list(h.history.keys())[0], list(output_history.history.keys())[0], list(output_history.history.keys())[1]]\nplt_val = [h.history[plt_key[0]], output_history.history[plt_key[1]], output_history.history[plt_key[2]]]\nf, ax = plt.subplots(1,3, figsize=(13,4), dpi=96)\nfor idx in range(3):\n ax[idx].plot(plt_val[idx])\n ax[idx].set_title(plt_key[idx])\nplt.show()", "_____no_output_____" ], [ "# Check the output distribution of the feature extractor\nX_train_flatten = X_train.reshape((X_train.shape[0], np.prod(X_train.shape[1:])))\nX_train_feat = feat_model_s1.predict(X_train_flatten)\nc = np.concatenate(((pose_pitch_yaw - np.min(pose_pitch_yaw, axis=0)) / np.max(pose_pitch_yaw, axis=0), np.zeros((num_data,1))), axis=1)\nplt.scatter(X_train_feat[:,0], X_train_feat[:,1], c=c)\nplt.show()", "_____no_output_____" ], [ "# Check the inverse model (i.e. predicting the action from consecutive observations)\nidx = 3\npred_action = inv_model.predict([feat_model_s1.predict(X_train_s1_flatten[[idx]]),\n feat_model_s1.predict(X_train_s2_flatten[[idx]])])\nf, ax = plt.subplots(1,2,figsize=(7.5,7.5), dpi=96)\nax[0].imshow(X_train_s1[idx])\nax[1].imshow(X_train_s2[idx])\nax[0].set_title('s_1')\nax[1].set_title('s_2')\nplt.show()\nprint(\"True: \" + str(X_train_actions[idx]) + \"\\n vs. \\nPred: \" + str(pred_action))\nprint(\"\\nDiff: \" + str(np.mean((X_train_actions[idx] - pred_action)**2)))", "_____no_output_____" ], [ "# Check the forward model (i.e. predicting the next obs. from current action and obs.)\nidx = 2\npred_obs = fwd_model.predict([X_train_actions[[idx]],\n feat_model_s1.predict(X_train_s1_flatten[[idx]])])\npred_obs_gt = feat_model_s1.predict(X_train_s2_flatten[[idx]])\nprint(\"True: \" + str(pred_obs_gt) + \"\\n vs. \\nPred: \" + str(pred_obs))\nprint(\"\\nReward: \" + str(np.mean((pred_obs_gt - pred_obs)**2)))", "True: [[ 0.26546863 -0.30335593]]\n vs. \nPred: [[ 0.21825394 -0.24688743]]\n\nReward: 0.0027089592\n" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7a7b13117948a4cb13ff068910f638ba683859

6,040

ipynb

Jupyter Notebook

src/Commenting_Bot.ipynb

zharmedia386/Automate_with_Python

bc263ee6dec564c908df29eb9c0ad03c299128e1

[ "MIT" ]

null

src/Commenting_Bot.ipynb

zharmedia386/Automate_with_Python

bc263ee6dec564c908df29eb9c0ad03c299128e1

[ "MIT" ]

null

src/Commenting_Bot.ipynb

zharmedia386/Automate_with_Python

bc263ee6dec564c908df29eb9c0ad03c299128e1

[ "MIT" ]

null

24.754098

143

0.549834

[ [ [ "# Commenting Bot to Particular Instagram Accounts", "_____no_output_____" ] ], [ [ "from selenium import webdriver\nfrom selenium.webdriver.common.keys import Keys\nfrom selenium.webdriver.support import expected_conditions as EC\nfrom selenium.webdriver.common.by import By\nfrom selenium.webdriver.support.wait import WebDriverWait\nimport time\nimport random", "_____no_output_____" ] ], [ [ "## Step 1 - Chrome Driver", "_____no_output_____" ] ], [ [ "driver = webdriver.Chrome('D:\\chromedriver_win32\\chromedriver.exe')\ndriver.get('https://www.instagram.com')", "_____no_output_____" ] ], [ [ "## Step 2 - Log in by Username and Password", "_____no_output_____" ] ], [ [ "time.sleep(3)\nusername = driver.find_element_by_css_selector(\"input[name='username']\").send_keys('My_Username')\npassword = driver.find_element_by_xpath(\"//input[@name='password']\").send_keys('My_Password')\ntime.sleep(1)\nbutton = WebDriverWait(driver,10).until(EC.element_to_be_clickable((By.XPATH, \"//button[@type='submit']\"))).click()", "_____no_output_____" ] ], [ [ "## Step 3 - Handle Alerts", "_____no_output_____" ] ], [ [ "time.sleep(3)\nnotNow_alert = WebDriverWait(driver,10).until(EC.element_to_be_clickable((By.XPATH, \"//button[contains(text(), 'Not Now')]\"))).click()", "_____no_output_____" ] ], [ [ "## Step 4 - Search Username", "_____no_output_____" ] ], [ [ "time.sleep(2)\nsearch_username = WebDriverWait(driver,10).until(EC.element_to_be_clickable((By.XPATH, \"//input[@placeholder='Search']\")))\nsearch_username.send_keys('m.azhar.alauddin')\ntime.sleep(3)\nsearch_username.send_keys(Keys.ENTER)\ntime.sleep(3)\nsearch_username.send_keys(Keys.ENTER)\ntime.sleep(3)", "_____no_output_____" ] ], [ [ "## Step 5 - Click Follow Button", "_____no_output_____" ] ], [ [ "# Optional\n# follow_button = WebDriverWait(driver,10).until(EC.element_to_be_clickable((By.XPATH, \"//button[contains(text(), 'Follow')]\")))\n# follow_button.click()", "_____no_output_____" ] ], [ [ "## Step 6 - Target Image Links", "_____no_output_____" ] ], [ [ "anchors = driver.find_elements_by_tag_name('a')\nanchors = [a.get_attribute('href') for a in anchors]\nanchors = [a for a in anchors if a.startswith(\"https://www.instagram.com/p/\")]\nanchors", "_____no_output_____" ] ], [ [ "## Step 7 - Redirect Page to the Particular Post", "_____no_output_____" ] ], [ [ "for a in anchors :\n driver.get(a)\n # Content\n greetings = [\"Hi!\", \"Hello!\", \"Hey!\", \"Heeey!\", \"Greetings!\"]\n desc = [\" I saw your photos - they are beautiful! wanna collab??\", \" That's cool bro\"]\n index_greet = random.randint(0, (len(greetings)-1))\n index_desc = random.randint(0, (len(desc)-1))\n # Send Message\n sendMessage = str(greetings[index_greet] + desc[index_desc])\n form = driver.find_element_by_tag_name(\"form\").click()\n time.sleep(2)\n commentBox = WebDriverWait(driver,10).until(EC.element_to_be_clickable((By.XPATH, \"//textarea[@placeholder='Add a comment…']\")))\n commentBox.send_keys(sendMessage)\n # Button Click\n time.sleep(1)\n buttonPost = WebDriverWait(driver,10).until(EC.element_to_be_clickable((By.XPATH, \"//button[contains(text(), 'Post')]\"))).click()", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

ec7a9398622449e6caed5095121c252c8db8115d

4,345

ipynb

Jupyter Notebook

module1-rnn-and-lstm/LS_DS_431_RNN_and_LSTM_Assignment.ipynb

martinclehman/DS-Unit-4-Sprint-3-Deep-Learning

2c7f8a856480c7610ac8664388dd6a71b11922ed

[ "MIT" ]

null

module1-rnn-and-lstm/LS_DS_431_RNN_and_LSTM_Assignment.ipynb

martinclehman/DS-Unit-4-Sprint-3-Deep-Learning

2c7f8a856480c7610ac8664388dd6a71b11922ed

[ "MIT" ]

null

module1-rnn-and-lstm/LS_DS_431_RNN_and_LSTM_Assignment.ipynb

martinclehman/DS-Unit-4-Sprint-3-Deep-Learning

2c7f8a856480c7610ac8664388dd6a71b11922ed

[ "MIT" ]

null

45.260417

318

0.671807

[ [ [ "<img align=\"left\" src=\"https://lever-client-logos.s3.amazonaws.com/864372b1-534c-480e-acd5-9711f850815c-1524247202159.png\" width=200>\n<br></br>\n<br></br>\n\n## *Data Science Unit 4 Sprint 3 Assignment 1*\n\n# Recurrent Neural Networks and Long Short Term Memory (LSTM)\n\n![Monkey at a typewriter](https://upload.wikimedia.org/wikipedia/commons/thumb/3/3c/Chimpanzee_seated_at_typewriter.jpg/603px-Chimpanzee_seated_at_typewriter.jpg)\n\nIt is said that [infinite monkeys typing for an infinite amount of time](https://en.wikipedia.org/wiki/Infinite_monkey_theorem) will eventually type, among other things, the complete works of Wiliam Shakespeare. Let's see if we can get there a bit faster, with the power of Recurrent Neural Networks and LSTM.\n\nThis text file contains the complete works of Shakespeare: https://www.gutenberg.org/files/100/100-0.txt\n\nUse it as training data for an RNN - you can keep it simple and train character level, and that is suggested as an initial approach.\n\nThen, use that trained RNN to generate Shakespearean-ish text. Your goal - a function that can take, as an argument, the size of text (e.g. number of characters or lines) to generate, and returns generated text of that size.\n\nNote - Shakespeare wrote an awful lot. It's OK, especially initially, to sample/use smaller data and parameters, so you can have a tighter feedback loop when you're trying to get things running. Then, once you've got a proof of concept - start pushing it more!", "_____no_output_____" ] ], [ [ "# TODO - Words, words, mere words, no matter from the heart.", "_____no_output_____" ] ], [ [ "# Resources and Stretch Goals", "_____no_output_____" ], [ "## Stretch goals:\n- Refine the training and generation of text to be able to ask for different genres/styles of Shakespearean text (e.g. plays versus sonnets)\n- Train a classification model that takes text and returns which work of Shakespeare it is most likely to be from\n- Make it more performant! Many possible routes here - lean on Keras, optimize the code, and/or use more resources (AWS, etc.)\n- Revisit the news example from class, and improve it - use categories or tags to refine the model/generation, or train a news classifier\n- Run on bigger, better data\n\n## Resources:\n- [The Unreasonable Effectiveness of Recurrent Neural Networks](https://karpathy.github.io/2015/05/21/rnn-effectiveness/) - a seminal writeup demonstrating a simple but effective character-level NLP RNN\n- [Simple NumPy implementation of RNN](https://github.com/JY-Yoon/RNN-Implementation-using-NumPy/blob/master/RNN%20Implementation%20using%20NumPy.ipynb) - Python 3 version of the code from \"Unreasonable Effectiveness\"\n- [TensorFlow RNN Tutorial](https://github.com/tensorflow/models/tree/master/tutorials/rnn) - code for training a RNN on the Penn Tree Bank language dataset\n- [4 part tutorial on RNN](http://www.wildml.com/2015/09/recurrent-neural-networks-tutorial-part-1-introduction-to-rnns/) - relates RNN to the vanishing gradient problem, and provides example implementation\n- [RNN training tips and tricks](https://github.com/karpathy/char-rnn#tips-and-tricks) - some rules of thumb for parameterizing and training your RNN", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown" ]

[ [ "markdown" ], [ "code" ], [ "markdown", "markdown" ] ]

ec7aa11a7b85491e574900de23cdaf3b6dc3a822

6,560

ipynb

Jupyter Notebook

automated_fire_mapping.ipynb

jdilger/drc-fire

8991f7c62a99acb16740c805ef29af09995370a6

[ "MIT" ]

null

automated_fire_mapping.ipynb

jdilger/drc-fire

8991f7c62a99acb16740c805ef29af09995370a6

[ "MIT" ]

null

automated_fire_mapping.ipynb

jdilger/drc-fire

8991f7c62a99acb16740c805ef29af09995370a6

[ "MIT" ]

null

31.690821

216

0.602896

[ [ [ "\nimport ee\n\nfrom fire_module_2 import step2\nfrom fire_module import step1\nee.Initialize()\n", "_____no_output_____" ] ], [ [ "The fire scripts are broken down into 3 steps:\n1. Prepare and export baseline spatial and temporal mean and standard deviation\n - This is specific to the year and land cover type\n2. Prepare and export NBR anomalies \n3. Export Annual burn map by land cover\n\n<img src=\"imgs/pipeline.png\" alt=\"example of workflow as a pipeline\" width=\"200\"/>", "_____no_output_____" ], [ "step1(analysisYear, geometry, cover, covername)\n\nClass for initializing the first steps in the burn anomalies mapping process and key input into the second step for generating burn maps.\n\nArgs:\n- paramtersIO (object): Parameters for cloud and shadow masking and dictionary of cover names and export paths.\n- analysisYear (int): The year to generate burn products.\n- geometry (ee.FeatureCollection): A feature collection of at least 1 geometry for the ROI. If there are multiple features in a collection the geometry is found with the .geometry() method.\n- cover (ee.Image): The land cover image that is described in paramterIO cover dictionary.\n- coverName (str): The land cover name to generate burn products for.\n ", "_____no_output_____" ] ], [ [ "# Note: use test geom if wanted to run test exports it much smaller (and thus faster)\ntest_geom = ee.FeatureCollection(\n \"projects/sig-misc-ee/assets/drc_fire/test_areas/test_area\")\nDRC_border = ee.FeatureCollection(\n \"projects/ee-karistenneson/assets/BurnedBiomass/DRC_Training/DRC_Border\")\n\n# cover for drc\ndrc_2000 = ee.Image('projects/sig-ee/FIRE/DRC/DIAF_2000forest')\n\n\n# drc cover name options:\n# Forests without dry forest\n# Dry forest or open forest\n\ncover = drc_2000\ncovername = \"Forests without dry forest\"\nyear = 2010\nregion = DRC_border\n\nfire = step1(year, region, cover, covername)", "_____no_output_____" ] ], [ [ "The preparing baseline step will take the longest to export. Test runs for DRC for 2016 typically finish ~12-19h.\n\nprepare_script1() returns an imagecollection of the baseline mean, and stddevs \n\nexport_image_collection takes the collection to export, and a function for how to export a specific collection in this case since we are making the baseline we pass in the export_baseline_landcover function.\n\noptionally: you can set test = True to export a single image, and change the export scale exportScale=Int\nexport_image_collection(image_collection, export function, test: bool = False, exportScale: int = 30 )", "_____no_output_____" ] ], [ [ "prep_baseline = fire.prepare_script1()\n\nfire.export_image_collection(\n prep_baseline, fire.export_baseline_landcover)", "_____no_output_____" ] ], [ [ "Next the anomalies are created for each analysis period for our analysis year. Test runs for DRC for 2016 typically finish in ~9-15h.\n\nIf you neeeded to restart the notebook that is ok, but make sure the fire object has be initiated for the anomalies that need to be processed.\n\nLikewise with the previous step test and exportScale can be adjusted", "_____no_output_____" ] ], [ [ "anomaly_col = fire.script1()\n\nfire.export_image_collection(\n anomaly_col, fire.export_nbr_anomalies)", "_____no_output_____" ] ], [ [ "Lastly the burn product cna be created for the analysis year using the 2nd step.\n\n\n main(alpha, pVal, year, cover, optional[expected_size])\n \n Generate yearly burn product from NBR anomalies that incorporates\n MODIS hotspots.\n\n Args:\n alpha (float): p-value threshold for identifying fires.\n pVal (str): A string for the p-value image to use for thresholding.\n Can be pval_spatial or pval_temporal\n year (int): The year to export.\n cover (ee.Image): The land cover image that is described in \n paramterIO cover dictionary.\n expected_size (int, optional): The expected image collection \n size of anomaly images for all land covers being consoli-\n dated. Defaults to 24.\n\n Returns:\n ee.Image: A yearly burn product for all land covers from the \n input anomaly collection.\n\n", "_____no_output_____" ] ], [ [ "alpha = 0.05\npVal = 'pval_spatial'\nfire2 = step2()\nout = fire2.main(alpha, pVal, year, cover)\nfire2.export_burn_yearly(\n out, region)", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

ec7ac3138e6d0c98a9a3494020460888ad86a57c

545,574

ipynb

Jupyter Notebook

Play_Store_App_Review_Analysis.ipynb

dheerajnbhat/EDA_Playstore_app_data

e2e26eb50138ab720cb3f6a3cd91fc0e4c786d3b

[ "MIT" ]

null

Play_Store_App_Review_Analysis.ipynb

dheerajnbhat/EDA_Playstore_app_data

e2e26eb50138ab720cb3f6a3cd91fc0e4c786d3b

[ "MIT" ]

null

Play_Store_App_Review_Analysis.ipynb

dheerajnbhat/EDA_Playstore_app_data

e2e26eb50138ab720cb3f6a3cd91fc0e4c786d3b

[ "MIT" ]

null

545,574

0.921521

[ [ [ "### The Play Store apps data has enormous potential to drive app-making businesses to success. Actionable insights can be drawn for developers to work on and capture the Android market.", "_____no_output_____" ] ], [ [ "# from google.colab import drive\n# drive.mount('/content/drive')", "_____no_output_____" ], [ "import pandas as pd\nimport numpy as np\nimport seaborn as sns\nimport matplotlib\nimport matplotlib.pyplot as plt\n%matplotlib inline", "_____no_output_____" ] ], [ [ "# Playstore App Data Analysis", "_____no_output_____" ] ], [ [ "play_store_data = '/content/play_store_data.csv'\nplay_store_df = pd.read_csv(play_store_data)\nuser_review_data = '/content/user_reviews.csv'\nuser_review_df = pd.read_csv(user_review_data)", "_____no_output_____" ] ], [ [ "### Check the few sample of data", "_____no_output_____" ] ], [ [ "play_store_df.head()", "_____no_output_____" ], [ "user_review_df.head()", "_____no_output_____" ] ], [ [ "### Get the unique counts and other statistics of every column in the data", "_____no_output_____" ] ], [ [ "play_store_df.describe(include='all')", "_____no_output_____" ], [ "user_review_df.describe(include='all')", "_____no_output_____" ] ], [ [ "## Data preprocessing", "_____no_output_____" ], [ "Check for missing/null values in the data", "_____no_output_____" ] ], [ [ "play_store_df.isnull().sum()", "_____no_output_____" ] ], [ [ "Columns Rating, Type, Content Rating, Current Ver and Android Ver have 1474, 1, 1, 8, 3 null values in them respectively.\n\nSince Type, Content Rating, Current Ver and Android Ver have very few null values, we can simply drop the rows which contain null values in these columns or we can replace them with the most distributed values by looking at other samples from the data\n\nLet us remove these few samples from the dataset.", "_____no_output_____" ] ], [ [ "play_store_df.dropna(subset=['Type', 'Content Rating', 'Current Ver', 'Android Ver'], inplace=True)", "_____no_output_____" ] ], [ [ "Looking at the null values again", "_____no_output_____" ] ], [ [ "play_store_df.isnull().sum()", "_____no_output_____" ] ], [ [ "Let us check the mean and median values for replacing the null values ", "_____no_output_____" ] ], [ [ "mean_value = play_store_df['Rating'].mean()\nprint('Mean value', mean_value)\nmedian_value = play_store_df['Rating'].median()\nprint('Median value', median_value)", "Mean value 4.191837606837612\nMedian value 4.3\n" ] ], [ [ "Since the mean and median both are very close to each other, we can replace missing values with the either of them.", "_____no_output_____" ] ], [ [ "play_store_df['Rating'].fillna(value=median_value, inplace=True)", "_____no_output_____" ] ], [ [ "Let us cross check if all null values are being replaced ", "_____no_output_____" ] ], [ [ "play_store_df.isnull().sum()", "_____no_output_____" ] ], [ [ "Now the data contains no missing values", "_____no_output_____" ], [ "Check the data types for all the columns", "_____no_output_____" ] ], [ [ "play_store_df.dtypes", "_____no_output_____" ] ], [ [ "Convert reviews column to int datatype", "_____no_output_____" ] ], [ [ "play_store_df['Reviews'] = play_store_df.Reviews.astype(int)", "_____no_output_____" ] ], [ [ "Size Column contains some of the special characters like , , + , M , K & also it has a some of the value as \"Varies with device\". We need to remove all of these and then convert it to int or float", "_____no_output_____" ] ], [ [ "play_store_df['Size'] = play_store_df.Size.apply(lambda x: x.strip('+')) # removing the + Sign\nplay_store_df['Size'] = play_store_df.Size.apply(lambda x: x.replace(',', '')) # removing the `,`\nplay_store_df['Size'] = play_store_df.Size.apply(lambda x: x.replace('M', 'e+6')) # converting the M to 1e+6\nplay_store_df['Size'] = play_store_df.Size.apply(lambda x: x.replace('k', 'e+3')) # converting the K to 1e+3\nplay_store_df['Size'] = play_store_df.Size.replace('Varies with device', np.NaN) # replacing varies with device with NaN\nplay_store_df['Size'] = pd.to_numeric(play_store_df['Size']) # Converting the string to Numeric type", "_____no_output_____" ] ], [ [ "Since we converted the Varies with device value to Nan , so we have to do something with those set of Nan values data. It will be a better idea to drop the Rows of the column Size having Nanvalues because it will be not an efficient idea to replace those values with mean or mode since the size of some apps would be too large and some of them too small.", "_____no_output_____" ] ], [ [ "play_store_df.dropna(subset=['Size'], inplace=True)", "_____no_output_____" ] ], [ [ "Converting Installs column from object to integer", "_____no_output_____" ] ], [ [ "play_store_df['Installs'] = play_store_df.Installs.apply(lambda x: x.strip('+'))\nplay_store_df['Installs'] = play_store_df.Installs.apply(lambda x: x.replace(',', ''))\nplay_store_df['Installs'] = pd.to_numeric(play_store_df['Installs'])", "_____no_output_____" ] ], [ [ "Converting Price column from object to integer. The values contain a special symbol $ which can be removed and then converted to the numeric type.", "_____no_output_____" ] ], [ [ "play_store_df['Price'] = play_store_df.Price.apply(lambda x: x.strip('$'))\nplay_store_df['Price'] = pd.to_numeric(play_store_df['Price'])", "_____no_output_____" ], [ "play_store_df.dtypes", "_____no_output_____" ] ], [ [ "## EDA", "_____no_output_____" ] ], [ [ "", "_____no_output_____" ] ], [ [ "### Check the correlation between data", "_____no_output_____" ] ], [ [ "f,ax = plt.subplots(figsize=(12, 12))\nsns.heatmap(play_store_df.corr(), annot=True, linewidths=.5, fmt= '.1f',ax=ax)\nplt.show()", "_____no_output_____" ] ], [ [ "As we can see, there is correlation between Reviews and Installs of 0.6", "_____no_output_____" ], [ "### Top categories in the play store which contains the highest number of apps", "_____no_output_____" ] ], [ [ "y = play_store_df['Category'].value_counts().index\nx = play_store_df['Category'].value_counts()\n\nxaxis = [x[i] for i in range(len(x))]\nyaxis = [y[i] for i in range(len(x))]\n\n\nplt.figure(figsize=(18,13))\nplt.xlabel(\"Count\")\nplt.ylabel(\"Category\")\n\ngraph = sns.barplot(x=xaxis, y=yaxis, palette=\"husl\")\ngraph.set_title(\"Top categories on Google Playstore\", fontsize=25);", "_____no_output_____" ] ], [ [ "There are all total of 33 categories in the dataset from the above output we can come to the conclusion that in the play store most of the apps are under Family & Game category and least are of Beauty & Comics Category", "_____no_output_____" ], [ "### Category of Apps from the ‘Content Rating’ column is found more on the play store", "_____no_output_____" ] ], [ [ "x = play_store_df['Content Rating'].value_counts().index\ny = play_store_df['Content Rating'].value_counts()\n\nxaxis = [x[i] for i in range(len(x))]\nyaxis = [y[i] for i in range(len(x))]\n\nplt.figure(figsize=(12,10))\nplt.bar(xaxis, yaxis, width=0.8, color=['#15244C','#FFFF48','#292734','#EF2920','#CD202D','#ECC5F2'], alpha=0.8);\nplt.title('Content Rating',size=20);\nplt.ylabel('Apps(Count)');\nplt.xlabel('Content Rating');", "_____no_output_____" ] ], [ [ "Everyone category has the highest number of apps.", "_____no_output_____" ], [ "### Distribution of the ratings of the data frame.", "_____no_output_____" ] ], [ [ "plt.figure(figsize=(15,9))\nplt.xlabel(\"Rating\")\nplt.ylabel(\"Frequency\")\ngraph = sns.kdeplot(play_store_df.Rating, color=\"Blue\", shade=True)\nplt.title('Distribution of Rating',size=20);", "_____no_output_____" ] ], [ [ "Most of the apps in the google play store are rated between 4 to 5", "_____no_output_____" ], [ "### Apps which are paid and free", "_____no_output_____" ] ], [ [ "plt.figure(figsize=(10,10))\nlabels = play_store_df['Type'].value_counts(sort=True).index\nsizes = play_store_df['Type'].value_counts(sort=True)\ncolors = [\"blue\",\"lightgreen\"]\nexplode = (0.2,0)\nplt.pie(sizes, explode=explode, labels=labels, colors=colors, autopct='%1.1f%%', shadow=True, startangle=0)\nplt.title('Percent of Free Vs Paid Apps in store',size = 20)\nplt.show()", "_____no_output_____" ] ], [ [ "Approximately 92% apps are free and 8% apps are paid ", "_____no_output_____" ], [ "### App’s having most number of installs", "_____no_output_____" ] ], [ [ "most_install_dfs = play_store_df.groupby('Category')[['Installs']].sum().sort_values(by='Installs', ascending=False)\n\nxaxis = [most_install_dfs.Installs[i] for i in range(len(most_install_dfs))]\nyaxis = [most_install_dfs.index[i] for i in range(len(most_install_dfs))]\n\nplt.figure(figsize=(18,13))\nplt.xlabel(\"Installs\")\nplt.ylabel(\"Category\")\ngraph = sns.barplot(x=xaxis, y=yaxis, alpha =0.9, palette= \"viridis\")\ngraph.set_title(\"Installs\", fontsize = 25);", "_____no_output_____" ] ], [ [ "The top categories with the highest installs are Game, Family, Communication, News & Magazines, & Tools.", "_____no_output_____" ], [ "### Top 25 installed apps in Game category", "_____no_output_____" ] ], [ [ "top = play_store_df[play_store_df['Category'] == 'GAME']\ntopapps = top.sort_values(by='Installs', ascending=False).head(25)\n# Top_Apps_in_art_and_design\nplt.figure(figsize=(15,12))\nplt.title('Top 25 Installed Apps',size = 20); \ngraph = sns.barplot(x=topapps.App, y=topapps.Installs)\ngraph.set_xticklabels(graph.get_xticklabels(), rotation= 45, horizontalalignment='right');", "_____no_output_____" ] ], [ [ "Top 10 expensive Apps in the play store", "_____no_output_____" ] ], [ [ "topPaidApps = play_store_df[play_store_df['Type'] == 'Paid'].sort_values(by='Price', ascending=False).head(11)\ntopPaidApps_df = topPaidApps[['App', 'Installs']].drop(9934)\n\nplt.figure(figsize=(15,12));\nplt.pie(topPaidApps_df.Installs, explode=None, labels=topPaidApps_df.App, autopct='%1.1f%%', startangle=0);\nplt.title('Top Expensive Apps Distribution',size = 20);\nplt.legend(topPaidApps_df.App, \n loc=\"lower right\",\n title=\"Apps\",\n fontsize = \"xx-small\"\n );", "/usr/local/lib/python3.7/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 128142 missing from current font.\n font.set_text(s, 0.0, flags=flags)\n/usr/local/lib/python3.7/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 128142 missing from current font.\n font.set_text(s, 0, flags=flags)\n" ] ], [ [ "I am rich is the most expensive app in the google play store followed by I am Rich Premium. we also had to drop one-row data for this visualization because the language of the app was Chinese and it was messing with the pie chart, visualization", "_____no_output_____" ], [ "### Count of apps in different genres", "_____no_output_____" ] ], [ [ "topAppsinGenres = play_store_df['Genres'].value_counts().head(50)\n\nxaxis = [topAppsinGenres.index[i] for i in range(len(topAppsinGenres))]\nyaxis = [topAppsinGenres[i] for i in range(len(topAppsinGenres))]\n\nplt.figure(figsize=(15,9))\nplt.ylabel('Genres(App Count)')\nplt.xlabel('Genres')\ngraph = sns.barplot(x=xaxis, y=yaxis, palette=\"deep\")\ngraph.set_xticklabels(graph.get_xticklabels(), rotation=90, fontsize=12)\ngraph.set_title(\"Top Genres in the Playstore\", fontsize = 20);", "_____no_output_____" ] ], [ [ "Highest Number of Apps are found in the Tools and Entertainment genres followed by Education, Medical and many more.", "_____no_output_____" ], [ "Apps that have made the highest-earning", "_____no_output_____" ] ], [ [ "Paid_Apps_df = play_store_df[play_store_df['Type'] == 'Paid']\nearning_df = Paid_Apps_df[['App', 'Installs', 'Price']]\nearning_df['Earnings'] = earning_df['Installs'] * earning_df['Price'];\nearning_df_sorted_by_Earnings = earning_df.sort_values(by='Earnings', ascending=False).head(50)\nearning_df_sorted_by_Price = earning_df_sorted_by_Earnings.sort_values(by='Price', ascending=False)\n\nplt.figure(figsize=(15,9))\nplt.bar(earning_df_sorted_by_Price.App, earning_df_sorted_by_Price.Earnings, width=1.1, label=earning_df_sorted_by_Price.Earnings)\nplt.xlabel(\"Apps\")\nplt.ylabel(\"Earnings\")\nplt.tick_params(rotation=90)\nplt.title(\"Top Earning Apps\");", "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:3: SettingWithCopyWarning: \nA value is trying to be set on a copy of a slice from a DataFrame.\nTry using .loc[row_indexer,col_indexer] = value instead\n\nSee the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n This is separate from the ipykernel package so we can avoid doing imports until\n/usr/local/lib/python3.7/dist-packages/matplotlib/backends/backend_agg.py:214: RuntimeWarning: Glyph 128142 missing from current font.\n font.set_text(s, 0.0, flags=flags)\n/usr/local/lib/python3.7/dist-packages/matplotlib/backends/backend_agg.py:183: RuntimeWarning: Glyph 128142 missing from current font.\n font.set_text(s, 0, flags=flags)\n" ] ], [ [ "The top five apps with the highest earnings found on google play store are: I am Rich, I am Rich Premium, Hitman Sniper, Grand Theft Auto: San Andreas, Facetune - For Free", "_____no_output_____" ] ], [ [ "# Join and merge the two dataframe\nmerged_df = pd.merge(play_store_df, user_review_df, on='App', how = \"inner\")\n\n# Drop NA values from Sentiment and Translated_Review columns\nmerged_df = merged_df.dropna(subset=['Sentiment', 'Translated_Review'])\n\nsns.set_style('ticks')\nfig, ax = plt.subplots()\nfig.set_size_inches(11, 8)\n\n# User review sentiment polarity for paid vs. free apps\nax = sns.boxplot(x = 'Type', y = 'Sentiment_Polarity', data = merged_df)\nax.set_title('Sentiment Polarity Distribution')", "_____no_output_____" ] ], [ [ "Free apps receive a lot of negative comments, as indicated by the outliers on the negative y-axis. Reviews for paid apps appear never to be extremely negative. This helps indicate about the app quality, i.e. paid apps being of higher quality than free apps on average.", "_____no_output_____" ] ], [ [ "user_review_df1 = user_review_df.dropna(inplace=False)\n\n# Create the Likert Scale\nlikert = {\n \"Negative\": -1,\n \"Neutral\": 0,\n \"Positive\": 1\n}\n\n# Transform the Sentument column to match the Likert Scale value\nuser_review_df1.Sentiment = user_review_df1.Sentiment.apply(lambda x: likert[x]).copy()\n\n# Here, we obtain the mean for each app by grouping the data\nreviews_mean = user_review_df1.groupby(\"App\").mean().copy()", "/usr/local/lib/python3.7/dist-packages/pandas/core/generic.py:5516: SettingWithCopyWarning: \nA value is trying to be set on a copy of a slice from a DataFrame.\nTry using .loc[row_indexer,col_indexer] = value instead\n\nSee the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n self[name] = value\n" ], [ "complete_data = pd.merge(left=play_store_df, right=reviews_mean, on=\"App\").copy()\n\n# Drop duplicates\ncomplete_data.drop_duplicates(\"App\", inplace=True)\n\n# Reset the index since now we have a different number of observations\ncomplete_data = complete_data.reset_index().drop(\"index\", axis=1).copy()\n\n# Select columns that will be used\ncolumns = [0, 1, 2, 3, 5, 6, 8, 9, 13, 14, 15]\ncomplete_data = complete_data.iloc[:,columns].copy()\n", "_____no_output_____" ] ], [ [ "### Analyzing how sentiment influences the rating of the app", "_____no_output_____" ] ], [ [ "# Create the plot for the histograms\nfig, ax = plt.subplots(nrows=1, ncols=2, figsize=(12, 6), sharey=True)\n\n# Create the histograms\nax[0].hist(complete_data.Sentiment, bins=20)\nax[1].hist(complete_data.Rating, bins=20)\nax[1].set_xlim(0, 5)\n\n# Add titles\nax[0].set_title(\"Sentiment\")\nax[1].set_title(\"Rating\")\n\nplt.show()", "_____no_output_____" ] ], [ [ "In the Sentiment aspect, the majority of the values are above neutral (0), at around 0,5 for a range of -1 to 1. But for the Rating column, this trend looks way more significant, with a lot of the values being around 4 and 5 for a range of 0 to 5", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ] ]

ec7ac6b0ef8d83763f7be1a648815faec4eae917

702

ipynb

Jupyter Notebook

HelloGitHub.ipynb

vonTita/dw_matrix

e8e4708bfa4c14a6a60b0b3f31a1eeac0971fe6c

[ "MIT" ]

null

HelloGitHub.ipynb

vonTita/dw_matrix

e8e4708bfa4c14a6a60b0b3f31a1eeac0971fe6c

[ "MIT" ]

null

HelloGitHub.ipynb

vonTita/dw_matrix

e8e4708bfa4c14a6a60b0b3f31a1eeac0971fe6c

[ "MIT" ]

null

702

0.695157

[ [ [ "print(\"Hello GitHub\")", "Hello GitHub\n" ] ] ]

[ "code" ]

[ [ "code" ] ]

ec7ac86961f474b68b02c98c80a547de7ab0bb95

8,250

ipynb

Jupyter Notebook

Resources/csv to html.ipynb

bjouellette/Web-Design-Challenge

7a05244c1d4b1b73bfe749cd2cce2ba3919edfac

[ "ADSL" ]

1

2022-01-27T00:04:24.000Z

Resources/csv to html.ipynb

bjouellette/Web_Design_Challenge

7a05244c1d4b1b73bfe749cd2cce2ba3919edfac

[ "ADSL" ]

null

Resources/csv to html.ipynb

bjouellette/Web_Design_Challenge

7a05244c1d4b1b73bfe749cd2cce2ba3919edfac

[ "ADSL" ]

null

30.555556

91

0.308242

[ [ [ "import pandas as pd\ndf = pd.read_csv(\"cities.csv\")\ndf", "_____no_output_____" ], [ "df.to_html('cities.html')", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code" ] ]

ec7ad01e87aa2b02738b2529bcff1ea0b4bec9d2

21,570

ipynb

Jupyter Notebook

week5.ipynb

jitendra80830/Python-Programming

2aed091559503d61d78c19854d23c5f94f2464ef

[ "Apache-2.0" ]

1

2020-10-31T12:43:48.000Z

week5.ipynb

jitendra80830/Python-Programming

2aed091559503d61d78c19854d23c5f94f2464ef

[ "Apache-2.0" ]

null

week5.ipynb

jitendra80830/Python-Programming

2aed091559503d61d78c19854d23c5f94f2464ef

[ "Apache-2.0" ]

null

23.860619

704

0.493

[ [ [ "#using a while loop with list and dictionary\n\n#Moving items from one list to another:-\nunconfermed_users=['ram','shayam','krishna']\nconfermed_users=[]\nwhile unconfermed_users:\n current_user=unconfermed_users.pop()\n \n print(f\"Verifying user:\" + current_user.title())\n confermed_users.append(current_user)\n \nprint(\"\\nFollowing users have been confermed.\")\nfor confermed_user in confermed_users:\n print(confermed_user.title())", "Verifying user:Krishna\nVerifying user:Shayam\nVerifying user:Ram\n\nFollowing users have been confermed.\nKrishna\nShayam\nRam\n" ], [ "# Removing all instance of specific value fropm list:-\npets=['dog','cat','dog','golgfish','cat','rabit','cat']\nprint(pets)\nwhile 'cat' in pets:\n pets.remove('cat')\nprint(pets)", "['dog', 'cat', 'dog', 'golgfish', 'cat', 'rabit', 'cat']\n['dog', 'dog', 'golgfish', 'rabit']\n" ], [ "#filling a dict wuth user input:-\nresponses={}\npolling_active=True\nwhile polling_active:\n name=input(\"\\nWhat is your name?\")\n response=input('Which mountain would you like to climb someday?')\n responses[name]=response\n repeat=input(\"would you like to let another person response?(yes/no)\")\n if repeat=='no':\n polling_active=False\nprint(\"\\n.....poll Result....\")\nfor name,response in responses.items():\n print(f\"{name} would like to climb {response}\") \n \n ", "\nWhat is your name?ram\nWhich mountain would you like to climb someday?kuleshwari\nwould you like to let another person response?(yes/no)yes\n\nWhat is your name?jitendra\nWhich mountain would you like to climb someday?bageshwari\nwould you like to let another person response?(yes/no)no\n\n.....poll Result....\nram would like to climb kuleshwari\njitendra would like to climb bageshwari\n" ], [ "# table without loop:-\nn=int(input(\"Enter a number\"))\nprint(\"Table of\",n)\nprint(n,'x',1,'=',n*1)\nprint(n,'x',1,'=',n*2)\nprint(n,'x',1,'=',n*3)\nprint(n,'x',1,'=',n*4)\nprint(n,'x',1,'=',n*5)\nprint(n,'x',1,'=',n*6)\nprint(n,'x',1,'=',n*7)\nprint(n,'x',1,'=',n*8)\nprint(n,'x',1,'=',n*9)\nprint(n,'x',1,'=',n*10)", "Enter a number15\nTable of 15\n15 x 1 = 15\n15 x 1 = 30\n15 x 1 = 45\n15 x 1 = 60\n15 x 1 = 75\n15 x 1 = 90\n15 x 1 = 105\n15 x 1 = 120\n15 x 1 = 135\n15 x 1 = 150\n" ], [ "# table with while loop:-\nn=int(input(\"enter a number \"))\nprint(\"Table of\",n)\ni=1\nwhile (i<=10):\n print(n,'x',i,'=',n*i)\n i+=1\nprint(\"done\")", "enter a number 5\nTable of 5\n5 x 1 = 5\n5 x 2 = 10\n5 x 3 = 15\n5 x 4 = 20\n5 x 5 = 25\n5 x 6 = 30\n5 x 7 = 35\n5 x 8 = 40\n5 x 9 = 45\n5 x 10 = 50\ndone\n" ], [ "# Traking the loop\nn=int(input())\nwhile (n!=-1): #read reamining number inside the loop.\n n=int(input())", "22\n3\n4\n56\n7\n-1\n" ], [ "# fing largest no until -1\nn=int(input())\nmax=n\nwhile (n!=-1):\n n=int(input())\n if max<n:\n max=n", "6\n7\n-1\n" ], [ "#program\nn=int(input())\nx=1\nwhile (x<=n):\n if(x%3==0) or (x%5==0):\n print(x)\n x+=1", "7\n" ], [ "# L-14\n# BREAK AND CONTINUE:\n#BREAK:-\nsum=0\nfor i in range(100):\n value=int(input())\n if (value<0):\n break\n sum=sum+value\nprint(\"sum=\",sum)", "3\n5\n6\n8\n-1\nsum= 22\n" ], [ "#CONTUNUE:-\nsum=0\nfor i in range(5):\n value=int(input())\n if (value<0):\n continue\n sum=sum+value\nprint(\"sum=\",sum)", "5\n3\n-2\n6\n-1\nsum= 14\n" ], [ "#print all odd numbers\ni=1\nwhile i<=10:\n if i%2==0:\n continue\n print(i,end=' ')\n i=i+1\n", "1 " ], [ "#FUNTION:-\ndef greet_user():\n print(\"hello\")\ngreet_user()", "hello\n" ], [ "#example\ndef greet_user(username):\n print(f\"hello,{username.title()}\")\ngreet_user('jitendra kumar')", "hello,Jitendra Kumar\n" ], [ "# Example of multiple parameter and arguments:-\ndef describe_pet(animal_type,pet_name):\n print(f\"\\ni have a {animal_type}\")\n print(f\"My {animal_type}'s name is {pet_name.title()}\")\ndescribe_pet('cat','victory')\ndescribe_pet('dog','willie')", "\ni have a cat\nMy cat's name is Victory\n\ni have a dog\nMy dog's name is Willie\n" ], [ "# POSITION Example\ndef describe_pet(animal_type,pet_name):\n print(f\"\\ni have a {animal_type}\")\n print(f\"My {animal_type}'s name is {pet_name.title()}\")\ndescribe_pet('victory','cat')", "\ni have a victory\nMy victory's name is Cat\n" ], [ "#Keyword Argument:-\ndef describe_pet(animal_type,pet_name):\n print(f\"\\ni have a {animal_type}\")\n print(f\"My {animal_type}'s name is {pet_name.title()}\")\ndescribe_pet(animal_type='cat',pet_name='victory')\ndescribe_pet(pet_name='victory',animal_type='cat')", "\ni have a cat\nMy cat's name is Victory\n\ni have a cat\nMy cat's name is Victory\n" ], [ "# Default value\ndef describe_pet(pet_name,animal_type='dog'):\n print(f\"\\ni have a {animal_type}\")\n print(f\"My {animal_type}'s name is {pet_name.title()}\")\ndescribe_pet(pet_name='willie')", "\ni have a dog\nMy dog's name is Willie\n" ], [ "# L=15\n#Avoiding argument error\ndef describe_pet(animal_type,pet_name):\n print(f\"\\ni have a {animal_type}.\")\n print(f\"My {aniaml_type}'s name is {pet_name.title()}\")\ndescribe_pet()", "_____no_output_____" ], [ "# Return value with funtion\ndef get_formatted_name(first_name,last_name):\n full_name=(f\"{first_name} {last_name}\")\n return full_name.title()\nmusician=get_formatted_name('ravi','shankar')\nprint(musician)", "Ravi Shankar\n" ], [ "# Example\ndef get_formatted_name(first_name,middle_name,last_name):\n full_name=(f\"{first_name} {middle_name} {last_name}\")\n return full_name.title()\n\nmusician=get_formatted_name('a','r','rehman')\nprint(musician)", "A R Rehman\n" ], [ "# Example with if condition\ndef get_formatted_name(first_name,last_name,middle_name=''):\n if middle_name:\n full_name=(f\"{first_name} {middle_name} {last_name}\")\n else:\n full_name=(f\"{first_name} {last_name}\")\n return full_name.title()\nmusician=get_formatted_name('ravi','shankar')\n\nmusician=get_formatted_name('a','r','rehman')\nprint(musician)", "A Rehman R\n" ], [ "# example with dictionary\ndef build_person(first_name,last_name):\n person={'first':first_name,'last': last_name}\n return person\nmusician=build_person('ravi','shankar')\nprint(musician)", "{'first': 'ravi', 'last': 'shankar'}\n" ], [ "#example with person'age\ndef build_person(first_name,last_name,age='non'):\n person={'first':first_name,'last': last_name}\n if age:\n person['age']=age\n return person\nmusician=build_person('ravi','shankar',age=92)\nprint(musician)\n ", "{'first': 'ravi', 'last': 'shankar', 'age': 92}\n" ], [ " # Example fo quit at any time\ndef get_foematted_name(first_name,last_name):\n full_naame=(f\"{first_name} {last_name}\")\n return full_name.title()\nwhile True:\n print(\"\\nPlease tell me your name\")\n print(\"(enter 'q' at any time to quit)\")\n f_name=input('first_name')\n if f_name=='q':\n break\n l_name=input('last_name')\n if l_name=='q':\n break\nformatted_name=get_formatted_name(f_name,l_name)\nprint(f\"\\nHello, {formatted_name}\")\n ", "\nPlease tell me your name\n(enter 'q' at any time to quit)\nfirst_namejitendra\nlast_nameraj\n\nPlease tell me your name\n(enter 'q' at any time to quit)\nfirst_nameq\n\nHello, Q Raj\n" ], [ "# PASSING A LIST WITH FUNTIION:-\ndef greet_users(names):\n for name in names:\n msg=(f\"Hello,{name.title()}\")\n print(msg)\nusernames=['dhruv','ishita','prachi']\ngreet_users(usernames)", "Hello,Dhruv\nHello,Ishita\nHello,Prachi\n" ], [ "#modifying a list\nunprinted_designs=['phhone case','robot pedent','cube']\ncompleted_models=[]\n\nwhile unprinted_designs:\n current_design=unprinted_designs.pop()\n print(f\"printing model: {current_design}\")\n completed_models.append(current_design)\nprint(\"\\nthe following models have been printed\")\nfor completed_model in completed_models:\n print(completed_model)", "printing model: cube\nprinting model: robot pedent\nprinting model: phhone case\n\nthe following models have been printed\ncube\nrobot pedent\nphhone case\n" ], [ "#Example with funtion\ndef print_models(unprinted_designs,completed_models):\n while unprinted_designs:\n current_design=unprinted_designs.pop()\n print(f\"printing models:{current_design}\")\n completed_models.append(current_design)\n \ndef show_completed_models(completed_models):\n print(\"\\nThe following mogels have been printed\")\n for completed_model in completed_models:\n print(completed_model)\n \nunprinted_designs=['phone case','robot pedant','code']\ncompleted_models=[]\nprint_models(unprinted_designs,completed_models)\nshow_completed_models(completed_models)", "printing models:code\nprinting models:robot pedant\nprinting models:phone case\n\nThe following mogels have been printed\ncode\nrobot pedant\nphone case\n" ], [ "#w-6\n#l-16\n# passing an arbitrary number of arguments\n#Example-\ndef make_pizza(*toppings):\n print(toppings)\nmake_pizza('veggie')\nmake_pizza('onion','tomato','cheese')\n ", "('veggie',)\n('onion', 'tomato', 'cheese')\n" ], [ "def make_pizza(*toppings):\n print(\"\\nmaking the pizza with the following toppings\")\n for topping in toppings:\n print(f\"-{topping}\")\nmake_pizza('veggie')\nmake_pizza('onion','tomato','cheese')\n ", "\nmaking the pizza with the following toppings\n-veggie\n\nmaking the pizza with the following toppings\n-onion\n-tomato\n-cheese\n" ], [ "#Mixing positional and arbitrary arguments\ndef make_pizza(size,*toppings):\n print(f\"\\nmaking a {size}-inch the pizza with the following toppings\")\n for topping in toppings:\n print(f\"-{topping}\")\nmake_pizza(16,'veggie')\nmake_pizza(12,'onion','tomato','cheese')\n ", "\nmaking a 16-inch the pizza with the following toppings\n-veggie\n\nmaking a 12-inch the pizza with the following toppings\n-onion\n-tomato\n-cheese\n" ], [ "# using arbitrary keyword arguments:\ndef build_profile(first,last,**user_info):\n user_info['first_name']=first\n user_info['last_name']=last\n return user_info\nuser_profile=build_profile('rahul','garg',location='iitkanpur',field='ict')\nprint(user_profile)", "{'location': 'iitkanpur', 'field': 'ict', 'first_name': 'rahul', 'last_name': 'garg'}\n" ], [ "# Storing your fun in moduls:\n", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7aeb80d8c4ce69a4873115c86456f087aa6165

27,342

ipynb

Jupyter Notebook

MIPVerify.jl/examples/01_importing_your_own_neural_net.ipynb

anonymous2398384/provable_robustness_max_linear_regions

529165d9047261813bc068997415f668c9675119

[ "BSD-3-Clause" ]

34

2019-03-10T22:16:24.000Z

2021-09-23T22:22:27.000Z

MIPVerify.jl/examples/01_importing_your_own_neural_net.ipynb

anonymous2398384/provable_robustness_max_linear_regions

529165d9047261813bc068997415f668c9675119

[ "BSD-3-Clause" ]

2

2019-09-24T16:18:55.000Z

2021-03-06T20:57:33.000Z

MIPVerify.jl/examples/01_importing_your_own_neural_net.ipynb

anonymous2398384/provable_robustness_max_linear_regions

529165d9047261813bc068997415f668c9675119

[ "BSD-3-Clause" ]

9

2019-03-13T17:35:36.000Z

2021-01-15T02:37:23.000Z

36.799462

342

0.50139

[ [ [ "empty" ] ] ]

[ "empty" ]

[ [ "empty" ] ]

ec7af1f93a7761300a9421183ee1509bbf898dcf

14,092

ipynb

Jupyter Notebook

T-Test, Correlation, Chi-Square Test Implemenation/.ipynb_checkpoints/T-Test Implementation-checkpoint.ipynb

yashrajjain726/Machine-Learning-Hands-On-Practice

32075012c95d06386902da7ffa5a905e2ef22a40

[ "Unlicense" ]

6

2020-07-04T07:00:27.000Z

2021-08-17T02:49:45.000Z

T-Test, Correlation, Chi-Square Test Implemenation/T-Test Implementation.ipynb

yashrajjain726/Machine-Learning-Hands-On-Practice

32075012c95d06386902da7ffa5a905e2ef22a40

[ "Unlicense" ]

null

T-Test, Correlation, Chi-Square Test Implemenation/T-Test Implementation.ipynb

yashrajjain726/Machine-Learning-Hands-On-Practice

32075012c95d06386902da7ffa5a905e2ef22a40

[ "Unlicense" ]

null

22.5472

322

0.433934

[ [ [ "## T-Test Implementation with 1- Sample T-test", "_____no_output_____" ], [ "#### 1-Sample T-Test will tell us whether mean's of the sample and population are different or not.", "_____no_output_____" ] ], [ [ "import numpy as np", "_____no_output_____" ], [ "ages = [10,20,30,40,35,94,35,51,3,66,1,21,5,84,3,54,21,24,71,15,61,45,87,12,64,75,15,2,48,12,23,14] ", "_____no_output_____" ], [ "len(ages)", "_____no_output_____" ], [ "ages_mean = np.mean(ages)", "_____no_output_____" ], [ "ages_mean", "_____no_output_____" ], [ "sample_mean_size = 10", "_____no_output_____" ], [ "sample_ages = np.random.choice(ages,sample_mean_size)", "_____no_output_____" ], [ "from scipy.stats import ttest_1samp", "_____no_output_____" ], [ "t_test, p_value = ttest_1samp(sample_ages,35)", "_____no_output_____" ], [ "p_value", "_____no_output_____" ], [ "if p_value < 0.05:\n print(\"We are Rejecting Null Hypothesis.\")\nelse:\n print(\"We are Accepting Null Hypothesis.\")", "We are Accepting Null Hypothesis.\n" ] ], [ [ "### More Problems on 1 Sample T-Test Implementation", "_____no_output_____" ] ], [ [ "import pandas as pd\nimport scipy.stats as stats\nimport numpy as np\nimport matplotlib.pyplot as plt\nimport math\nnp.random.seed(0)", "_____no_output_____" ], [ "school_age = stats.poisson.rvs(18,35,1500)\nclass_a_age = stats.poisson.rvs(18,30,60)", "_____no_output_____" ], [ "school_age_mean = school_age.mean()", "_____no_output_____" ], [ "school_age_mean", "_____no_output_____" ], [ "_, p_value = ttest_1samp(class_a_age,school_age_mean)", "_____no_output_____" ], [ "p_value", "_____no_output_____" ], [ "if p_value <= 0.05:\n print(\"We are Rejecting Null Hypothesis.\")\nelse:\n print(\"We are Accepting Null Hypothesis.\")", "We are Rejecting Null Hypothesis.\n" ] ], [ [ "## T-Test Implementation with 2-Sample T-Test", "_____no_output_____" ], [ "#### 2-Sample T-Test compares the mean of two independent group's in order to determine whether there is statistical evidence that the associated population means are significantly different. The independent sample T-Test is a parametric test. <span style=\"color:red\">It is also known as Independent T-Test</span>.", "_____no_output_____" ] ], [ [ "np.random.seed(0)\nclass_b_age = stats.poisson.rvs(18,33,60)", "_____no_output_____" ], [ "ttest, p_value = stats.ttest_ind(class_a_age,class_b_age,equal_var = False)", "_____no_output_____" ], [ "p_value", "_____no_output_____" ], [ "if p_value <= 0.05:\n print(\"We are Rejecting Null Hypothesis.\")\nelse:\n print(\"We are Accepting Null Hypothesis.\")", "We are Rejecting Null Hypothesis.\n" ] ], [ [ "## Paired T-Test with Python", "_____no_output_____" ], [ "#### Paired T-Test is used when you want to check how different sample's from the same group are.", "_____no_output_____" ] ], [ [ "weight_now = [25,23,26,34,20,15,35,16,23,30,15,25,32,45,40]\nweight_after_a_month = weight_now + stats.norm.rvs(scale = 5,loc = -1.25,size = 15)", "_____no_output_____" ], [ "weight_after_a_month", "_____no_output_____" ], [ "weight_dataframe = pd.DataFrame({\"Weight Now\":np.array(weight_now),\n \"Weight After a Month\":np.array(weight_after_a_month),\n \"Weight Change\":np.array(weight_after_a_month) - np.array(weight_now)})", "_____no_output_____" ], [ "weight_dataframe", "_____no_output_____" ], [ "_,p_value = stats.ttest_rel(weight_now,weight_after_a_month)", "_____no_output_____" ], [ "p_value", "_____no_output_____" ], [ "if p_value < 0.05:\n print(\"We are Rejecting Null Hypothesis.\")\nelse:\n print(\"We are Accepting Null Hypothesis.\")", "We are Accepting Null Hypothesis.\n" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code" ], [ "markdown", "markdown" ], [ "code", "code", "code", "code" ], [ "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code", "code" ] ]

ec7af5471c8fd3a5c44dcd06ccb31732c3415939

284,344

ipynb

Jupyter Notebook

forecast.ipynb

AnnaBorisova/python-test

61a66a6b457009e52ecc22cd93de5d926c747f2f

[ "MIT" ]

null

forecast.ipynb

AnnaBorisova/python-test

61a66a6b457009e52ecc22cd93de5d926c747f2f

[ "MIT" ]

null

forecast.ipynb

AnnaBorisova/python-test

61a66a6b457009e52ecc22cd93de5d926c747f2f

[ "MIT" ]

null

517.930783

66,748

0.942457

[ [ [ "import pandas as pd\nimport numpy as np\nimport matplotlib\nimport datetime\nimport pylab as plt\nfrom statsmodels.graphics.tsaplots import plot_acf\n\nimport functions as func\n\nfrom sklearn.linear_model import LinearRegression\nimport matplotlib.pyplot as plt\nfrom patsy import dmatrices\nimport statsmodels.formula.api as smf\n\nimport warnings\nwarnings.filterwarnings('ignore')", "_____no_output_____" ] ], [ [ "# I. Exploratory Data Analysis", "_____no_output_____" ] ], [ [ "df = pd.read_csv('energy.dat')\ndf['Date'] = pd.to_datetime(df['Date'], format = '%d-%b-%y')\ndf = df.set_index('Date')", "_____no_output_____" ], [ "df.plot()", "_____no_output_____" ] ], [ [ "The chart of the energy consumption shows pronounced annual seasonality.", "_____no_output_____" ] ], [ [ "df['weekend'] = df.apply(lambda x: 1 if x.name.dayofweek > 4 else 0, axis = 1)\ndf['week'] = df.index.week", "_____no_output_____" ], [ "ax = df[df.weekend==0].plot.scatter(x='week',\n y='Consumption',\n c='DarkBlue', label = 'working_day')\ndf[df.weekend==1].plot.scatter(x='week',\n y='Consumption',\n c='Green', label = 'weekend', ax = ax)", "_____no_output_____" ] ], [ [ "The scatter plots of demand by week number shows that demand is higher during working days. However this relationship is less pronounced in winter as other factors, such as the number of HDD, play a more important role.", "_____no_output_____" ] ], [ [ "plot_acf(df.Consumption, lags = 30)", "_____no_output_____" ], [ "plot_acf(df.Consumption, lags = 365)", "_____no_output_____" ] ], [ [ "Energy demand data show strong autocorrelation.", "_____no_output_____" ], [ "# II. Select explanatory variables and train the model", "_____no_output_____" ] ], [ [ "#Add trend variable and seasonality variables\ndf['trend'] = df.apply(lambda x: (x.name - datetime.datetime(2015,4,1)).days, axis = 1)\n\ndf = func.seasonal_var (df, 365, 3)\n#df = func.seasonal_var (df, 7, 3)", "_____no_output_____" ], [ "#Define specification of the model\nmod = 'Consumption~ '\nfor i in df.columns[3:]:\n mod = mod + i +' + '\nmod = mod[:-3]\nprint(mod)", "Consumption~ trend + seas_365_1a + seas_365_1b + seas_365_2a + seas_365_2b + seas_365_3a + seas_365_3b\n" ], [ "#Train the model\nfitted_reg = pd.DataFrame()\nresiduals = pd.DataFrame()\n\nlr = LinearRegression()\ny,X = dmatrices (mod, df, return_type = 'dataframe')\nlr.fit(X,y)\nmodel = smf.ols(formula = mod, data = df).fit()\n\nprint(model.summary())\n\nnp.savetxt('params/'+mod.split('~')[0]+'_coef.csv', lr.coef_) #, delimeter = ','\nnp.savetxt('params/'+mod.split('~')[0]+'_intercept.csv', lr.intercept_)\nnp.savetxt('params/'+mod.split('~')[0]+'_model.csv', [mod], fmt = '%s')\n\nfitted_reg[mod.split('~')[0]] = model.fittedvalues\nresiduals[mod.split('~')[0]] = model.resid", " OLS Regression Results \n==============================================================================\nDep. Variable: Consumption R-squared: 0.714\nModel: OLS Adj. R-squared: 0.714\nMethod: Least Squares F-statistic: 1140.\nDate: Sun, 14 Nov 2021 Prob (F-statistic): 0.00\nTime: 19:33:31 Log-Likelihood: -9878.7\nNo. Observations: 1827 AIC: 1.977e+04\nDf Residuals: 1822 BIC: 1.979e+04\nDf Model: 4 \nCovariance Type: nonrobust \n===============================================================================\n coef std err t P>|t| [0.025 0.975]\n-------------------------------------------------------------------------------\nIntercept 296.2598 2.559 115.770 0.000 291.241 301.279\ntrend -0.0420 0.002 -17.246 0.000 -0.047 -0.037\nseas_365_1a -60.6192 0.905 -66.953 0.000 -62.395 -58.843\nseas_365_1b -60.6192 0.905 -66.953 0.000 -62.395 -58.843\nseas_365_2a -3.6573 0.897 -4.077 0.000 -5.417 -1.898\nseas_365_2b -3.6573 0.897 -4.077 0.000 -5.417 -1.898\nseas_365_3a -4.9441 0.896 -5.521 0.000 -6.700 -3.188\nseas_365_3b -4.9441 0.896 -5.521 0.000 -6.700 -3.188\n==============================================================================\nOmnibus: 121.068 Durbin-Watson: 0.268\nProb(Omnibus): 0.000 Jarque-Bera (JB): 183.859\nSkew: 0.534 Prob(JB): 1.19e-40\nKurtosis: 4.128 Cond. No. 4.63e+19\n==============================================================================\n\nWarnings:\n[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n[2] The smallest eigenvalue is 9.49e-31. This might indicate that there are\nstrong multicollinearity problems or that the design matrix is singular.\n" ], [ "plt.figure()\nplt.subplot(211)\nplt.plot(df[mod.split('~')[0]].loc[df[mod.split('~')[0]].first_valid_index():df[mod.split('~')[0]].last_valid_index()])\nplt.plot(model.fittedvalues.loc[model.fittedvalues.first_valid_index():model.fittedvalues.last_valid_index()])\nplt.legend(['Data', 'Fitted Model'])\nplt.title(mod.split('~')[0])\nplt.subplot(212)\nplt.plot(model.resid.loc[model.resid.first_valid_index():model.resid.last_valid_index()], 'ro')\nplt.legend(['Residuals'])\nplt.show()", "_____no_output_____" ] ], [ [ "# III. Produce the forecast", "_____no_output_____" ] ], [ [ "output = pd.DataFrame(index = pd.date_range(start = '2020-04-01', end = '2022-03-31'), columns = ['trend'])\noutput['trend'] = output.apply(lambda x: (x.name - datetime.datetime(2015,4,1)).days, axis = 1)\noutput = func.seasonal_var (output, 365, 3)\noutput['Consumption'] = 0", "_____no_output_____" ], [ "mod.split('~')[0]", "_____no_output_____" ], [ "demand_forecast = pd.DataFrame()\ny,X = dmatrices(mod, output, return_type = 'dataframe')\nlr = LinearRegression()\ncoefs = np.genfromtxt('params/'+mod.split('~')[0]+'_coef.csv')\nintercepts = np.genfromtxt('params/'+mod.split('~')[0]+'_intercept.csv')\nlr.intercept_ = intercepts\nlr.coef_ = coefs\nf = lr.predict(X)\ndemand_forecast = pd.DataFrame(f, index = X.index, columns=['Forecast'])", "_____no_output_____" ], [ "demand_forecast.plot()", "_____no_output_____" ], [ "fin = pd.DataFrame(index = pd.date_range(start = '2015-04-01', end = '2022-03-31'), columns = ['Actual', 'Forecast'])\nfin.Actual = fin.apply(lambda x: df[df.index==x.name]['Consumption'].iloc[0] if not(df[df.index==x.name]['Consumption'].empty) else np.nan, axis = 1)\nfin.Forecast = fin.apply(lambda x: demand_forecast[demand_forecast.index==x.name]['Forecast'].iloc[0] if not(demand_forecast[demand_forecast.index==x.name]['Forecast'].empty) else np.nan, axis = 1)", "_____no_output_____" ], [ "#Plot actual consumption and produced forecast\nfin.plot(figsize = (15,5))", "_____no_output_____" ], [ "fin.to_csv('output/output.csv')", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code" ] ]

ec7afd4edd95bdea0131f4d67f9a3151fca49df1

510,850

ipynb

Jupyter Notebook

CTC_IGFL_FR.ipynb

ksugar/CTC-IGFL-FR

b73d94311319ccd1c30f5f833c64095d53b537de

[ "BSD-2-Clause" ]

null

CTC_IGFL_FR.ipynb

ksugar/CTC-IGFL-FR

b73d94311319ccd1c30f5f833c64095d53b537de

[ "BSD-2-Clause" ]

null

CTC_IGFL_FR.ipynb

ksugar/CTC-IGFL-FR

b73d94311319ccd1c30f5f833c64095d53b537de

[ "BSD-2-Clause" ]

null

86.746476

2,216

0.529574

[ [ [ "# Download CTC-IGFL-FR", "_____no_output_____" ] ], [ [ "from pathlib import Path\n\n%cd /content\nif Path('CTC-IGFL-FR-main').exists():\n print('CTC-IGFL-FR-main directory alredy exists')\nelse:\n !wget https://github.com/ksugar/CTC-IGFL-FR/archive/refs/heads/main.zip\n !unzip -q main.zip && rm main.zip\n%cd CTC-IGFL-FR-main/src/SW", "/content\n--2022-01-10 11:39:39-- https://github.com/ksugar/CTC-IGFL-FR/archive/refs/heads/main.zip\nResolving github.com (github.com)... 52.69.186.44\nConnecting to github.com (github.com)|52.69.186.44|:443... connected.\nHTTP request sent, awaiting response... 302 Found\nLocation: https://codeload.github.com/ksugar/CTC-IGFL-FR/zip/refs/heads/main [following]\n--2022-01-10 11:39:39-- https://codeload.github.com/ksugar/CTC-IGFL-FR/zip/refs/heads/main\nResolving codeload.github.com (codeload.github.com)... 52.193.111.178\nConnecting to codeload.github.com (codeload.github.com)|52.193.111.178|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: unspecified [application/zip]\nSaving to: ‘main.zip’\n\nmain.zip [ <=> ] 347.52K 1.85MB/s in 0.2s \n\n2022-01-10 11:39:40 (1.85 MB/s) - ‘main.zip’ saved [355864]\n\n/content/CTC-IGFL-FR-main/src/SW\n" ] ], [ [ "# Download data", "_____no_output_____" ] ], [ [ "!./download_data.sh BF-C2DL-HSC", "--2022-01-10 11:39:40-- http://data.celltrackingchallenge.net/training-datasets/BF-C2DL-HSC.zip\nResolving data.celltrackingchallenge.net (data.celltrackingchallenge.net)... 147.251.52.183\nConnecting to data.celltrackingchallenge.net (data.celltrackingchallenge.net)|147.251.52.183|:80... connected.\nHTTP request sent, awaiting response... 301 Moved Permanently\nLocation: https://data.celltrackingchallenge.net/training-datasets/BF-C2DL-HSC.zip [following]\n--2022-01-10 11:39:41-- https://data.celltrackingchallenge.net/training-datasets/BF-C2DL-HSC.zip\nConnecting to data.celltrackingchallenge.net (data.celltrackingchallenge.net)|147.251.52.183|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 1707088358 (1.6G) [application/zip]\nSaving to: ‘../../Data/BF-C2DL-HSC.zip’\n\n../../Data/BF-C2DL- 100%[===================>] 1.59G 10.9MB/s in 2m 34s \n\n2022-01-10 11:42:16 (10.6 MB/s) - ‘../../Data/BF-C2DL-HSC.zip’ saved [1707088358/1707088358]\n\n" ] ], [ [ "# Create a Conda environment", "_____no_output_____" ] ], [ [ "!./create_env.sh", "--2022-01-10 11:42:33-- https://repo.anaconda.com/miniconda/Miniconda3-py38_4.8.3-Linux-x86_64.sh\nResolving repo.anaconda.com (repo.anaconda.com)... 104.16.130.3, 104.16.131.3, 2606:4700::6810:8203, ...\nConnecting to repo.anaconda.com (repo.anaconda.com)|104.16.130.3|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 93052469 (89M) [application/x-sh]\nSaving to: ‘miniconda.sh’\n\nminiconda.sh 100%[===================>] 88.74M 14.4MB/s in 6.2s \n\n2022-01-10 11:42:40 (14.3 MB/s) - ‘miniconda.sh’ saved [93052469/93052469]\n\nPREFIX=/content/CTC-IGFL-FR-main/src/SW/miniconda\nUnpacking payload ...\nCollecting package metadata (current_repodata.json): - \b\b\\ \b\b| \b\bdone\nSolving environment: - \b\b\\ \b\bdone\n\n## Package Plan ##\n\n environment location: /content/CTC-IGFL-FR-main/src/SW/miniconda\n\n added / updated specs:\n - _libgcc_mutex==0.1=main\n - ca-certificates==2020.1.1=0\n - certifi==2020.4.5.1=py38_0\n - cffi==1.14.0=py38he30daa8_1\n - chardet==3.0.4=py38_1003\n - conda-package-handling==1.6.1=py38h7b6447c_0\n - conda==4.8.3=py38_0\n - cryptography==2.9.2=py38h1ba5d50_0\n - idna==2.9=py_1\n - ld_impl_linux-64==2.33.1=h53a641e_7\n - libedit==3.1.20181209=hc058e9b_0\n - libffi==3.3=he6710b0_1\n - libgcc-ng==9.1.0=hdf63c60_0\n - libstdcxx-ng==9.1.0=hdf63c60_0\n - ncurses==6.2=he6710b0_1\n - openssl==1.1.1g=h7b6447c_0\n - pip==20.0.2=py38_3\n - pycosat==0.6.3=py38h7b6447c_1\n - pycparser==2.20=py_0\n - pyopenssl==19.1.0=py38_0\n - pysocks==1.7.1=py38_0\n - python==3.8.3=hcff3b4d_0\n - readline==8.0=h7b6447c_0\n - requests==2.23.0=py38_0\n - ruamel_yaml==0.15.87=py38h7b6447c_0\n - setuptools==46.4.0=py38_0\n - six==1.14.0=py38_0\n - sqlite==3.31.1=h62c20be_1\n - tk==8.6.8=hbc83047_0\n - tqdm==4.46.0=py_0\n - urllib3==1.25.8=py38_0\n - wheel==0.34.2=py38_0\n - xz==5.2.5=h7b6447c_0\n - yaml==0.1.7=had09818_2\n - zlib==1.2.11=h7b6447c_3\n\n\nThe following NEW packages will be INSTALLED:\n\n _libgcc_mutex pkgs/main/linux-64::_libgcc_mutex-0.1-main\n ca-certificates pkgs/main/linux-64::ca-certificates-2020.1.1-0\n certifi pkgs/main/linux-64::certifi-2020.4.5.1-py38_0\n cffi pkgs/main/linux-64::cffi-1.14.0-py38he30daa8_1\n chardet pkgs/main/linux-64::chardet-3.0.4-py38_1003\n conda pkgs/main/linux-64::conda-4.8.3-py38_0\n conda-package-han~ pkgs/main/linux-64::conda-package-handling-1.6.1-py38h7b6447c_0\n cryptography pkgs/main/linux-64::cryptography-2.9.2-py38h1ba5d50_0\n idna pkgs/main/noarch::idna-2.9-py_1\n ld_impl_linux-64 pkgs/main/linux-64::ld_impl_linux-64-2.33.1-h53a641e_7\n libedit pkgs/main/linux-64::libedit-3.1.20181209-hc058e9b_0\n libffi pkgs/main/linux-64::libffi-3.3-he6710b0_1\n libgcc-ng pkgs/main/linux-64::libgcc-ng-9.1.0-hdf63c60_0\n libstdcxx-ng pkgs/main/linux-64::libstdcxx-ng-9.1.0-hdf63c60_0\n ncurses pkgs/main/linux-64::ncurses-6.2-he6710b0_1\n openssl pkgs/main/linux-64::openssl-1.1.1g-h7b6447c_0\n pip pkgs/main/linux-64::pip-20.0.2-py38_3\n pycosat pkgs/main/linux-64::pycosat-0.6.3-py38h7b6447c_1\n pycparser pkgs/main/noarch::pycparser-2.20-py_0\n pyopenssl pkgs/main/linux-64::pyopenssl-19.1.0-py38_0\n pysocks pkgs/main/linux-64::pysocks-1.7.1-py38_0\n python pkgs/main/linux-64::python-3.8.3-hcff3b4d_0\n readline pkgs/main/linux-64::readline-8.0-h7b6447c_0\n requests pkgs/main/linux-64::requests-2.23.0-py38_0\n ruamel_yaml pkgs/main/linux-64::ruamel_yaml-0.15.87-py38h7b6447c_0\n setuptools pkgs/main/linux-64::setuptools-46.4.0-py38_0\n six pkgs/main/linux-64::six-1.14.0-py38_0\n sqlite pkgs/main/linux-64::sqlite-3.31.1-h62c20be_1\n tk pkgs/main/linux-64::tk-8.6.8-hbc83047_0\n tqdm pkgs/main/noarch::tqdm-4.46.0-py_0\n urllib3 pkgs/main/linux-64::urllib3-1.25.8-py38_0\n wheel pkgs/main/linux-64::wheel-0.34.2-py38_0\n xz pkgs/main/linux-64::xz-5.2.5-h7b6447c_0\n yaml pkgs/main/linux-64::yaml-0.1.7-had09818_2\n zlib pkgs/main/linux-64::zlib-1.2.11-h7b6447c_3\n\n\nPreparing transaction: / \b\b- \b\b\\ \b\bdone\nExecuting transaction: / \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\bdone\ninstallation finished.\nWARNING:\n You currently have a PYTHONPATH environment variable set. This may cause\n unexpected behavior when running the Python interpreter in Miniconda3.\n For best results, please verify that your PYTHONPATH only points to\n directories of packages that are compatible with the Python interpreter\n in Miniconda3: /content/CTC-IGFL-FR-main/src/SW/miniconda\nCollecting package metadata (repodata.json): - \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\bdone\nSolving environment: / \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\bdone\n\n\n==> WARNING: A newer version of conda exists. <==\n current version: 4.8.3\n latest version: 4.11.0\n\nPlease update conda by running\n\n $ conda update -n base -c defaults conda\n\n\n\nDownloading and Extracting Packages\nptyprocess-0.6.0 | 23 KB | : 100% 1.0/1 [00:02<00:00, 2.38s/it]\nlibstdcxx-ng-9.1.0 | 4.0 MB | : 100% 1.0/1 [00:04<00:00, 4.61s/it]\nlibarchive-3.4.2 | 1.6 MB | : 100% 1.0/1 [00:03<00:00, 3.66s/it]\nfilelock-3.0.12 | 12 KB | : 100% 1.0/1 [00:02<00:00, 2.33s/it]\npython-dateutil-2.8. | 221 KB | : 100% 1.0/1 [00:02<00:00, 2.23s/it]\nzstd-1.4.5 | 716 KB | : 100% 1.0/1 [00:03<00:00, 3.39s/it]\nmatplotlib-base-3.2. | 7.1 MB | : 100% 1.0/1 [00:04<00:00, 4.44s/it]\nlcms2-2.11 | 419 KB | : 100% 1.0/1 [00:03<00:00, 3.03s/it]\ntornado-6.1 | 647 KB | : 100% 1.0/1 [00:03<00:00, 3.46s/it]\ncryptography-2.8 | 612 KB | : 100% 1.0/1 [00:03<00:00, 3.48s/it]\nintel-openmp-2019.4 | 876 KB | : 100% 1.0/1 [00:03<00:00, 3.56s/it]\nyaml-0.1.7 | 85 KB | : 100% 1.0/1 [00:01<00:00, 1.78s/it]\npysocks-1.7.1 | 30 KB | : 100% 1.0/1 [00:02<00:00, 2.31s/it]\nlibzopfli-1.0.3 | 179 KB | : 100% 1.0/1 [00:02<00:00, 2.03s/it]\nscikit-learn-0.23.1 | 6.8 MB | : 100% 1.0/1 [00:04<00:00, 4.53s/it]\nchardet-3.0.4 | 173 KB | : 100% 1.0/1 [00:02<00:00, 2.83s/it]\ncloudpickle-1.6.0 | 29 KB | : 100% 1.0/1 [00:02<00:00, 2.38s/it]\npygments-2.5.2 | 672 KB | : 100% 1.0/1 [00:03<00:00, 3.44s/it]\nfreetype-2.9.1 | 822 KB | : 100% 1.0/1 [00:02<00:00, 2.59s/it]\nbackcall-0.1.0 | 20 KB | : 100% 1.0/1 [00:01<00:00, 1.54s/it]\nblosc-1.21.0 | 74 KB | : 100% 1.0/1 [00:01<00:00, 1.81s/it]\ntqdm-4.48.2 | 63 KB | : 100% 1.0/1 [00:02<00:00, 2.59s/it]\ncffi-1.13.0 | 224 KB | : 100% 1.0/1 [00:03<00:00, 3.06s/it]\nmarkupsafe-1.1.1 | 29 KB | : 100% 1.0/1 [00:01<00:00, 1.54s/it]\nasn1crypto-1.2.0 | 162 KB | : 100% 1.0/1 [00:02<00:00, 2.84s/it]\npytz-2019.3 | 231 KB | : 100% 1.0/1 [00:03<00:00, 3.11s/it]\nreadline-7.0 | 392 KB | : 100% 1.0/1 [00:03<00:00, 3.31s/it]\nsnappy-1.1.8 | 43 KB | : 100% 1.0/1 [00:01<00:00, 1.83s/it]\nmkl-2019.4 | 204.1 MB | : 100% 1.0/1 [00:45<00:00, 45.59s/it]\npy-lief-0.9.0 | 1.5 MB | : 100% 1.0/1 [00:03<00:00, 3.06s/it]\nsetuptools-41.4.0 | 651 KB | : 100% 1.0/1 [00:03<00:00, 3.70s/it]\ncytoolz-0.11.0 | 367 KB | : 100% 1.0/1 [00:02<00:00, 2.31s/it]\nlibwebp-1.0.1 | 913 KB | : 100% 1.0/1 [00:02<00:00, 2.86s/it]\njoblib-1.0.1 | 207 KB | : 100% 1.0/1 [00:03<00:00, 3.13s/it]\nconda-4.9.2 | 3.1 MB | : 100% 1.0/1 [00:03<00:00, 3.57s/it]\nidna-2.8 | 101 KB | : 100% 1.0/1 [00:02<00:00, 2.02s/it]\nbeautifulsoup4-4.8.2 | 161 KB | : 100% 1.0/1 [00:02<00:00, 2.10s/it]\nlzo-2.10 | 313 KB | : 100% 1.0/1 [00:02<00:00, 2.33s/it]\njxrlib-1.1 | 238 KB | : 100% 1.0/1 [00:02<00:00, 2.99s/it]\nrequests-2.22.0 | 89 KB | : 100% 1.0/1 [00:00<00:00, 1.49it/s]\nblas-1.0 | 6 KB | : 100% 1.0/1 [00:01<00:00, 1.33s/it]\npywavelets-1.1.1 | 4.4 MB | : 100% 1.0/1 [00:04<00:00, 4.58s/it]\nparso-0.5.2 | 69 KB | : 100% 1.0/1 [00:02<00:00, 2.61s/it]\nwheel-0.33.6 | 40 KB | : 100% 1.0/1 [00:02<00:00, 2.53s/it]\nscipy-1.4.1 | 18.9 MB | : 100% 1.0/1 [00:07<00:00, 7.51s/it]\ndecorator-4.4.1 | 13 KB | : 100% 1.0/1 [00:02<00:00, 2.42s/it]\nipython_genutils-0.2 | 39 KB | : 100% 1.0/1 [00:02<00:00, 2.36s/it]\nurllib3-1.24.2 | 153 KB | : 100% 1.0/1 [00:01<00:00, 1.98s/it]\ncudatoolkit-10.1.243 | 513.2 MB | : 100% 1.0/1 [01:27<00:00, 87.91s/it]\nmkl_fft-1.0.15 | 172 KB | : 100% 1.0/1 [00:02<00:00, 2.87s/it]\nlibaec-1.0.4 | 35 KB | : 100% 1.0/1 [00:02<00:00, 2.36s/it]\njedi-0.15.2 | 759 KB | : 100% 1.0/1 [00:02<00:00, 3.00s/it]\nprompt_toolkit-3.0.2 | 234 KB | : 100% 1.0/1 [00:02<00:00, 2.35s/it]\nnumpy-1.17.4 | 4 KB | : 100% 1.0/1 [00:02<00:00, 2.06s/it]\npexpect-4.7.0 | 82 KB | : 100% 1.0/1 [00:01<00:00, 1.83s/it]\njpeg-9b | 248 KB | : 100% 1.0/1 [00:03<00:00, 3.04s/it]\nlibgfortran-ng-7.3.0 | 1.3 MB | : 100% 1.0/1 [00:02<00:00, 2.95s/it]\ndask-core-2021.2.0 | 681 KB | : 100% 1.0/1 [00:03<00:00, 3.35s/it]\npyopenssl-19.0.0 | 82 KB | : 100% 1.0/1 [00:02<00:00, 2.57s/it]\npyyaml-5.2 | 190 KB | : 100% 1.0/1 [00:01<00:00, 1.97s/it]\nopenssl-1.1.1i | 3.8 MB | : 100% 1.0/1 [00:04<00:00, 4.52s/it]\nmkl_random-1.1.0 | 376 KB | : 100% 1.0/1 [00:03<00:00, 3.12s/it]\nipython-7.11.1 | 1.1 MB | : 100% 1.0/1 [00:03<00:00, 3.73s/it]\npycosat-0.6.3 | 105 KB | : 100% 1.0/1 [00:01<00:00, 1.99s/it]\nlibffi-3.2.1 | 43 KB | : 100% 1.0/1 [00:02<00:00, 2.56s/it]\nliblief-0.9.0 | 4.4 MB | : 100% 1.0/1 [00:04<00:00, 4.69s/it]\nopenjpeg-2.3.0 | 456 KB | : 100% 1.0/1 [00:03<00:00, 3.26s/it]\ntifffile-2021.1.14 | 125 KB | : 100% 1.0/1 [00:02<00:00, 2.03s/it]\nnetworkx-2.5 | 1.2 MB | : 100% 1.0/1 [00:03<00:00, 3.66s/it]\npytorch-1.4.0 | 432.9 MB | : 100% 1.0/1 [01:22<00:00, 2653.98s/it] \npillow-7.0.0 | 657 KB | : 100% 1.0/1 [00:03<00:00, 3.38s/it]\nicu-58.2 | 22.7 MB | : 100% 1.0/1 [00:07<00:00, 7.52s/it]\nscikit-image-0.17.2 | 10.7 MB | : 100% 1.0/1 [00:04<00:00, 4.82s/it]\nzlib-1.2.11 | 120 KB | : 100% 1.0/1 [00:02<00:00, 2.86s/it]\npyparsing-2.4.7 | 59 KB | : 100% 1.0/1 [00:01<00:00, 1.81s/it]\ncycler-0.10.0 | 13 KB | : 100% 1.0/1 [00:02<00:00, 2.36s/it]\nlibxml2-2.9.9 | 2.0 MB | : 100% 1.0/1 [00:04<00:00, 4.10s/it]\nkiwisolver-1.3.1 | 86 KB | : 100% 1.0/1 [00:02<00:00, 2.53s/it]\nconda-build-3.18.11 | 526 KB | : 100% 1.0/1 [00:02<00:00, 2.57s/it]\ntk-8.6.8 | 3.1 MB | : 100% 1.0/1 [00:04<00:00, 4.15s/it]\npython-3.7.4 | 36.5 MB | : 100% 1.0/1 [00:09<00:00, 9.17s/it]\ncertifi-2020.12.5 | 143 KB | : 100% 1.0/1 [00:02<00:00, 2.01s/it]\nlz4-c-1.9.3 | 216 KB | : 100% 1.0/1 [00:03<00:00, 3.04s/it]\n_libgcc_mutex-0.1 | 3 KB | : 100% 1.0/1 [00:01<00:00, 1.39s/it]\npkginfo-1.5.0.1 | 43 KB | : 100% 1.0/1 [00:02<00:00, 2.56s/it]\ntraitlets-4.3.3 | 138 KB | : 100% 1.0/1 [00:02<00:00, 2.06s/it]\nlibtiff-4.1.0 | 607 KB | : 100% 1.0/1 [00:02<00:00, 2.61s/it]\nlibpng-1.6.37 | 364 KB | : 100% 1.0/1 [00:02<00:00, 2.34s/it]\nripgrep-11.0.2 | 1.5 MB | : 100% 1.0/1 [00:04<00:00, 4.47s/it]\npython-libarchive-c- | 23 KB | : 100% 1.0/1 [00:02<00:00, 2.34s/it]\nimagecodecs-2020.5.3 | 6.1 MB | : 100% 1.0/1 [00:04<00:00, 4.15s/it]\njinja2-2.10.3 | 95 KB | : 100% 1.0/1 [00:01<00:00, 1.76s/it]\npsutil-5.6.7 | 329 KB | : 100% 1.0/1 [00:03<00:00, 3.13s/it]\nmkl-service-2.3.0 | 208 KB | : 100% 1.0/1 [00:02<00:00, 2.30s/it]\nsix-1.12.0 | 22 KB | : 100% 1.0/1 [00:02<00:00, 2.31s/it]\nxz-5.2.5 | 438 KB | : 100% 1.0/1 [00:03<00:00, 3.34s/it]\ngiflib-5.1.4 | 78 KB | : 100% 1.0/1 [00:02<00:00, 2.63s/it]\ntorchvision-0.5.0 | 9.1 MB | : 100% 1.0/1 [00:04<00:00, 4.76s/it] \nwcwidth-0.1.7 | 23 KB | : 100% 1.0/1 [00:02<00:00, 2.52s/it]\npickleshare-0.7.5 | 13 KB | : 100% 1.0/1 [00:02<00:00, 2.35s/it]\nbrotli-1.0.9 | 402 KB | : 100% 1.0/1 [00:03<00:00, 3.07s/it]\nca-certificates-2021 | 128 KB | : 100% 1.0/1 [00:02<00:00, 2.76s/it]\ncharls-2.1.0 | 153 KB | : 100% 1.0/1 [00:02<00:00, 2.85s/it]\nimageio-2.9.0 | 3.1 MB | : 100% 1.0/1 [00:03<00:00, 3.38s/it]\nbzip2-1.0.8 | 105 KB | : 100% 1.0/1 [00:02<00:00, 2.04s/it]\nlibedit-3.1.20181209 | 188 KB | : 100% 1.0/1 [00:02<00:00, 2.95s/it]\nlibgcc-ng-9.1.0 | 8.1 MB | : 100% 1.0/1 [00:05<00:00, 5.26s/it]\npycparser-2.19 | 172 KB | : 100% 1.0/1 [00:02<00:00, 2.15s/it]\nsoupsieve-1.9.5 | 61 KB | : 100% 1.0/1 [00:01<00:00, 1.82s/it]\npip-19.3.1 | 1.9 MB | : 100% 1.0/1 [00:03<00:00, 3.33s/it]\nglob2-0.7 | 14 KB | : 100% 1.0/1 [00:02<00:00, 2.29s/it]\nruamel_yaml-0.15.46 | 245 KB | : 100% 1.0/1 [00:02<00:00, 2.30s/it]\nthreadpoolctl-2.1.0 | 16 KB | : 100% 1.0/1 [00:01<00:00, 1.52s/it]\nncurses-6.1 | 958 KB | : 100% 1.0/1 [00:03<00:00, 3.83s/it]\nninja-1.9.0 | 1.6 MB | : 100% 1.0/1 [00:02<00:00, 2.95s/it]\nnumpy-base-1.17.4 | 5.2 MB | : 100% 1.0/1 [00:04<00:00, 4.96s/it]\nsqlite-3.30.0 | 1.9 MB | : 100% 1.0/1 [00:03<00:00, 3.15s/it]\ntoolz-0.11.1 | 46 KB | : 100% 1.0/1 [00:02<00:00, 2.60s/it]\nconda-package-handli | 872 KB | : 100% 1.0/1 [00:03<00:00, 3.45s/it]\nolefile-0.46 | 48 KB | : 100% 1.0/1 [00:01<00:00, 1.83s/it]\npatchelf-0.10 | 78 KB | : 100% 1.0/1 [00:01<00:00, 1.84s/it]\nPreparing transaction: | \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\bdone\nVerifying transaction: - \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\bdone\nExecuting transaction: | \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\bdone\nRan pip subprocess with arguments:\n['/content/CTC-IGFL-FR-main/src/SW/miniconda/bin/python', '-m', 'pip', 'install', '-U', '-r', '/content/CTC-IGFL-FR-main/src/SW/condaenv.a_xir_p_.requirements.txt']\nPip subprocess output:\nProcessing ./elephant-core\nBuilding wheels for collected packages: elephant\n Building wheel for elephant (setup.py): started\n Building wheel for elephant (setup.py): finished with status 'done'\n Created wheel for elephant: filename=elephant-0.2.0-cp37-none-any.whl size=38137 sha256=b468f9426dfe42d9db1d77725f9a829b7cd7c1a88c53eef0f1591fdf8cd57e16\n Stored in directory: /root/.cache/pip/wheels/55/c3/79/36997bc76070e5066655d88bb260d9dc13f474fe5b27767e44\nSuccessfully built elephant\nInstalling collected packages: elephant\nSuccessfully installed elephant-0.2.0\n\n#\n# To activate this environment, use\n#\n# $ conda activate base\n#\n# To deactivate an active environment, use\n#\n# $ conda deactivate\n\nCollecting package metadata (repodata.json): - \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\bdone\nSolving environment: \\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\bdone\n\n\n==> WARNING: A newer version of conda exists. <==\n current version: 4.9.2\n latest version: 4.11.0\n\nPlease update conda by running\n\n $ conda update -n base conda\n\n\n\nDownloading and Extracting Packages\nfasteners-0.16 | 25 KB | : 100% 1.0/1 [00:00<00:00, 15.67it/s]\nasciitree-0.3.3 | 6 KB | : 100% 1.0/1 [00:00<00:00, 3.16it/s]\nzarr-2.4.0 | 95 KB | : 100% 1.0/1 [00:00<00:00, 2.88it/s] \nnumcodecs-0.7.2 | 960 KB | : 100% 1.0/1 [00:00<00:00, 1.35it/s]\nmonotonic-1.5 | 9 KB | : 100% 1.0/1 [00:00<00:00, 28.55it/s]\nmsgpack-python-1.0.0 | 91 KB | : 100% 1.0/1 [00:00<00:00, 3.08it/s] \npython_abi-3.7 | 4 KB | : 100% 1.0/1 [00:00<00:00, 39.54it/s]\ntensorboardx-2.1 | 80 KB | : 100% 1.0/1 [00:00<00:00, 2.91it/s] \nopenssl-1.1.1h | 2.1 MB | : 100% 1.0/1 [00:00<00:00, 3.21it/s]\nprotobuf-3.13.0.1 | 704 KB | : 100% 1.0/1 [00:00<00:00, 1.48it/s]\ncertifi-2021.10.8 | 145 KB | : 100% 1.0/1 [00:00<00:00, 19.51it/s]\nca-certificates-2021 | 139 KB | : 100% 1.0/1 [00:00<00:00, 14.18it/s]\nlibprotobuf-3.13.0.1 | 2.3 MB | : 100% 1.0/1 [00:00<00:00, 2.34it/s]\nPreparing transaction: / \b\bdone\nVerifying transaction: \\ \b\bdone\nExecuting transaction: / \b\b- \b\b\\ \b\b| \b\b/ \b\b- \b\b\\ \b\b| \b\bdone\n#\n# To activate this environment, use\n#\n# $ conda activate base\n#\n# To deactivate an active environment, use\n#\n# $ conda deactivate\n\n" ] ], [ [ "# Generate training datasets", "_____no_output_____" ] ], [ [ "!miniconda/bin/python prep/generate_seg_labels.py ../../Data/BF-C2DL-HSC train_data/BF-C2DL-HSC", "2D data found at ../../Data/BF-C2DL-HSC/01_GT/SEG\n(49, 1010, 1010) u1\n100% 49/49 [00:01<00:00, 25.00it/s]\n2D data found at ../../Data/BF-C2DL-HSC/02_GT/SEG\n(8, 1010, 1010) u1\n100% 8/8 [00:00<00:00, 13.72it/s]\n2D data found at ../../Data/BF-C2DL-HSC/01_ST/SEG\n(1764, 1010, 1010) u1\n100% 1764/1764 [00:59<00:00, 29.84it/s]\n2D data found at ../../Data/BF-C2DL-HSC/02_ST/SEG\n(1764, 1010, 1010) u1\n100% 1764/1764 [02:02<00:00, 14.39it/s]\n2D data found at ../../Data/BF-C2DL-HSC/01_GT/SEG\n(1764, 1010, 1010) u1\n100% 49/49 [00:01<00:00, 32.49it/s]\n100% 1764/1764 [00:52<00:00, 33.44it/s]\n2D data found at ../../Data/BF-C2DL-HSC/02_GT/SEG\n(1764, 1010, 1010) u1\n100% 8/8 [00:00<00:00, 14.54it/s]\n100% 1764/1764 [01:54<00:00, 15.43it/s]\n" ] ], [ [ "Please note that if your dataset contains image files with sparse annotations, you need to specify them using the `--sparse` argument.\n\ne.g. Fluo-C2DL-MSC_sparse.json\n\n```json\n{\n \"01\": [\n \"man_seg009.tif\",\n \"man_seg028.tif\",\n \"man_seg030.tif\",\n \"man_seg031.tif\",\n \"man_seg036.tif\",\n \"man_seg046.tif\",\n \"man_seg047.tif\"\n ],\n \"02\": [\n \"man_seg009.tif\",\n \"man_seg013.tif\",\n \"man_seg015.tif\",\n \"man_seg016.tif\"\n ]\n}\n```\n\n```bash\n!miniconda/bin/python prep/generate_seg_labels.py ../../Data/Fluo-C2DL-MSC train_data/Fluo-C2DL-MSC --sparse Fluo-C2DL-MSC_sparse.json\n```", "_____no_output_____" ], [ "# Prepare a training config file", "_____no_output_____" ] ], [ [ "!miniconda/bin/python generate_train_config.py training.json --dataset BF-C2DL-HSC/01-GT-seg BF-C2DL-HSC/02-GT-seg --model_name BF-C2DL-HSC-GT-seg.pth --log_dir BF-C2DL-HSC-GT-seg --n_epochs 10", "_____no_output_____" ] ], [ [ "# Run a training script", "_____no_output_____" ] ], [ [ "!miniconda/bin/python train.py seg training.json", "Train Epoch: 0 [0/1 (0%)]\tLoss: 0.455832\nTrain Epoch: 1 [0/1 (0%)]\tLoss: 0.298484\nTrain Epoch: 2 [0/1 (0%)]\tLoss: 0.268938\nauto_bg_thresh: 0.0\nbatch_size: 1\nc_ratio: None\nclass_weights: (1.0, 10.0, 10.0)\ncontrast: 0.0\ncrop_box: None\ncrop_size: (384, 384)\ndataset_name: BF-C2DL-HSC/01-GT-seg\ndebug: False\ndevice: cuda\nfalse_weight: None\nis_3d: False\nis_livemode: False\nis_pad: False\nkeep_axials: (True, True, True, True)\nlog_dir: logs/BF-C2DL-HSC-GT-seg\nlr: 0.0005\nmodel_path: models/BF-C2DL-HSC-GT-seg.pth\nn_crops: 1\nn_epochs: 10\noutput_prediction: False\np_thresh: None\npatch_size: [128, 128]\nr_max: None\nr_min: None\nrotation_angle: 0.0\nscale_factor_base: 0.0\nscales: None\ntimepoint: None\nuse_2d: False\nuse_median: None\nzpath_input: train_data/BF-C2DL-HSC/01-GT-seg/imgs.zarr\nzpath_seg_label: train_data/BF-C2DL-HSC/01-GT-seg/seg_labels.zarr\nzpath_seg_label_vis: train_data/BF-C2DL-HSC/01-GT-seg/seg_labels_vis.zarr\nzpath_seg_output: train_data/BF-C2DL-HSC/01-GT-seg/seg_outputs.zarr\nauto_bg_thresh: 0.0\nbatch_size: 1\nc_ratio: None\nclass_weights: (1.0, 10.0, 10.0)\ncontrast: 0.0\ncrop_box: None\ncrop_size: (384, 384)\ndataset_name: BF-C2DL-HSC/02-GT-seg\ndebug: False\ndevice: cuda\nfalse_weight: None\nis_3d: False\nis_livemode: False\nis_pad: False\nkeep_axials: (True, True, True, True)\nlog_dir: logs/BF-C2DL-HSC-GT-seg\nlr: 0.0005\nmodel_path: models/BF-C2DL-HSC-GT-seg.pth\nn_crops: 1\nn_epochs: 10\noutput_prediction: False\np_thresh: None\npatch_size: [128, 128]\nr_max: None\nr_min: None\nrotation_angle: 0.0\nscale_factor_base: 0.0\nscales: None\ntimepoint: None\nuse_2d: False\nuse_median: None\nzpath_input: train_data/BF-C2DL-HSC/02-GT-seg/imgs.zarr\nzpath_seg_label: train_data/BF-C2DL-HSC/02-GT-seg/seg_labels.zarr\nzpath_seg_label_vis: train_data/BF-C2DL-HSC/02-GT-seg/seg_labels_vis.zarr\nzpath_seg_output: train_data/BF-C2DL-HSC/02-GT-seg/seg_outputs.zarr\nTrain Epoch: 0 [0/53 (0%)]\tLoss: 1.038014\tNLL Loss: 1.380138\tCenter Dice Loss: 1.000000\tSmooth Loss: 0.032649\nEval Epoch: 0 \tLoss: 0.996351\nTrain Epoch: 1 [0/53 (0%)]\tLoss: 0.995269\tNLL Loss: 0.979194\tCenter Dice Loss: 0.997055\tSmooth Loss: 0.013280\nEval Epoch: 1 \tLoss: 0.984174\nTrain Epoch: 2 [0/53 (0%)]\tLoss: 0.983624\tNLL Loss: 0.860272\tCenter Dice Loss: 0.997330\tSmooth Loss: 0.011360\nEval Epoch: 2 \tLoss: 0.971131\nTrain Epoch: 3 [0/53 (0%)]\tLoss: 0.968924\tNLL Loss: 0.721982\tCenter Dice Loss: 0.996362\tSmooth Loss: 0.010758\nEval Epoch: 3 \tLoss: 0.956835\nTrain Epoch: 4 [0/53 (0%)]\tLoss: 0.953921\tNLL Loss: 0.587261\tCenter Dice Loss: 0.994661\tSmooth Loss: 0.009493\nEval Epoch: 4 \tLoss: 0.942152\nTrain Epoch: 5 [0/53 (0%)]\tLoss: 0.937996\tNLL Loss: 0.453298\tCenter Dice Loss: 0.991852\tSmooth Loss: 0.009450\nEval Epoch: 5 \tLoss: 0.928954\nTrain Epoch: 6 [0/53 (0%)]\tLoss: 0.921112\tNLL Loss: 0.337566\tCenter Dice Loss: 0.985950\tSmooth Loss: 0.008326\nEval Epoch: 6 \tLoss: 0.917517\nTrain Epoch: 7 [0/53 (0%)]\tLoss: 0.914061\tNLL Loss: 0.294186\tCenter Dice Loss: 0.982936\tSmooth Loss: 0.008099\nEval Epoch: 7 \tLoss: 0.906349\nTrain Epoch: 8 [0/53 (0%)]\tLoss: 0.890280\tNLL Loss: 0.219754\tCenter Dice Loss: 0.964782\tSmooth Loss: 0.008508\nEval Epoch: 8 \tLoss: 0.890048\nTrain Epoch: 9 [0/53 (0%)]\tLoss: 0.829467\tNLL Loss: 0.214251\tCenter Dice Loss: 0.897824\tSmooth Loss: 0.004281\nEval Epoch: 9 \tLoss: 0.864157\n" ] ], [ [ "# Download required libraries for inference", "_____no_output_____" ] ], [ [ "!./download_libraries.sh", "Downloading dependencies for Java\n--2022-01-10 12:17:30-- https://repo1.maven.org/maven2/net/imglib2/imglib2/5.6.3/imglib2-5.6.3.jar\nResolving repo1.maven.org (repo1.maven.org)... 199.232.192.209, 199.232.196.209\nConnecting to repo1.maven.org (repo1.maven.org)|199.232.192.209|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 762387 (745K) [application/java-archive]\nSaving to: ‘lib/imglib2-5.6.3.jar’\n\nimglib2-5.6.3.jar 100%[===================>] 744.52K 1.36MB/s in 0.5s \n\n2022-01-10 12:17:31 (1.36 MB/s) - ‘lib/imglib2-5.6.3.jar’ saved [762387/762387]\n\n--2022-01-10 12:17:31-- https://repo1.maven.org/maven2/gov/nist/math/jama/1.0.3/jama-1.0.3.jar\nResolving repo1.maven.org (repo1.maven.org)... 199.232.192.209, 199.232.196.209\nConnecting to repo1.maven.org (repo1.maven.org)|199.232.192.209|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 37424 (37K) [application/java-archive]\nSaving to: ‘lib/jama-1.0.3.jar’\n\njama-1.0.3.jar 100%[===================>] 36.55K --.-KB/s in 0.05s \n\n2022-01-10 12:17:32 (802 KB/s) - ‘lib/jama-1.0.3.jar’ saved [37424/37424]\n\n--2022-01-10 12:17:32-- http://maven.imagej.net/content/repositories/releases/org/mastodon/mastodon-collection/1.0.0-beta-17/mastodon-collection-1.0.0-beta-17.jar\nResolving maven.imagej.net (maven.imagej.net)... 144.92.48.199\nConnecting to maven.imagej.net (maven.imagej.net)|144.92.48.199|:80... connected.\nHTTP request sent, awaiting response... 301 Moved Permanently\nLocation: https://maven.scijava.org/content/repositories/releases/org/mastodon/mastodon-collection/1.0.0-beta-17/mastodon-collection-1.0.0-beta-17.jar [following]\n--2022-01-10 12:17:33-- https://maven.scijava.org/content/repositories/releases/org/mastodon/mastodon-collection/1.0.0-beta-17/mastodon-collection-1.0.0-beta-17.jar\nResolving maven.scijava.org (maven.scijava.org)... 144.92.48.199\nConnecting to maven.scijava.org (maven.scijava.org)|144.92.48.199|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 373090 (364K) [application/java-archive]\nSaving to: ‘lib/mastodon-collection-1.0.0-beta-17.jar’\n\nmastodon-collection 100%[===================>] 364.35K 519KB/s in 0.7s \n\n2022-01-10 12:17:34 (519 KB/s) - ‘lib/mastodon-collection-1.0.0-beta-17.jar’ saved [373090/373090]\n\n--2022-01-10 12:17:34-- http://maven.imagej.net/content/repositories/releases/org/mastodon/mastodon-graph/1.0.0-beta-16/mastodon-graph-1.0.0-beta-16.jar\nResolving maven.imagej.net (maven.imagej.net)... 144.92.48.199\nConnecting to maven.imagej.net (maven.imagej.net)|144.92.48.199|:80... connected.\nHTTP request sent, awaiting response... 301 Moved Permanently\nLocation: https://maven.scijava.org/content/repositories/releases/org/mastodon/mastodon-graph/1.0.0-beta-16/mastodon-graph-1.0.0-beta-16.jar [following]\n--2022-01-10 12:17:35-- https://maven.scijava.org/content/repositories/releases/org/mastodon/mastodon-graph/1.0.0-beta-16/mastodon-graph-1.0.0-beta-16.jar\nResolving maven.scijava.org (maven.scijava.org)... 144.92.48.199\nConnecting to maven.scijava.org (maven.scijava.org)|144.92.48.199|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 168824 (165K) [application/java-archive]\nSaving to: ‘lib/mastodon-graph-1.0.0-beta-16.jar’\n\nmastodon-graph-1.0. 100%[===================>] 164.87K 314KB/s in 0.5s \n\n2022-01-10 12:17:36 (314 KB/s) - ‘lib/mastodon-graph-1.0.0-beta-16.jar’ saved [168824/168824]\n\nFile ‘lib/mastodon-graph-1.0.0-beta-16.jar’ already there; not retrieving.\n\n--2022-01-10 12:17:36-- https://repo1.maven.org/maven2/com/eclipsesource/minimal-json/minimal-json/0.9.5/minimal-json-0.9.5.jar\nResolving repo1.maven.org (repo1.maven.org)... 199.232.192.209, 199.232.196.209\nConnecting to repo1.maven.org (repo1.maven.org)|199.232.192.209|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 34221 (33K) [application/java-archive]\nSaving to: ‘lib/minimal-json-0.9.5.jar’\n\nminimal-json-0.9.5. 100%[===================>] 33.42K --.-KB/s in 0.04s \n\n2022-01-10 12:17:36 (808 KB/s) - ‘lib/minimal-json-0.9.5.jar’ saved [34221/34221]\n\n--2022-01-10 12:17:36-- https://repo1.maven.org/maven2/com/opencsv/opencsv/3.9/opencsv-3.9.jar\nResolving repo1.maven.org (repo1.maven.org)... 199.232.192.209, 199.232.196.209\nConnecting to repo1.maven.org (repo1.maven.org)|199.232.192.209|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 80720 (79K) [application/java-archive]\nSaving to: ‘lib/opencsv-3.9.jar’\n\nopencsv-3.9.jar 100%[===================>] 78.83K 484KB/s in 0.2s \n\n2022-01-10 12:17:37 (484 KB/s) - ‘lib/opencsv-3.9.jar’ saved [80720/80720]\n\nFile ‘lib/opencsv-3.9.jar’ already there; not retrieving.\n\n--2022-01-10 12:17:37-- https://repo1.maven.org/maven2/net/sf/trove4j/trove4j/3.0.3/trove4j-3.0.3.jar\nResolving repo1.maven.org (repo1.maven.org)... 199.232.192.209, 199.232.196.209\nConnecting to repo1.maven.org (repo1.maven.org)|199.232.192.209|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 2523218 (2.4M) [application/java-archive]\nSaving to: ‘lib/trove4j-3.0.3.jar’\n\ntrove4j-3.0.3.jar 100%[===================>] 2.41M 3.21MB/s in 0.8s \n\n2022-01-10 12:17:38 (3.21 MB/s) - ‘lib/trove4j-3.0.3.jar’ saved [2523218/2523218]\n\n--2022-01-10 12:17:38-- https://sites.imagej.net/Mastodonpreview/jars/trackmate-1.0.0-beta-13.jar-20190320130043\nResolving sites.imagej.net (sites.imagej.net)... 144.92.48.186\nConnecting to sites.imagej.net (sites.imagej.net)|144.92.48.186|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 1084822 (1.0M)\nSaving to: ‘lib/trackmate-1.0.0-beta-13.jar’\n\nlib/trackmate-1.0.0 100%[===================>] 1.03M 1003KB/s in 1.1s \n\n2022-01-10 12:17:41 (1003 KB/s) - ‘lib/trackmate-1.0.0-beta-13.jar’ saved [1084822/1084822]\n\n" ] ], [ [ "# Prepare inference config files", "_____no_output_____" ] ], [ [ "!miniconda/bin/python generate_run_config.py run.json --seg_model models/BF-C2DL-HSC-GT-seg.pth --scales 0.645 0.645 --c_ratio 0.3 --p_thresh 0.8 --r_min 1 --r_max 50 --use_interpolation", "_____no_output_____" ] ], [ [ "# Run a Java program for inference", "_____no_output_____" ] ], [ [ "!java -jar elephant-ctc-0.1.0.jar \"../../Data/BF-C2DL-HSC/01\" \"../../Data/BF-C2DL-HSC/01_RES-allGT\" \"run.json\"", "\u001b[1;30;43mStreaming output truncated to the last 5000 lines.\u001b[0m\nIncrease radii from ([0.5324 1.438675]) to ([0.58564 1.5825425])\nIncrease radii from ([0.58564 1.5825425]) to ([0.644204 1.74079675])\nIncrease radii from ([0.4 1.05178458]) to ([0.44 1.15696304])\nIncrease radii from ([0.44 1.15696304]) to ([0.484 1.27265934])\nIncrease radii from ([0.484 1.27265934]) to ([0.5324 1.39992527])\nIncrease radii from ([0.5324 1.39992527]) to ([0.58564 1.5399178])\nIncrease radii from ([0.58564 1.5399178]) to ([0.644204 1.69390958])\nIncrease radii from ([0.4 1.09060011]) to ([0.44 1.19966012])\nIncrease radii from ([0.44 1.19966012]) to ([0.484 1.31962613])\nIncrease radii from ([0.484 1.31962613]) to ([0.5324 1.45158874])\nIncrease radii from ([0.5324 1.45158874]) to ([0.58564 1.59674762])\nIncrease radii from ([0.58564 1.59674762]) to ([0.644204 1.75642238])\nIncrease radii from ([0.35 0.94578559]) to ([0.385 1.04036415])\nIncrease radii from ([0.385 1.04036415]) to ([0.4235 1.14440057])\nIncrease radii from ([0.4235 1.14440057]) to ([0.46585 1.25884063])\nIncrease radii from ([0.46585 1.25884063]) to ([0.512435 1.38472469])\nIncrease radii from ([0.512435 1.38472469]) to ([0.5636785 1.52319716])\nIncrease radii from ([0.5636785 1.52319716]) to ([0.62004635 1.67551687])\nIncrease radii from ([0.35 0.92031151]) to ([0.385 1.01234266])\nIncrease radii from ([0.385 1.01234266]) to ([0.4235 1.11357692])\nIncrease radii from ([0.4235 1.11357692]) to ([0.46585 1.22493462])\nIncrease radii from ([0.46585 1.22493462]) to ([0.512435 1.34742808])\nIncrease radii from ([0.512435 1.34742808]) to ([0.5636785 1.48217088])\nIncrease radii from ([0.5636785 1.48217088]) to ([0.62004635 1.63038797])\nIncrease radii from ([0.35 0.95427509]) to ([0.385 1.0497026])\nIncrease radii from ([0.385 1.0497026]) to ([0.4235 1.15467286])\nIncrease radii from ([0.4235 1.15467286]) to ([0.46585 1.27014015])\nIncrease radii from ([0.46585 1.27014015]) to ([0.512435 1.39715417])\nIncrease radii from ([0.512435 1.39715417]) to ([0.5636785 1.53686958])\nIncrease radii from ([0.5636785 1.53686958]) to ([0.62004635 1.69055654])\nIncrease radii from ([0.3 0.81067337]) to ([0.33 0.8917407])\nIncrease radii from ([0.33 0.8917407]) to ([0.363 0.98091477])\nIncrease radii from ([0.363 0.98091477]) to ([0.3993 1.07900625])\nIncrease radii from ([0.3993 1.07900625]) to ([0.43923 1.18690688])\nIncrease radii from ([0.43923 1.18690688]) to ([0.483153 1.30559756])\nIncrease radii from ([0.483153 1.30559756]) to ([0.5314683 1.43615732])\nIncrease radii from ([0.5314683 1.43615732]) to ([0.58461513 1.57977305])\nIncrease radii from ([0.58461513 1.57977305]) to ([0.64307664 1.73775036])\nIncrease radii from ([0.3 1.54638339]) to ([0.33 1.70102173])\nIncrease radii from ([0.3 1.14671225]) to ([0.33 1.26138347])\nIncrease radii from ([0.33 1.26138347]) to ([0.363 1.38752182])\nIncrease radii from ([0.3 0.78883843]) to ([0.33 0.86772228])\nIncrease radii from ([0.33 0.86772228]) to ([0.363 0.95449451])\nIncrease radii from ([0.363 0.95449451]) to ([0.3993 1.04994396])\nIncrease radii from ([0.3993 1.04994396]) to ([0.43923 1.15493835])\nIncrease radii from ([0.43923 1.15493835]) to ([0.483153 1.27043219])\nIncrease radii from ([0.483153 1.27043219]) to ([0.5314683 1.39747541])\nIncrease radii from ([0.5314683 1.39747541]) to ([0.58461513 1.53722295])\nIncrease radii from ([0.58461513 1.53722295]) to ([0.64307664 1.69094524])\nIncrease radii from ([0.3 0.81795008]) to ([0.33 0.89974509])\nIncrease radii from ([0.33 0.89974509]) to ([0.363 0.9897196])\nIncrease radii from ([0.363 0.9897196]) to ([0.3993 1.08869156])\nIncrease radii from ([0.3993 1.08869156]) to ([0.43923 1.19756071])\nIncrease radii from ([0.43923 1.19756071]) to ([0.483153 1.31731678])\nIncrease radii from ([0.483153 1.31731678]) to ([0.5314683 1.44904846])\nIncrease radii from ([0.5314683 1.44904846]) to ([0.58461513 1.59395331])\nIncrease radii from ([0.58461513 1.59395331]) to ([0.64307664 1.75334864])\nIncrease radii from ([0.25 0.67556114]) to ([0.275 0.74311725])\nIncrease radii from ([0.275 0.74311725]) to ([0.3025 0.81742898])\nIncrease radii from ([0.3025 0.81742898]) to ([0.33275 0.89917188])\nIncrease radii from ([0.33275 0.89917188]) to ([0.366025 0.98908906])\nIncrease radii from ([0.366025 0.98908906]) to ([0.4026275 1.08799797])\nIncrease radii from ([0.4026275 1.08799797]) to ([0.44289025 1.19679777])\nIncrease radii from ([0.44289025 1.19679777]) to ([0.48717928 1.31647754])\nIncrease radii from ([0.48717928 1.31647754]) to ([0.5358972 1.4481253])\nIncrease radii from ([0.5358972 1.4481253]) to ([0.58948692 1.59293783])\nIncrease radii from ([0.58948692 1.59293783]) to ([0.64843562 1.75223161])\nIncrease radii from ([0.25 1.28865282]) to ([0.275 1.41751811])\nIncrease radii from ([0.275 1.41751811]) to ([0.3025 1.55926992])\nIncrease radii from ([0.3025 1.55926992]) to ([0.33275 1.71519691])\nIncrease radii from ([0.25 0.95559354]) to ([0.275 1.05115289])\nIncrease radii from ([0.275 1.05115289]) to ([0.3025 1.15626818])\nIncrease radii from ([0.3025 1.15626818]) to ([0.33275 1.271895])\nIncrease radii from ([0.33275 1.271895]) to ([0.366025 1.3990845])\nIncrease radii from ([0.25 0.65736536]) to ([0.275 0.7231019])\nIncrease radii from ([0.275 0.7231019]) to ([0.3025 0.79541209])\nIncrease radii from ([0.3025 0.79541209]) to ([0.33275 0.8749533])\nIncrease radii from ([0.33275 0.8749533]) to ([0.366025 0.96244863])\nIncrease radii from ([0.366025 0.96244863]) to ([0.4026275 1.05869349])\nIncrease radii from ([0.4026275 1.05869349]) to ([0.44289025 1.16456284])\nIncrease radii from ([0.44289025 1.16456284]) to ([0.48717928 1.28101912])\nIncrease radii from ([0.48717928 1.28101912]) to ([0.5358972 1.40912103])\nIncrease radii from ([0.5358972 1.40912103]) to ([0.58948692 1.55003314])\nIncrease radii from ([0.58948692 1.55003314]) to ([0.64843562 1.70503645])\nIncrease radii from ([0.25 0.68162507]) to ([0.275 0.74978757])\nIncrease radii from ([0.275 0.74978757]) to ([0.3025 0.82476633])\nIncrease radii from ([0.3025 0.82476633]) to ([0.33275 0.90724296])\nIncrease radii from ([0.33275 0.90724296]) to ([0.366025 0.99796726])\nIncrease radii from ([0.366025 0.99796726]) to ([0.4026275 1.09776399])\nIncrease radii from ([0.4026275 1.09776399]) to ([0.44289025 1.20754039])\nIncrease radii from ([0.44289025 1.20754039]) to ([0.48717928 1.32829442])\nIncrease radii from ([0.48717928 1.32829442]) to ([0.5358972 1.46112387])\nIncrease radii from ([0.5358972 1.46112387]) to ([0.58948692 1.60723625])\nIncrease radii from ([0.58948692 1.60723625]) to ([0.64843562 1.76795988])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1700.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.57767687]) to ([0.66 1.73544456])\nIncrease radii from ([0.6 1.55512011]) to ([0.66 1.71063212])\nIncrease radii from ([0.6 1.59073913]) to ([0.66 1.74981305])\nIncrease radii from ([0.55 1.4462038]) to ([0.605 1.59082418])\nIncrease radii from ([0.605 1.59082418]) to ([0.6655 1.74990659])\nIncrease radii from ([0.55 1.42552677]) to ([0.605 1.56807945])\nIncrease radii from ([0.605 1.56807945]) to ([0.6655 1.72488739])\nIncrease radii from ([0.55 1.45817754]) to ([0.605 1.60399529])\nIncrease radii from ([0.605 1.60399529]) to ([0.6655 1.76439482])\nIncrease radii from ([0.5 1.31473072]) to ([0.55 1.4462038])\nIncrease radii from ([0.55 1.4462038]) to ([0.605 1.59082418])\nIncrease radii from ([0.605 1.59082418]) to ([0.6655 1.74990659])\nIncrease radii from ([0.5 1.29593343]) to ([0.55 1.42552677])\nIncrease radii from ([0.55 1.42552677]) to ([0.605 1.56807945])\nIncrease radii from ([0.605 1.56807945]) to ([0.6655 1.72488739])\nIncrease radii from ([0.5 1.32561594]) to ([0.55 1.45817754])\nIncrease radii from ([0.55 1.45817754]) to ([0.605 1.60399529])\nIncrease radii from ([0.605 1.60399529]) to ([0.6655 1.76439482])\nIncrease radii from ([0.45 1.18325765]) to ([0.495 1.30158342])\nIncrease radii from ([0.495 1.30158342]) to ([0.5445 1.43174176])\nIncrease radii from ([0.5445 1.43174176]) to ([0.59895 1.57491593])\nIncrease radii from ([0.59895 1.57491593]) to ([0.658845 1.73240753])\nIncrease radii from ([0.45 1.16634008]) to ([0.495 1.28297409])\nIncrease radii from ([0.495 1.28297409]) to ([0.5445 1.4112715])\nIncrease radii from ([0.5445 1.4112715]) to ([0.59895 1.55239865])\nIncrease radii from ([0.59895 1.55239865]) to ([0.658845 1.70763852])\nIncrease radii from ([0.45 1.19305435]) to ([0.495 1.31235978])\nIncrease radii from ([0.495 1.31235978]) to ([0.5445 1.44359576])\nIncrease radii from ([0.5445 1.44359576]) to ([0.59895 1.58795534])\nIncrease radii from ([0.59895 1.58795534]) to ([0.658845 1.74675087])\nIncrease radii from ([0.4 1.05178458]) to ([0.44 1.15696304])\nIncrease radii from ([0.44 1.15696304]) to ([0.484 1.27265934])\nIncrease radii from ([0.484 1.27265934]) to ([0.5324 1.39992527])\nIncrease radii from ([0.5324 1.39992527]) to ([0.58564 1.5399178])\nIncrease radii from ([0.58564 1.5399178]) to ([0.644204 1.69390958])\nIncrease radii from ([0.4 1.03674674]) to ([0.44 1.14042142])\nIncrease radii from ([0.44 1.14042142]) to ([0.484 1.25446356])\nIncrease radii from ([0.484 1.25446356]) to ([0.5324 1.37990991])\nIncrease radii from ([0.5324 1.37990991]) to ([0.58564 1.51790091])\nIncrease radii from ([0.58564 1.51790091]) to ([0.644204 1.669691])\nIncrease radii from ([0.4 1.06049276]) to ([0.44 1.16654203])\nIncrease radii from ([0.44 1.16654203]) to ([0.484 1.28319623])\nIncrease radii from ([0.484 1.28319623]) to ([0.5324 1.41151586])\nIncrease radii from ([0.5324 1.41151586]) to ([0.58564 1.55266744])\nIncrease radii from ([0.58564 1.55266744]) to ([0.644204 1.70793419])\nIncrease radii from ([0.35 0.92031151]) to ([0.385 1.01234266])\nIncrease radii from ([0.385 1.01234266]) to ([0.4235 1.11357692])\nIncrease radii from ([0.4235 1.11357692]) to ([0.46585 1.22493462])\nIncrease radii from ([0.46585 1.22493462]) to ([0.512435 1.34742808])\nIncrease radii from ([0.512435 1.34742808]) to ([0.5636785 1.48217088])\nIncrease radii from ([0.5636785 1.48217088]) to ([0.62004635 1.63038797])\nIncrease radii from ([0.35 0.9071534]) to ([0.385 0.99786874])\nIncrease radii from ([0.385 0.99786874]) to ([0.4235 1.09765561])\nIncrease radii from ([0.4235 1.09765561]) to ([0.46585 1.20742117])\nIncrease radii from ([0.46585 1.20742117]) to ([0.512435 1.32816329])\nIncrease radii from ([0.512435 1.32816329]) to ([0.5636785 1.46097962])\nIncrease radii from ([0.5636785 1.46097962]) to ([0.62004635 1.60707758])\nIncrease radii from ([0.62004635 1.60707758]) to ([0.68205099 1.76778534])\nIncrease radii from ([0.35 0.92793116]) to ([0.385 1.02072428])\nIncrease radii from ([0.385 1.02072428]) to ([0.4235 1.1227967])\nIncrease radii from ([0.4235 1.1227967]) to ([0.46585 1.23507637])\nIncrease radii from ([0.46585 1.23507637]) to ([0.512435 1.35858401])\nIncrease radii from ([0.512435 1.35858401]) to ([0.5636785 1.49444241])\nIncrease radii from ([0.5636785 1.49444241]) to ([0.62004635 1.64388666])\nIncrease radii from ([0.3 1.54638339]) to ([0.33 1.70102173])\nIncrease radii from ([0.3 0.78883843]) to ([0.33 0.86772228])\nIncrease radii from ([0.33 0.86772228]) to ([0.363 0.95449451])\nIncrease radii from ([0.363 0.95449451]) to ([0.3993 1.04994396])\nIncrease radii from ([0.3993 1.04994396]) to ([0.43923 1.15493835])\nIncrease radii from ([0.43923 1.15493835]) to ([0.483153 1.27043219])\nIncrease radii from ([0.483153 1.27043219]) to ([0.5314683 1.39747541])\nIncrease radii from ([0.5314683 1.39747541]) to ([0.58461513 1.53722295])\nIncrease radii from ([0.58461513 1.53722295]) to ([0.64307664 1.69094524])\nIncrease radii from ([0.3 0.77756006]) to ([0.33 0.85531606])\nIncrease radii from ([0.33 0.85531606]) to ([0.363 0.94084767])\nIncrease radii from ([0.363 0.94084767]) to ([0.3993 1.03493244])\nIncrease radii from ([0.3993 1.03493244]) to ([0.43923 1.13842568])\nIncrease radii from ([0.43923 1.13842568]) to ([0.483153 1.25226825])\nIncrease radii from ([0.483153 1.25226825]) to ([0.5314683 1.37749507])\nIncrease radii from ([0.5314683 1.37749507]) to ([0.58461513 1.51524458])\nIncrease radii from ([0.58461513 1.51524458]) to ([0.64307664 1.66676904])\nIncrease radii from ([0.3 0.79536957]) to ([0.33 0.87490652])\nIncrease radii from ([0.33 0.87490652]) to ([0.363 0.96239718])\nIncrease radii from ([0.363 0.96239718]) to ([0.3993 1.05863689])\nIncrease radii from ([0.3993 1.05863689]) to ([0.43923 1.16450058])\nIncrease radii from ([0.43923 1.16450058]) to ([0.483153 1.28095064])\nIncrease radii from ([0.483153 1.28095064]) to ([0.5314683 1.4090457])\nIncrease radii from ([0.5314683 1.4090457]) to ([0.58461513 1.54995027])\nIncrease radii from ([0.58461513 1.54995027]) to ([0.64307664 1.7049453 ])\nIncrease radii from ([0.25 1.28865282]) to ([0.275 1.41751811])\nIncrease radii from ([0.275 1.41751811]) to ([0.3025 1.55926992])\nIncrease radii from ([0.3025 1.55926992]) to ([0.33275 1.71519691])\nIncrease radii from ([0.25 0.65736536]) to ([0.275 0.7231019])\nIncrease radii from ([0.275 0.7231019]) to ([0.3025 0.79541209])\nIncrease radii from ([0.3025 0.79541209]) to ([0.33275 0.8749533])\nIncrease radii from ([0.33275 0.8749533]) to ([0.366025 0.96244863])\nIncrease radii from ([0.366025 0.96244863]) to ([0.4026275 1.05869349])\nIncrease radii from ([0.4026275 1.05869349]) to ([0.44289025 1.16456284])\nIncrease radii from ([0.44289025 1.16456284]) to ([0.48717928 1.28101912])\nIncrease radii from ([0.48717928 1.28101912]) to ([0.5358972 1.40912103])\nIncrease radii from ([0.5358972 1.40912103]) to ([0.58948692 1.55003314])\nIncrease radii from ([0.58948692 1.55003314]) to ([0.64843562 1.70503645])\nIncrease radii from ([0.25 0.64796671]) to ([0.275 0.71276339])\nIncrease radii from ([0.275 0.71276339]) to ([0.3025 0.78403972])\nIncrease radii from ([0.3025 0.78403972]) to ([0.33275 0.8624437])\nIncrease radii from ([0.33275 0.8624437]) to ([0.366025 0.94868807])\nIncrease radii from ([0.366025 0.94868807]) to ([0.4026275 1.04355687])\nIncrease radii from ([0.4026275 1.04355687]) to ([0.44289025 1.14791256])\nIncrease radii from ([0.44289025 1.14791256]) to ([0.48717928 1.26270382])\nIncrease radii from ([0.48717928 1.26270382]) to ([0.5358972 1.3889742])\nIncrease radii from ([0.5358972 1.3889742]) to ([0.58948692 1.52787162])\nIncrease radii from ([0.58948692 1.52787162]) to ([0.64843562 1.68065878])\nIncrease radii from ([0.25 0.66280797]) to ([0.275 0.72908877])\nIncrease radii from ([0.275 0.72908877]) to ([0.3025 0.80199765])\nIncrease radii from ([0.3025 0.80199765]) to ([0.33275 0.88219741])\nIncrease radii from ([0.33275 0.88219741]) to ([0.366025 0.97041715])\nIncrease radii from ([0.366025 0.97041715]) to ([0.4026275 1.06745887])\nIncrease radii from ([0.4026275 1.06745887]) to ([0.44289025 1.17420475])\nIncrease radii from ([0.44289025 1.17420475]) to ([0.48717928 1.29162523])\nIncrease radii from ([0.48717928 1.29162523]) to ([0.5358972 1.42078775])\nIncrease radii from ([0.5358972 1.42078775]) to ([0.58948692 1.56286653])\nIncrease radii from ([0.58948692 1.56286653]) to ([0.64843562 1.71915318])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1701.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.60309105]) to ([0.66 1.76340015])\nIncrease radii from ([0.55 1.46950013]) to ([0.605 1.61645014])\nIncrease radii from ([0.5 1.33590921]) to ([0.55 1.46950013])\nIncrease radii from ([0.55 1.46950013]) to ([0.605 1.61645014])\nIncrease radii from ([0.45 1.20231829]) to ([0.495 1.32255012])\nIncrease radii from ([0.495 1.32255012]) to ([0.5445 1.45480513])\nIncrease radii from ([0.5445 1.45480513]) to ([0.59895 1.60028564])\nIncrease radii from ([0.59895 1.60028564]) to ([0.658845 1.7603142])\nIncrease radii from ([0.4 1.06872737]) to ([0.44 1.1756001])\nIncrease radii from ([0.44 1.1756001]) to ([0.484 1.29316011])\nIncrease radii from ([0.484 1.29316011]) to ([0.5324 1.42247612])\nIncrease radii from ([0.5324 1.42247612]) to ([0.58564 1.56472374])\nIncrease radii from ([0.58564 1.56472374]) to ([0.644204 1.72119611])\nIncrease radii from ([0.35 0.93513645]) to ([0.385 1.02865009])\nIncrease radii from ([0.385 1.02865009]) to ([0.4235 1.1315151])\nIncrease radii from ([0.4235 1.1315151]) to ([0.46585 1.24466661])\nIncrease radii from ([0.46585 1.24466661]) to ([0.512435 1.36913327])\nIncrease radii from ([0.512435 1.36913327]) to ([0.5636785 1.5060466])\nIncrease radii from ([0.5636785 1.5060466]) to ([0.62004635 1.65665126])\nIncrease radii from ([0.3 1.57503745]) to ([0.33 1.7325412])\nIncrease radii from ([0.3 1.15602576]) to ([0.33 1.27162834])\nIncrease radii from ([0.33 1.27162834]) to ([0.363 1.39879117])\nIncrease radii from ([0.3 1.15418797]) to ([0.33 1.26960677])\nIncrease radii from ([0.33 1.26960677]) to ([0.363 1.39656744])\nIncrease radii from ([0.3 0.80154552]) to ([0.33 0.88170008])\nIncrease radii from ([0.33 0.88170008]) to ([0.363 0.96987008])\nIncrease radii from ([0.363 0.96987008]) to ([0.3993 1.06685709])\nIncrease radii from ([0.3993 1.06685709]) to ([0.43923 1.1735428])\nIncrease radii from ([0.43923 1.1735428]) to ([0.483153 1.29089708])\nIncrease radii from ([0.483153 1.29089708]) to ([0.5314683 1.41998679])\nIncrease radii from ([0.5314683 1.41998679]) to ([0.58461513 1.56198547])\nIncrease radii from ([0.58461513 1.56198547]) to ([0.64307664 1.71818402])\nIncrease radii from ([0.25 1.31253121]) to ([0.275 1.44378433])\nIncrease radii from ([0.275 1.44378433]) to ([0.3025 1.58816276])\nIncrease radii from ([0.3025 1.58816276]) to ([0.33275 1.74697904])\nIncrease radii from ([0.25 0.9633548]) to ([0.275 1.05969028])\nIncrease radii from ([0.275 1.05969028]) to ([0.3025 1.16565931])\nIncrease radii from ([0.3025 1.16565931]) to ([0.33275 1.28222524])\nIncrease radii from ([0.33275 1.28222524]) to ([0.366025 1.41044776])\nIncrease radii from ([0.25 0.96182331]) to ([0.275 1.05800564])\nIncrease radii from ([0.275 1.05800564]) to ([0.3025 1.1638062])\nIncrease radii from ([0.3025 1.1638062]) to ([0.33275 1.28018682])\nIncrease radii from ([0.33275 1.28018682]) to ([0.366025 1.40820551])\nIncrease radii from ([0.43591192 0.75502159]) to ([0.47950311 0.83052375])\nIncrease radii from ([0.25 0.6679546]) to ([0.275 0.73475006])\nIncrease radii from ([0.275 0.73475006]) to ([0.3025 0.80822507])\nIncrease radii from ([0.3025 0.80822507]) to ([0.33275 0.88904758])\nIncrease radii from ([0.33275 0.88904758]) to ([0.366025 0.97795234])\nIncrease radii from ([0.366025 0.97795234]) to ([0.4026275 1.07574757])\nIncrease radii from ([0.4026275 1.07574757]) to ([0.44289025 1.18332233])\nIncrease radii from ([0.44289025 1.18332233]) to ([0.48717928 1.30165456])\nIncrease radii from ([0.48717928 1.30165456]) to ([0.5358972 1.43182001])\nIncrease radii from ([0.5358972 1.43182001]) to ([0.58948692 1.57500202])\nIncrease radii from ([0.58948692 1.57500202]) to ([0.64843562 1.73250222])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1702.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.58736605]) to ([0.66 1.74610265])\nIncrease radii from ([0.55 1.45508554]) to ([0.605 1.6005941])\nIncrease radii from ([0.605 1.6005941]) to ([0.6655 1.76065351])\nIncrease radii from ([0.55 1.49659162]) to ([0.605 1.64625079])\nIncrease radii from ([0.5 1.32280504]) to ([0.55 1.45508554])\nIncrease radii from ([0.55 1.45508554]) to ([0.605 1.6005941])\nIncrease radii from ([0.605 1.6005941]) to ([0.6655 1.76065351])\nIncrease radii from ([0.5 1.36053784]) to ([0.55 1.49659162])\nIncrease radii from ([0.55 1.49659162]) to ([0.605 1.64625079])\nIncrease radii from ([0.45 1.19052453]) to ([0.495 1.30957699])\nIncrease radii from ([0.495 1.30957699]) to ([0.5445 1.44053469])\nIncrease radii from ([0.5445 1.44053469]) to ([0.59895 1.58458816])\nIncrease radii from ([0.59895 1.58458816]) to ([0.658845 1.74304697])\nIncrease radii from ([0.45 1.22448406]) to ([0.495 1.34693246])\nIncrease radii from ([0.495 1.34693246]) to ([0.5445 1.48162571])\nIncrease radii from ([0.5445 1.48162571]) to ([0.59895 1.62978828])\nIncrease radii from ([0.4 1.05824403]) to ([0.44 1.16406843])\nIncrease radii from ([0.44 1.16406843]) to ([0.484 1.28047528])\nIncrease radii from ([0.484 1.28047528]) to ([0.5324 1.4085228])\nIncrease radii from ([0.5324 1.4085228]) to ([0.58564 1.54937509])\nIncrease radii from ([0.58564 1.54937509]) to ([0.644204 1.70431259])\nIncrease radii from ([0.4 1.08843027]) to ([0.44 1.1972733])\nIncrease radii from ([0.44 1.1972733]) to ([0.484 1.31700063])\nIncrease radii from ([0.484 1.31700063]) to ([0.5324 1.44870069])\nIncrease radii from ([0.5324 1.44870069]) to ([0.58564 1.59357076])\nIncrease radii from ([0.58564 1.59357076]) to ([0.644204 1.75292784])\nIncrease radii from ([0.35 0.92596353]) to ([0.385 1.01855988])\nIncrease radii from ([0.385 1.01855988]) to ([0.4235 1.12041587])\nIncrease radii from ([0.4235 1.12041587]) to ([0.46585 1.23245745])\nIncrease radii from ([0.46585 1.23245745]) to ([0.512435 1.3557032])\nIncrease radii from ([0.512435 1.3557032]) to ([0.5636785 1.49127352])\nIncrease radii from ([0.5636785 1.49127352]) to ([0.62004635 1.64040087])\nIncrease radii from ([0.35 0.95237649]) to ([0.385 1.04761414])\nIncrease radii from ([0.385 1.04761414]) to ([0.4235 1.15237555])\nIncrease radii from ([0.4235 1.15237555]) to ([0.46585 1.26761311])\nIncrease radii from ([0.46585 1.26761311]) to ([0.512435 1.39437442])\nIncrease radii from ([0.512435 1.39437442]) to ([0.5636785 1.53381186])\nIncrease radii from ([0.5636785 1.53381186]) to ([0.62004635 1.68719304])\nIncrease radii from ([0.3 1.57963328]) to ([0.33 1.73759661])\nIncrease radii from ([0.3 0.79368302]) to ([0.33 0.87305133])\nIncrease radii from ([0.33 0.87305133]) to ([0.363 0.96035646])\nIncrease radii from ([0.363 0.96035646]) to ([0.3993 1.0563921])\nIncrease radii from ([0.3993 1.0563921]) to ([0.43923 1.16203131])\nIncrease radii from ([0.43923 1.16203131]) to ([0.483153 1.27823445])\nIncrease radii from ([0.483153 1.27823445]) to ([0.5314683 1.40605789])\nIncrease radii from ([0.5314683 1.40605789]) to ([0.58461513 1.54666368])\nIncrease radii from ([0.58461513 1.54666368]) to ([0.64307664 1.70133005])\nIncrease radii from ([0.3 1.16669408]) to ([0.33 1.28336349])\nIncrease radii from ([0.33 1.28336349]) to ([0.363 1.41169984])\nIncrease radii from ([0.3 0.8163227]) to ([0.33 0.89795497])\nIncrease radii from ([0.33 0.89795497]) to ([0.363 0.98775047])\nIncrease radii from ([0.363 0.98775047]) to ([0.3993 1.08652552])\nIncrease radii from ([0.3993 1.08652552]) to ([0.43923 1.19517807])\nIncrease radii from ([0.43923 1.19517807]) to ([0.483153 1.31469588])\nIncrease radii from ([0.483153 1.31469588]) to ([0.5314683 1.44616547])\nIncrease radii from ([0.5314683 1.44616547]) to ([0.58461513 1.59078201])\nIncrease radii from ([0.58461513 1.59078201]) to ([0.64307664 1.74986021])\nIncrease radii from ([0.25 1.31636107]) to ([0.275 1.44799717])\nIncrease radii from ([0.275 1.44799717]) to ([0.3025 1.59279689])\nIncrease radii from ([0.3025 1.59279689]) to ([0.33275 1.75207658])\nIncrease radii from ([0.25 0.66140252]) to ([0.275 0.72754277])\nIncrease radii from ([0.275 0.72754277]) to ([0.3025 0.80029705])\nIncrease radii from ([0.3025 0.80029705]) to ([0.33275 0.88032675])\nIncrease radii from ([0.33275 0.88032675]) to ([0.366025 0.96835943])\nIncrease radii from ([0.366025 0.96835943]) to ([0.4026275 1.06519537])\nIncrease radii from ([0.4026275 1.06519537]) to ([0.44289025 1.17171491])\nIncrease radii from ([0.44289025 1.17171491]) to ([0.48717928 1.2888864 ])\nIncrease radii from ([0.48717928 1.2888864 ]) to ([0.5358972 1.41777504])\nIncrease radii from ([0.5358972 1.41777504]) to ([0.58948692 1.55955254])\nIncrease radii from ([0.58948692 1.55955254]) to ([0.64843562 1.7155078 ])\nIncrease radii from ([0.25 0.97224507]) to ([0.275 1.06946957])\nIncrease radii from ([0.275 1.06946957]) to ([0.3025 1.17641653])\nIncrease radii from ([0.3025 1.17641653]) to ([0.33275 1.29405818])\nIncrease radii from ([0.25 0.68026892]) to ([0.275 0.74829581])\nIncrease radii from ([0.275 0.74829581]) to ([0.3025 0.82312539])\nIncrease radii from ([0.3025 0.82312539]) to ([0.33275 0.90543793])\nIncrease radii from ([0.33275 0.90543793]) to ([0.366025 0.99598173])\nIncrease radii from ([0.366025 0.99598173]) to ([0.4026275 1.0955799])\nIncrease radii from ([0.4026275 1.0955799]) to ([0.44289025 1.20513789])\nIncrease radii from ([0.44289025 1.20513789]) to ([0.48717928 1.32565168])\nIncrease radii from ([0.48717928 1.32565168]) to ([0.5358972 1.45821685])\nIncrease radii from ([0.5358972 1.45821685]) to ([0.58948692 1.60403853])\nIncrease radii from ([0.58948692 1.60403853]) to ([0.64843562 1.76444238])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1703.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.58736605]) to ([0.66 1.74610265])\nIncrease radii from ([0.6 1.57613092]) to ([0.66 1.73374401])\nIncrease radii from ([0.55 1.45508554]) to ([0.605 1.6005941])\nIncrease radii from ([0.605 1.6005941]) to ([0.6655 1.76065351])\nIncrease radii from ([0.55 1.44478668]) to ([0.605 1.58926534])\nIncrease radii from ([0.605 1.58926534]) to ([0.6655 1.74819188])\nIncrease radii from ([0.55 1.49659162]) to ([0.605 1.64625079])\nIncrease radii from ([0.5 1.32280504]) to ([0.55 1.45508554])\nIncrease radii from ([0.55 1.45508554]) to ([0.605 1.6005941])\nIncrease radii from ([0.605 1.6005941]) to ([0.6655 1.76065351])\nIncrease radii from ([0.5 1.31344243]) to ([0.55 1.44478668])\nIncrease radii from ([0.55 1.44478668]) to ([0.605 1.58926534])\nIncrease radii from ([0.605 1.58926534]) to ([0.6655 1.74819188])\nIncrease radii from ([0.5 1.36053784]) to ([0.55 1.49659162])\nIncrease radii from ([0.55 1.49659162]) to ([0.605 1.64625079])\nIncrease radii from ([0.45 1.19052453]) to ([0.495 1.30957699])\nIncrease radii from ([0.495 1.30957699]) to ([0.5445 1.44053469])\nIncrease radii from ([0.5445 1.44053469]) to ([0.59895 1.58458816])\nIncrease radii from ([0.59895 1.58458816]) to ([0.658845 1.74304697])\nIncrease radii from ([0.45 1.18209819]) to ([0.495 1.30030801])\nIncrease radii from ([0.495 1.30030801]) to ([0.5445 1.43033881])\nIncrease radii from ([0.5445 1.43033881]) to ([0.59895 1.57337269])\nIncrease radii from ([0.59895 1.57337269]) to ([0.658845 1.73070996])\nIncrease radii from ([0.45 1.22448406]) to ([0.495 1.34693246])\nIncrease radii from ([0.495 1.34693246]) to ([0.5445 1.48162571])\nIncrease radii from ([0.5445 1.48162571]) to ([0.59895 1.62978828])\nIncrease radii from ([0.4 1.05824403]) to ([0.44 1.16406843])\nIncrease radii from ([0.44 1.16406843]) to ([0.484 1.28047528])\nIncrease radii from ([0.484 1.28047528]) to ([0.5324 1.4085228])\nIncrease radii from ([0.5324 1.4085228]) to ([0.58564 1.54937509])\nIncrease radii from ([0.58564 1.54937509]) to ([0.644204 1.70431259])\nIncrease radii from ([0.4 1.05075395]) to ([0.44 1.15582934])\nIncrease radii from ([0.44 1.15582934]) to ([0.484 1.27141227])\nIncrease radii from ([0.484 1.27141227]) to ([0.5324 1.3985535])\nIncrease radii from ([0.5324 1.3985535]) to ([0.58564 1.53840885])\nIncrease radii from ([0.58564 1.53840885]) to ([0.644204 1.69224974])\nIncrease radii from ([0.4 1.08843027]) to ([0.44 1.1972733])\nIncrease radii from ([0.44 1.1972733]) to ([0.484 1.31700063])\nIncrease radii from ([0.484 1.31700063]) to ([0.5324 1.44870069])\nIncrease radii from ([0.5324 1.44870069]) to ([0.58564 1.59357076])\nIncrease radii from ([0.58564 1.59357076]) to ([0.644204 1.75292784])\nIncrease radii from ([0.35 0.92596353]) to ([0.385 1.01855988])\nIncrease radii from ([0.385 1.01855988]) to ([0.4235 1.12041587])\nIncrease radii from ([0.4235 1.12041587]) to ([0.46585 1.23245745])\nIncrease radii from ([0.46585 1.23245745]) to ([0.512435 1.3557032])\nIncrease radii from ([0.512435 1.3557032]) to ([0.5636785 1.49127352])\nIncrease radii from ([0.5636785 1.49127352]) to ([0.62004635 1.64040087])\nIncrease radii from ([0.35 0.9194097]) to ([0.385 1.01135067])\nIncrease radii from ([0.385 1.01135067]) to ([0.4235 1.11248574])\nIncrease radii from ([0.4235 1.11248574]) to ([0.46585 1.22373431])\nIncrease radii from ([0.46585 1.22373431]) to ([0.512435 1.34610775])\nIncrease radii from ([0.512435 1.34610775]) to ([0.5636785 1.48071852])\nIncrease radii from ([0.5636785 1.48071852]) to ([0.62004635 1.62879037])\nIncrease radii from ([0.35 0.95237649]) to ([0.385 1.04761414])\nIncrease radii from ([0.385 1.04761414]) to ([0.4235 1.15237555])\nIncrease radii from ([0.4235 1.15237555]) to ([0.46585 1.26761311])\nIncrease radii from ([0.46585 1.26761311]) to ([0.512435 1.39437442])\nIncrease radii from ([0.512435 1.39437442]) to ([0.5636785 1.53381186])\nIncrease radii from ([0.5636785 1.53381186]) to ([0.62004635 1.68719304])\nIncrease radii from ([0.3 1.16903319]) to ([0.33 1.28593651])\nIncrease radii from ([0.33 1.28593651]) to ([0.363 1.41453016])\nIncrease radii from ([0.3 0.79368302]) to ([0.33 0.87305133])\nIncrease radii from ([0.33 0.87305133]) to ([0.363 0.96035646])\nIncrease radii from ([0.363 0.96035646]) to ([0.3993 1.0563921])\nIncrease radii from ([0.3993 1.0563921]) to ([0.43923 1.16203131])\nIncrease radii from ([0.43923 1.16203131]) to ([0.483153 1.27823445])\nIncrease radii from ([0.483153 1.27823445]) to ([0.5314683 1.40605789])\nIncrease radii from ([0.5314683 1.40605789]) to ([0.58461513 1.54666368])\nIncrease radii from ([0.58461513 1.54666368]) to ([0.64307664 1.70133005])\nIncrease radii from ([0.3 0.78806546]) to ([0.33 0.86687201])\nIncrease radii from ([0.33 0.86687201]) to ([0.363 0.95355921])\nIncrease radii from ([0.363 0.95355921]) to ([0.3993 1.04891513])\nIncrease radii from ([0.3993 1.04891513]) to ([0.43923 1.15380664])\nIncrease radii from ([0.43923 1.15380664]) to ([0.483153 1.2691873])\nIncrease radii from ([0.483153 1.2691873]) to ([0.5314683 1.39610603])\nIncrease radii from ([0.5314683 1.39610603]) to ([0.58461513 1.53571664])\nIncrease radii from ([0.58461513 1.53571664]) to ([0.64307664 1.6892883 ])\nIncrease radii from ([0.3 0.8163227]) to ([0.33 0.89795497])\nIncrease radii from ([0.33 0.89795497]) to ([0.363 0.98775047])\nIncrease radii from ([0.363 0.98775047]) to ([0.3993 1.08652552])\nIncrease radii from ([0.3993 1.08652552]) to ([0.43923 1.19517807])\nIncrease radii from ([0.43923 1.19517807]) to ([0.483153 1.31469588])\nIncrease radii from ([0.483153 1.31469588]) to ([0.5314683 1.44616547])\nIncrease radii from ([0.5314683 1.44616547]) to ([0.58461513 1.59078201])\nIncrease radii from ([0.58461513 1.59078201]) to ([0.64307664 1.74986021])\nIncrease radii from ([0.25 0.97419433]) to ([0.275 1.07161376])\nIncrease radii from ([0.275 1.07161376]) to ([0.3025 1.17877514])\nIncrease radii from ([0.3025 1.17877514]) to ([0.33275 1.29665265])\nIncrease radii from ([0.44337701 0.7679515 ]) to ([0.48771471 0.84474665])\nIncrease radii from ([0.25 0.66140252]) to ([0.275 0.72754277])\nIncrease radii from ([0.275 0.72754277]) to ([0.3025 0.80029705])\nIncrease radii from ([0.3025 0.80029705]) to ([0.33275 0.88032675])\nIncrease radii from ([0.33275 0.88032675]) to ([0.366025 0.96835943])\nIncrease radii from ([0.366025 0.96835943]) to ([0.4026275 1.06519537])\nIncrease radii from ([0.4026275 1.06519537]) to ([0.44289025 1.17171491])\nIncrease radii from ([0.44289025 1.17171491]) to ([0.48717928 1.2888864 ])\nIncrease radii from ([0.48717928 1.2888864 ]) to ([0.5358972 1.41777504])\nIncrease radii from ([0.5358972 1.41777504]) to ([0.58948692 1.55955254])\nIncrease radii from ([0.58948692 1.55955254]) to ([0.64843562 1.7155078 ])\nIncrease radii from ([0.25 0.92568421]) to ([0.275 1.01825263])\nIncrease radii from ([0.275 1.01825263]) to ([0.3025 1.12007789])\nIncrease radii from ([0.25 0.65672122]) to ([0.275 0.72239334])\nIncrease radii from ([0.275 0.72239334]) to ([0.3025 0.79463267])\nIncrease radii from ([0.3025 0.79463267]) to ([0.33275 0.87409594])\nIncrease radii from ([0.33275 0.87409594]) to ([0.366025 0.96150553])\nIncrease radii from ([0.366025 0.96150553]) to ([0.4026275 1.05765609])\nIncrease radii from ([0.4026275 1.05765609]) to ([0.44289025 1.16342169])\nIncrease radii from ([0.44289025 1.16342169]) to ([0.48717928 1.27976386])\nIncrease radii from ([0.48717928 1.27976386]) to ([0.5358972 1.40774025])\nIncrease radii from ([0.5358972 1.40774025]) to ([0.58948692 1.54851428])\nIncrease radii from ([0.58948692 1.54851428]) to ([0.64843562 1.7033657 ])\nIncrease radii from ([0.25 0.68026892]) to ([0.275 0.74829581])\nIncrease radii from ([0.275 0.74829581]) to ([0.3025 0.82312539])\nIncrease radii from ([0.3025 0.82312539]) to ([0.33275 0.90543793])\nIncrease radii from ([0.33275 0.90543793]) to ([0.366025 0.99598173])\nIncrease radii from ([0.366025 0.99598173]) to ([0.4026275 1.0955799])\nIncrease radii from ([0.4026275 1.0955799]) to ([0.44289025 1.20513789])\nIncrease radii from ([0.44289025 1.20513789]) to ([0.48717928 1.32565168])\nIncrease radii from ([0.48717928 1.32565168]) to ([0.5358972 1.45821685])\nIncrease radii from ([0.5358972 1.45821685]) to ([0.58948692 1.60403853])\nIncrease radii from ([0.58948692 1.60403853]) to ([0.64843562 1.76444238])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1704.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.52150754]) to ([0.605 1.67365829])\nIncrease radii from ([0.5 1.38318867]) to ([0.55 1.52150754])\nIncrease radii from ([0.55 1.52150754]) to ([0.605 1.67365829])\nIncrease radii from ([0.45 1.2448698]) to ([0.495 1.36935678])\nIncrease radii from ([0.495 1.36935678]) to ([0.5445 1.50629246])\nIncrease radii from ([0.5445 1.50629246]) to ([0.59895 1.65692171])\nIncrease radii from ([0.4 1.10655094]) to ([0.44 1.21720603])\nIncrease radii from ([0.44 1.21720603]) to ([0.484 1.33892663])\nIncrease radii from ([0.484 1.33892663]) to ([0.5324 1.4728193])\nIncrease radii from ([0.5324 1.4728193]) to ([0.58564 1.62010123])\nIncrease radii from ([0.35 0.96823207]) to ([0.385 1.06505528])\nIncrease radii from ([0.385 1.06505528]) to ([0.4235 1.1715608])\nIncrease radii from ([0.4235 1.1715608]) to ([0.46585 1.28871688])\nIncrease radii from ([0.46585 1.28871688]) to ([0.512435 1.41758857])\nIncrease radii from ([0.512435 1.41758857]) to ([0.5636785 1.55934743])\nIncrease radii from ([0.5636785 1.55934743]) to ([0.62004635 1.71528217])\nIncrease radii from ([0.3 0.8299132]) to ([0.33 0.91290452])\nIncrease radii from ([0.33 0.91290452]) to ([0.363 1.00419498])\nIncrease radii from ([0.363 1.00419498]) to ([0.3993 1.10461447])\nIncrease radii from ([0.3993 1.10461447]) to ([0.43923 1.21507592])\nIncrease radii from ([0.43923 1.21507592]) to ([0.483153 1.33658351])\nIncrease radii from ([0.483153 1.33658351]) to ([0.5314683 1.47024186])\nIncrease radii from ([0.5314683 1.47024186]) to ([0.58461513 1.61726605])\nIncrease radii from ([0.3 1.5813815]) to ([0.33 1.73951965])\nIncrease radii from ([0.25 0.69159434]) to ([0.275 0.76075377])\nIncrease radii from ([0.275 0.76075377]) to ([0.3025 0.83682915])\nIncrease radii from ([0.3025 0.83682915]) to ([0.33275 0.92051206])\nIncrease radii from ([0.33275 0.92051206]) to ([0.366025 1.01256327])\nIncrease radii from ([0.366025 1.01256327]) to ([0.4026275 1.11381959])\nIncrease radii from ([0.4026275 1.11381959]) to ([0.44289025 1.22520155])\nIncrease radii from ([0.44289025 1.22520155]) to ([0.48717928 1.34772171])\nIncrease radii from ([0.48717928 1.34772171]) to ([0.5358972 1.48249388])\nIncrease radii from ([0.5358972 1.48249388]) to ([0.58948692 1.63074327])\nIncrease radii from ([0.25 1.31781792]) to ([0.275 1.44959971])\nIncrease radii from ([0.275 1.44959971]) to ([0.3025 1.59455968])\nIncrease radii from ([0.3025 1.59455968]) to ([0.33275 1.75401565])\nIncrease radii from ([0.25 0.96523498]) to ([0.275 1.06175848])\nIncrease radii from ([0.275 1.06175848]) to ([0.3025 1.16793433])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1705.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.59810327]) to ([0.66 1.7579136])\nIncrease radii from ([0.6 1.57467193]) to ([0.66 1.73213912])\nIncrease radii from ([0.55 1.464928]) to ([0.605 1.6114208])\nIncrease radii from ([0.605 1.6114208]) to ([0.6655 1.77256288])\nIncrease radii from ([0.55 1.44344927]) to ([0.605 1.5877942])\nIncrease radii from ([0.605 1.5877942]) to ([0.6655 1.74657362])\nIncrease radii from ([0.5 1.33175273]) to ([0.55 1.464928])\nIncrease radii from ([0.55 1.464928]) to ([0.605 1.6114208])\nIncrease radii from ([0.605 1.6114208]) to ([0.6655 1.77256288])\nIncrease radii from ([0.5 1.31222661]) to ([0.55 1.44344927])\nIncrease radii from ([0.55 1.44344927]) to ([0.605 1.5877942])\nIncrease radii from ([0.605 1.5877942]) to ([0.6655 1.74657362])\nIncrease radii from ([0.45 1.19857745]) to ([0.495 1.3184352])\nIncrease radii from ([0.495 1.3184352]) to ([0.5445 1.45027872])\nIncrease radii from ([0.5445 1.45027872]) to ([0.59895 1.59530659])\nIncrease radii from ([0.59895 1.59530659]) to ([0.658845 1.75483725])\nIncrease radii from ([0.45 1.18100395]) to ([0.495 1.29910434])\nIncrease radii from ([0.495 1.29910434]) to ([0.5445 1.42901478])\nIncrease radii from ([0.5445 1.42901478]) to ([0.59895 1.57191626])\nIncrease radii from ([0.59895 1.57191626]) to ([0.658845 1.72910788])\nIncrease radii from ([0.4 1.06540218]) to ([0.44 1.1719424])\nIncrease radii from ([0.44 1.1719424]) to ([0.484 1.28913664])\nIncrease radii from ([0.484 1.28913664]) to ([0.5324 1.4180503])\nIncrease radii from ([0.5324 1.4180503]) to ([0.58564 1.55985533])\nIncrease radii from ([0.58564 1.55985533]) to ([0.644204 1.71584087])\nIncrease radii from ([0.4 1.04978129]) to ([0.44 1.15475942])\nIncrease radii from ([0.44 1.15475942]) to ([0.484 1.27023536])\nIncrease radii from ([0.484 1.27023536]) to ([0.5324 1.39725889])\nIncrease radii from ([0.5324 1.39725889]) to ([0.58564 1.53698478])\nIncrease radii from ([0.58564 1.53698478]) to ([0.644204 1.69068326])\nIncrease radii from ([0.35 0.93222691]) to ([0.385 1.0254496])\nIncrease radii from ([0.385 1.0254496]) to ([0.4235 1.12799456])\nIncrease radii from ([0.4235 1.12799456]) to ([0.46585 1.24079402])\nIncrease radii from ([0.46585 1.24079402]) to ([0.512435 1.36487342])\nIncrease radii from ([0.512435 1.36487342]) to ([0.5636785 1.50136076])\nIncrease radii from ([0.5636785 1.50136076]) to ([0.62004635 1.65149683])\nIncrease radii from ([0.35 0.91855863]) to ([0.385 1.01041449])\nIncrease radii from ([0.385 1.01041449]) to ([0.4235 1.11145594])\nIncrease radii from ([0.4235 1.11145594]) to ([0.46585 1.22260153])\nIncrease radii from ([0.46585 1.22260153]) to ([0.512435 1.34486168])\nIncrease radii from ([0.512435 1.34486168]) to ([0.5636785 1.47934785])\nIncrease radii from ([0.5636785 1.47934785]) to ([0.62004635 1.62728264])\nIncrease radii from ([0.3 1.5813815]) to ([0.33 1.73951965])\nIncrease radii from ([0.3 0.79905164]) to ([0.33 0.8789568])\nIncrease radii from ([0.33 0.8789568]) to ([0.363 0.96685248])\nIncrease radii from ([0.363 0.96685248]) to ([0.3993 1.06353773])\nIncrease radii from ([0.3993 1.06353773]) to ([0.43923 1.1698915])\nIncrease radii from ([0.43923 1.1698915]) to ([0.483153 1.28688065])\nIncrease radii from ([0.483153 1.28688065]) to ([0.5314683 1.41556872])\nIncrease radii from ([0.5314683 1.41556872]) to ([0.58461513 1.55712559])\nIncrease radii from ([0.58461513 1.55712559]) to ([0.64307664 1.71283815])\nIncrease radii from ([0.3 0.78733597]) to ([0.33 0.86606956])\nIncrease radii from ([0.33 0.86606956]) to ([0.363 0.95267652])\nIncrease radii from ([0.363 0.95267652]) to ([0.3993 1.04794417])\nIncrease radii from ([0.3993 1.04794417]) to ([0.43923 1.15273859])\nIncrease radii from ([0.43923 1.15273859]) to ([0.483153 1.26801245])\nIncrease radii from ([0.483153 1.26801245]) to ([0.5314683 1.39481369])\nIncrease radii from ([0.5314683 1.39481369]) to ([0.58461513 1.53429506])\nIncrease radii from ([0.58461513 1.53429506]) to ([0.64307664 1.68772457])\nIncrease radii from ([0.25 1.31781792]) to ([0.275 1.44959971])\nIncrease radii from ([0.275 1.44959971]) to ([0.3025 1.59455968])\nIncrease radii from ([0.3025 1.59455968]) to ([0.33275 1.75401565])\nIncrease radii from ([0.25 0.93200425]) to ([0.275 1.02520467])\nIncrease radii from ([0.275 1.02520467]) to ([0.3025 1.12772514])\nIncrease radii from ([0.25 0.66587636]) to ([0.275 0.732464])\nIncrease radii from ([0.275 0.732464]) to ([0.3025 0.8057104])\nIncrease radii from ([0.3025 0.8057104]) to ([0.33275 0.88628144])\nIncrease radii from ([0.33275 0.88628144]) to ([0.366025 0.97490958])\nIncrease radii from ([0.366025 0.97490958]) to ([0.4026275 1.07240054])\nIncrease radii from ([0.4026275 1.07240054]) to ([0.44289025 1.1796406 ])\nIncrease radii from ([0.44289025 1.1796406 ]) to ([0.48717928 1.29760466])\nIncrease radii from ([0.48717928 1.29760466]) to ([0.5358972 1.42736512])\nIncrease radii from ([0.5358972 1.42736512]) to ([0.58948692 1.57010163])\nIncrease radii from ([0.58948692 1.57010163]) to ([0.64843562 1.7271118 ])\nIncrease radii from ([0.25 0.6561133]) to ([0.275 0.72172464])\nIncrease radii from ([0.275 0.72172464]) to ([0.3025 0.7938971])\nIncrease radii from ([0.3025 0.7938971]) to ([0.33275 0.87328681])\nIncrease radii from ([0.33275 0.87328681]) to ([0.366025 0.96061549])\nIncrease radii from ([0.366025 0.96061549]) to ([0.4026275 1.05667704])\nIncrease radii from ([0.4026275 1.05667704]) to ([0.44289025 1.16234474])\nIncrease radii from ([0.44289025 1.16234474]) to ([0.48717928 1.27857922])\nIncrease radii from ([0.48717928 1.27857922]) to ([0.5358972 1.40643714])\nIncrease radii from ([0.5358972 1.40643714]) to ([0.58948692 1.54708085])\nIncrease radii from ([0.58948692 1.54708085]) to ([0.64843562 1.70178894])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1706.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.57549058]) to ([0.66 1.73303964])\nIncrease radii from ([0.6 1.57667126]) to ([0.66 1.73433839])\nIncrease radii from ([0.6 1.57467193]) to ([0.66 1.73213912])\nIncrease radii from ([0.6 1.60736606]) to ([0.66 1.76810266])\nIncrease radii from ([0.55 1.50749914]) to ([0.605 1.65824906])\nIncrease radii from ([0.55 1.4441997]) to ([0.605 1.58861967])\nIncrease radii from ([0.605 1.58861967]) to ([0.6655 1.74748163])\nIncrease radii from ([0.55 1.44528199]) to ([0.605 1.58981019])\nIncrease radii from ([0.605 1.58981019]) to ([0.6655 1.74879121])\nIncrease radii from ([0.55 1.44344927]) to ([0.605 1.5877942])\nIncrease radii from ([0.605 1.5877942]) to ([0.6655 1.74657362])\nIncrease radii from ([0.55 1.47341888]) to ([0.605 1.62076077])\nIncrease radii from ([0.5 1.37045377]) to ([0.55 1.50749914])\nIncrease radii from ([0.55 1.50749914]) to ([0.605 1.65824906])\nIncrease radii from ([0.5 1.31290882]) to ([0.55 1.4441997])\nIncrease radii from ([0.55 1.4441997]) to ([0.605 1.58861967])\nIncrease radii from ([0.605 1.58861967]) to ([0.6655 1.74748163])\nIncrease radii from ([0.5 1.31389272]) to ([0.55 1.44528199])\nIncrease radii from ([0.55 1.44528199]) to ([0.605 1.58981019])\nIncrease radii from ([0.605 1.58981019]) to ([0.6655 1.74879121])\nIncrease radii from ([0.5 1.31222661]) to ([0.55 1.44344927])\nIncrease radii from ([0.55 1.44344927]) to ([0.605 1.5877942])\nIncrease radii from ([0.605 1.5877942]) to ([0.6655 1.74657362])\nIncrease radii from ([0.5 1.33947171]) to ([0.55 1.47341888])\nIncrease radii from ([0.55 1.47341888]) to ([0.605 1.62076077])\nIncrease radii from ([0.45 1.23340839]) to ([0.495 1.35674923])\nIncrease radii from ([0.495 1.35674923]) to ([0.5445 1.49242415])\nIncrease radii from ([0.5445 1.49242415]) to ([0.59895 1.64166657])\nIncrease radii from ([0.45 1.18161793]) to ([0.495 1.29977973])\nIncrease radii from ([0.495 1.29977973]) to ([0.5445 1.4297577])\nIncrease radii from ([0.5445 1.4297577]) to ([0.59895 1.57273347])\nIncrease radii from ([0.59895 1.57273347]) to ([0.658845 1.73000682])\nIncrease radii from ([0.45 1.18250345]) to ([0.495 1.30075379])\nIncrease radii from ([0.495 1.30075379]) to ([0.5445 1.43082917])\nIncrease radii from ([0.5445 1.43082917]) to ([0.59895 1.57391209])\nIncrease radii from ([0.59895 1.57391209]) to ([0.658845 1.7313033])\nIncrease radii from ([0.45 1.18100395]) to ([0.495 1.29910434])\nIncrease radii from ([0.495 1.29910434]) to ([0.5445 1.42901478])\nIncrease radii from ([0.5445 1.42901478]) to ([0.59895 1.57191626])\nIncrease radii from ([0.59895 1.57191626]) to ([0.658845 1.72910788])\nIncrease radii from ([0.45 1.20552454]) to ([0.495 1.326077])\nIncrease radii from ([0.495 1.326077]) to ([0.5445 1.4586847])\nIncrease radii from ([0.5445 1.4586847]) to ([0.59895 1.60455316])\nIncrease radii from ([0.59895 1.60455316]) to ([0.658845 1.76500848])\nIncrease radii from ([0.4 1.09636301]) to ([0.44 1.20599931])\nIncrease radii from ([0.44 1.20599931]) to ([0.484 1.32659925])\nIncrease radii from ([0.484 1.32659925]) to ([0.5324 1.45925917])\nIncrease radii from ([0.5324 1.45925917]) to ([0.58564 1.60518509])\nIncrease radii from ([0.58564 1.60518509]) to ([0.644204 1.7657036])\nIncrease radii from ([0.4 1.05032705]) to ([0.44 1.15535976])\nIncrease radii from ([0.44 1.15535976]) to ([0.484 1.27089573])\nIncrease radii from ([0.484 1.27089573]) to ([0.5324 1.39798531])\nIncrease radii from ([0.5324 1.39798531]) to ([0.58564 1.53778384])\nIncrease radii from ([0.58564 1.53778384]) to ([0.644204 1.69156222])\nIncrease radii from ([0.4 1.05111418]) to ([0.44 1.15622559])\nIncrease radii from ([0.44 1.15622559]) to ([0.484 1.27184815])\nIncrease radii from ([0.484 1.27184815]) to ([0.5324 1.39903297])\nIncrease radii from ([0.5324 1.39903297]) to ([0.58564 1.53893626])\nIncrease radii from ([0.58564 1.53893626]) to ([0.644204 1.69282989])\nIncrease radii from ([0.4 1.04978129]) to ([0.44 1.15475942])\nIncrease radii from ([0.44 1.15475942]) to ([0.484 1.27023536])\nIncrease radii from ([0.484 1.27023536]) to ([0.5324 1.39725889])\nIncrease radii from ([0.5324 1.39725889]) to ([0.58564 1.53698478])\nIncrease radii from ([0.58564 1.53698478]) to ([0.644204 1.69068326])\nIncrease radii from ([0.4 1.07157737]) to ([0.44 1.17873511])\nIncrease radii from ([0.44 1.17873511]) to ([0.484 1.29660862])\nIncrease radii from ([0.484 1.29660862]) to ([0.5324 1.42626948])\nIncrease radii from ([0.5324 1.42626948]) to ([0.58564 1.56889643])\nIncrease radii from ([0.58564 1.56889643]) to ([0.644204 1.72578607])\nIncrease radii from ([0.35 0.95931764]) to ([0.385 1.0552494])\nIncrease radii from ([0.385 1.0552494]) to ([0.4235 1.16077434])\nIncrease radii from ([0.4235 1.16077434]) to ([0.46585 1.27685177])\nIncrease radii from ([0.46585 1.27685177]) to ([0.512435 1.40453695])\nIncrease radii from ([0.512435 1.40453695]) to ([0.5636785 1.54499065])\nIncrease radii from ([0.5636785 1.54499065]) to ([0.62004635 1.69948971])\nIncrease radii from ([0.35 0.91903617]) to ([0.385 1.01093979])\nIncrease radii from ([0.385 1.01093979]) to ([0.4235 1.11203377])\nIncrease radii from ([0.4235 1.11203377]) to ([0.46585 1.22323714])\nIncrease radii from ([0.46585 1.22323714]) to ([0.512435 1.34556086])\nIncrease radii from ([0.512435 1.34556086]) to ([0.5636785 1.48011694])\nIncrease radii from ([0.5636785 1.48011694]) to ([0.62004635 1.62812864])\nIncrease radii from ([0.35 0.9197249]) to ([0.385 1.01169739])\nIncrease radii from ([0.385 1.01169739]) to ([0.4235 1.11286713])\nIncrease radii from ([0.4235 1.11286713]) to ([0.46585 1.22415385])\nIncrease radii from ([0.46585 1.22415385]) to ([0.512435 1.34656923])\nIncrease radii from ([0.512435 1.34656923]) to ([0.5636785 1.48122615])\nIncrease radii from ([0.5636785 1.48122615]) to ([0.62004635 1.62934877])\nIncrease radii from ([0.35 0.91855863]) to ([0.385 1.01041449])\nIncrease radii from ([0.385 1.01041449]) to ([0.4235 1.11145594])\nIncrease radii from ([0.4235 1.11145594]) to ([0.46585 1.22260153])\nIncrease radii from ([0.46585 1.22260153]) to ([0.512435 1.34486168])\nIncrease radii from ([0.512435 1.34486168]) to ([0.5636785 1.47934785])\nIncrease radii from ([0.5636785 1.47934785]) to ([0.62004635 1.62728264])\nIncrease radii from ([0.35 0.9376302]) to ([0.385 1.03139322])\nIncrease radii from ([0.385 1.03139322]) to ([0.4235 1.13453254])\nIncrease radii from ([0.4235 1.13453254]) to ([0.46585 1.24798579])\nIncrease radii from ([0.46585 1.24798579]) to ([0.512435 1.37278437])\nIncrease radii from ([0.512435 1.37278437]) to ([0.5636785 1.51006281])\nIncrease radii from ([0.5636785 1.51006281]) to ([0.62004635 1.66106909])\nIncrease radii from ([0.3 0.82227226]) to ([0.33 0.90449949])\nIncrease radii from ([0.33 0.90449949]) to ([0.363 0.99494943])\nIncrease radii from ([0.363 0.99494943]) to ([0.3993 1.09444438])\nIncrease radii from ([0.3993 1.09444438]) to ([0.43923 1.20388882])\nIncrease radii from ([0.43923 1.20388882]) to ([0.483153 1.3242777])\nIncrease radii from ([0.483153 1.3242777]) to ([0.5314683 1.45670547])\nIncrease radii from ([0.5314683 1.45670547]) to ([0.58461513 1.60237601])\nIncrease radii from ([0.58461513 1.60237601]) to ([0.64307664 1.76261361])\nIncrease radii from ([0.3 1.53900657]) to ([0.33 1.69290723])\nIncrease radii from ([0.3 0.78774529]) to ([0.33 0.86651982])\nIncrease radii from ([0.33 0.86651982]) to ([0.363 0.9531718])\nIncrease radii from ([0.363 0.9531718]) to ([0.3993 1.04848898])\nIncrease radii from ([0.3993 1.04848898]) to ([0.43923 1.15333788])\nIncrease radii from ([0.43923 1.15333788]) to ([0.483153 1.26867167])\nIncrease radii from ([0.483153 1.26867167]) to ([0.5314683 1.39553883])\nIncrease radii from ([0.5314683 1.39553883]) to ([0.58461513 1.53509272])\nIncrease radii from ([0.58461513 1.53509272]) to ([0.64307664 1.68860199])\nIncrease radii from ([0.3 0.78833563]) to ([0.33 0.86716919])\nIncrease radii from ([0.33 0.86716919]) to ([0.363 0.95388611])\nIncrease radii from ([0.363 0.95388611]) to ([0.3993 1.04927473])\nIncrease radii from ([0.3993 1.04927473]) to ([0.43923 1.1542022])\nIncrease radii from ([0.43923 1.1542022]) to ([0.483153 1.26962242])\nIncrease radii from ([0.483153 1.26962242]) to ([0.5314683 1.39658466])\nIncrease radii from ([0.5314683 1.39658466]) to ([0.58461513 1.53624313])\nIncrease radii from ([0.58461513 1.53624313]) to ([0.64307664 1.68986744])\nIncrease radii from ([0.3 0.78733597]) to ([0.33 0.86606956])\nIncrease radii from ([0.33 0.86606956]) to ([0.363 0.95267652])\nIncrease radii from ([0.363 0.95267652]) to ([0.3993 1.04794417])\nIncrease radii from ([0.3993 1.04794417]) to ([0.43923 1.15273859])\nIncrease radii from ([0.43923 1.15273859]) to ([0.483153 1.26801245])\nIncrease radii from ([0.483153 1.26801245]) to ([0.5314683 1.39481369])\nIncrease radii from ([0.5314683 1.39481369]) to ([0.58461513 1.53429506])\nIncrease radii from ([0.58461513 1.53429506]) to ([0.64307664 1.68772457])\nIncrease radii from ([0.3 1.13075823]) to ([0.33 1.24383406])\nIncrease radii from ([0.33 1.24383406]) to ([0.363 1.36821746])\nIncrease radii from ([0.3 0.80368303]) to ([0.33 0.88405133])\nIncrease radii from ([0.33 0.88405133]) to ([0.363 0.97245646])\nIncrease radii from ([0.363 0.97245646]) to ([0.3993 1.06970211])\nIncrease radii from ([0.3993 1.06970211]) to ([0.43923 1.17667232])\nIncrease radii from ([0.43923 1.17667232]) to ([0.483153 1.29433955])\nIncrease radii from ([0.483153 1.29433955]) to ([0.5314683 1.42377351])\nIncrease radii from ([0.5314683 1.42377351]) to ([0.58461513 1.56615086])\nIncrease radii from ([0.58461513 1.56615086]) to ([0.64307664 1.72276594])\nIncrease radii from ([0.25 0.68522688]) to ([0.275 0.75374957])\nIncrease radii from ([0.275 0.75374957]) to ([0.3025 0.82912453])\nIncrease radii from ([0.3025 0.82912453]) to ([0.33275 0.91203698])\nIncrease radii from ([0.33275 0.91203698]) to ([0.366025 1.00324068])\nIncrease radii from ([0.366025 1.00324068]) to ([0.4026275 1.10356475])\nIncrease radii from ([0.4026275 1.10356475]) to ([0.44289025 1.21392122])\nIncrease radii from ([0.44289025 1.21392122]) to ([0.48717928 1.33531334])\nIncrease radii from ([0.48717928 1.33531334]) to ([0.5358972 1.46884468])\nIncrease radii from ([0.5358972 1.46884468]) to ([0.58948692 1.61572915])\nIncrease radii from ([0.44268307 0.76674958]) to ([0.48695138 0.84342453])\nIncrease radii from ([0.25 1.28250548]) to ([0.275 1.41075602])\nIncrease radii from ([0.275 1.41075602]) to ([0.3025 1.55183163])\nIncrease radii from ([0.3025 1.55183163]) to ([0.33275 1.70701479])\nIncrease radii from ([0.25 0.93200425]) to ([0.275 1.02520467])\nIncrease radii from ([0.275 1.02520467]) to ([0.3025 1.12772514])\nIncrease radii from ([0.25 0.65645441]) to ([0.275 0.72209985])\nIncrease radii from ([0.275 0.72209985]) to ([0.3025 0.79430983])\nIncrease radii from ([0.3025 0.79430983]) to ([0.33275 0.87374082])\nIncrease radii from ([0.33275 0.87374082]) to ([0.366025 0.9611149])\nIncrease radii from ([0.366025 0.9611149]) to ([0.4026275 1.05722639])\nIncrease radii from ([0.4026275 1.05722639]) to ([0.44289025 1.16294903])\nIncrease radii from ([0.44289025 1.16294903]) to ([0.48717928 1.27924393])\nIncrease radii from ([0.48717928 1.27924393]) to ([0.5358972 1.40716832])\nIncrease radii from ([0.5358972 1.40716832]) to ([0.58948692 1.54788516])\nIncrease radii from ([0.58948692 1.54788516]) to ([0.64843562 1.70267367])\nIncrease radii from ([0.25 0.65694636]) to ([0.275 0.722641])\nIncrease radii from ([0.275 0.722641]) to ([0.3025 0.7949051])\nIncrease radii from ([0.3025 0.7949051]) to ([0.33275 0.8743956])\nIncrease radii from ([0.33275 0.8743956]) to ([0.366025 0.96183517])\nIncrease radii from ([0.366025 0.96183517]) to ([0.4026275 1.05801868])\nIncrease radii from ([0.4026275 1.05801868]) to ([0.44289025 1.16382055])\nIncrease radii from ([0.44289025 1.16382055]) to ([0.48717928 1.2802026 ])\nIncrease radii from ([0.48717928 1.2802026 ]) to ([0.5358972 1.40822287])\nIncrease radii from ([0.5358972 1.40822287]) to ([0.58948692 1.54904515])\nIncrease radii from ([0.58948692 1.54904515]) to ([0.64843562 1.70394967])\nIncrease radii from ([0.25 0.6561133]) to ([0.275 0.72172464])\nIncrease radii from ([0.275 0.72172464]) to ([0.3025 0.7938971])\nIncrease radii from ([0.3025 0.7938971]) to ([0.33275 0.87328681])\nIncrease radii from ([0.33275 0.87328681]) to ([0.366025 0.96061549])\nIncrease radii from ([0.366025 0.96061549]) to ([0.4026275 1.05667704])\nIncrease radii from ([0.4026275 1.05667704]) to ([0.44289025 1.16234474])\nIncrease radii from ([0.44289025 1.16234474]) to ([0.48717928 1.27857922])\nIncrease radii from ([0.48717928 1.27857922]) to ([0.5358972 1.40643714])\nIncrease radii from ([0.5358972 1.40643714]) to ([0.58948692 1.54708085])\nIncrease radii from ([0.58948692 1.54708085]) to ([0.64843562 1.70178894])\nIncrease radii from ([0.25 0.94229853]) to ([0.275 1.03652838])\nIncrease radii from ([0.275 1.03652838]) to ([0.3025 1.14018122])\nIncrease radii from ([0.3025 1.14018122]) to ([0.33275 1.25419934])\nIncrease radii from ([0.33275 1.25419934]) to ([0.366025 1.37961928])\nIncrease radii from ([0.25 0.66973586]) to ([0.275 0.73670944])\nIncrease radii from ([0.275 0.73670944]) to ([0.3025 0.81038039])\nIncrease radii from ([0.3025 0.81038039]) to ([0.33275 0.89141842])\nIncrease radii from ([0.33275 0.89141842]) to ([0.366025 0.98056027])\nIncrease radii from ([0.366025 0.98056027]) to ([0.4026275 1.07861629])\nIncrease radii from ([0.4026275 1.07861629]) to ([0.44289025 1.18647792])\nIncrease radii from ([0.44289025 1.18647792]) to ([0.48717928 1.30512572])\nIncrease radii from ([0.48717928 1.30512572]) to ([0.5358972 1.43563829])\nIncrease radii from ([0.5358972 1.43563829]) to ([0.58948692 1.57920212])\nIncrease radii from ([0.58948692 1.57920212]) to ([0.64843562 1.73712233])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1707.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.57549058]) to ([0.66 1.73303964])\nIncrease radii from ([0.6 1.60736606]) to ([0.66 1.76810266])\nIncrease radii from ([0.55 1.48631845]) to ([0.605 1.63495029])\nIncrease radii from ([0.55 1.4441997]) to ([0.605 1.58861967])\nIncrease radii from ([0.605 1.58861967]) to ([0.6655 1.74748163])\nIncrease radii from ([0.55 1.47341888]) to ([0.605 1.62076077])\nIncrease radii from ([0.5 1.35119859]) to ([0.55 1.48631845])\nIncrease radii from ([0.55 1.48631845]) to ([0.605 1.63495029])\nIncrease radii from ([0.5 1.31290882]) to ([0.55 1.4441997])\nIncrease radii from ([0.55 1.4441997]) to ([0.605 1.58861967])\nIncrease radii from ([0.605 1.58861967]) to ([0.6655 1.74748163])\nIncrease radii from ([0.5 1.33947171]) to ([0.55 1.47341888])\nIncrease radii from ([0.55 1.47341888]) to ([0.605 1.62076077])\nIncrease radii from ([0.45 1.21607873]) to ([0.495 1.3376866])\nIncrease radii from ([0.495 1.3376866]) to ([0.5445 1.47145526])\nIncrease radii from ([0.5445 1.47145526]) to ([0.59895 1.61860079])\nIncrease radii from ([0.45 1.18161793]) to ([0.495 1.29977973])\nIncrease radii from ([0.495 1.29977973]) to ([0.5445 1.4297577])\nIncrease radii from ([0.5445 1.4297577]) to ([0.59895 1.57273347])\nIncrease radii from ([0.59895 1.57273347]) to ([0.658845 1.73000682])\nIncrease radii from ([0.45 1.20552454]) to ([0.495 1.326077])\nIncrease radii from ([0.495 1.326077]) to ([0.5445 1.4586847])\nIncrease radii from ([0.5445 1.4586847]) to ([0.59895 1.60455316])\nIncrease radii from ([0.59895 1.60455316]) to ([0.658845 1.76500848])\nIncrease radii from ([0.4 1.08095887]) to ([0.44 1.18905476])\nIncrease radii from ([0.44 1.18905476]) to ([0.484 1.30796023])\nIncrease radii from ([0.484 1.30796023]) to ([0.5324 1.43875626])\nIncrease radii from ([0.5324 1.43875626]) to ([0.58564 1.58263188])\nIncrease radii from ([0.58564 1.58263188]) to ([0.644204 1.74089507])\nIncrease radii from ([0.4 1.05032705]) to ([0.44 1.15535976])\nIncrease radii from ([0.44 1.15535976]) to ([0.484 1.27089573])\nIncrease radii from ([0.484 1.27089573]) to ([0.5324 1.39798531])\nIncrease radii from ([0.5324 1.39798531]) to ([0.58564 1.53778384])\nIncrease radii from ([0.58564 1.53778384]) to ([0.644204 1.69156222])\nIncrease radii from ([0.4 1.07157737]) to ([0.44 1.17873511])\nIncrease radii from ([0.44 1.17873511]) to ([0.484 1.29660862])\nIncrease radii from ([0.484 1.29660862]) to ([0.5324 1.42626948])\nIncrease radii from ([0.5324 1.42626948]) to ([0.58564 1.56889643])\nIncrease radii from ([0.58564 1.56889643]) to ([0.644204 1.72578607])\nIncrease radii from ([0.35 0.94583901]) to ([0.385 1.04042291])\nIncrease radii from ([0.385 1.04042291]) to ([0.4235 1.14446521])\nIncrease radii from ([0.4235 1.14446521]) to ([0.46585 1.25891173])\nIncrease radii from ([0.46585 1.25891173]) to ([0.512435 1.3848029])\nIncrease radii from ([0.512435 1.3848029]) to ([0.5636785 1.52328319])\nIncrease radii from ([0.5636785 1.52328319]) to ([0.62004635 1.67561151])\nIncrease radii from ([0.35 0.91903617]) to ([0.385 1.01093979])\nIncrease radii from ([0.385 1.01093979]) to ([0.4235 1.11203377])\nIncrease radii from ([0.4235 1.11203377]) to ([0.46585 1.22323714])\nIncrease radii from ([0.46585 1.22323714]) to ([0.512435 1.34556086])\nIncrease radii from ([0.512435 1.34556086]) to ([0.5636785 1.48011694])\nIncrease radii from ([0.5636785 1.48011694]) to ([0.62004635 1.62812864])\nIncrease radii from ([0.35 0.9376302]) to ([0.385 1.03139322])\nIncrease radii from ([0.385 1.03139322]) to ([0.4235 1.13453254])\nIncrease radii from ([0.4235 1.13453254]) to ([0.46585 1.24798579])\nIncrease radii from ([0.46585 1.24798579]) to ([0.512435 1.37278437])\nIncrease radii from ([0.512435 1.37278437]) to ([0.5636785 1.51006281])\nIncrease radii from ([0.5636785 1.51006281]) to ([0.62004635 1.66106909])\nIncrease radii from ([0.3 0.81071915]) to ([0.33 0.89179107])\nIncrease radii from ([0.33 0.89179107]) to ([0.363 0.98097018])\nIncrease radii from ([0.363 0.98097018]) to ([0.3993 1.07906719])\nIncrease radii from ([0.3993 1.07906719]) to ([0.43923 1.18697391])\nIncrease radii from ([0.43923 1.18697391]) to ([0.483153 1.3056713])\nIncrease radii from ([0.483153 1.3056713]) to ([0.5314683 1.43623843])\nIncrease radii from ([0.5314683 1.43623843]) to ([0.58461513 1.57986228])\nIncrease radii from ([0.58461513 1.57986228]) to ([0.64307664 1.73784851])\nIncrease radii from ([0.3 1.59915102]) to ([0.33 1.75906613])\nIncrease radii from ([0.3 0.78774529]) to ([0.33 0.86651982])\nIncrease radii from ([0.33 0.86651982]) to ([0.363 0.9531718])\nIncrease radii from ([0.363 0.9531718]) to ([0.3993 1.04848898])\nIncrease radii from ([0.3993 1.04848898]) to ([0.43923 1.15333788])\nIncrease radii from ([0.43923 1.15333788]) to ([0.483153 1.26867167])\nIncrease radii from ([0.483153 1.26867167]) to ([0.5314683 1.39553883])\nIncrease radii from ([0.5314683 1.39553883]) to ([0.58461513 1.53509272])\nIncrease radii from ([0.58461513 1.53509272]) to ([0.64307664 1.68860199])\nIncrease radii from ([0.3 1.49840597]) to ([0.33 1.64824657])\nIncrease radii from ([0.3 0.80368303]) to ([0.33 0.88405133])\nIncrease radii from ([0.33 0.88405133]) to ([0.363 0.97245646])\nIncrease radii from ([0.363 0.97245646]) to ([0.3993 1.06970211])\nIncrease radii from ([0.3993 1.06970211]) to ([0.43923 1.17667232])\nIncrease radii from ([0.43923 1.17667232]) to ([0.483153 1.29433955])\nIncrease radii from ([0.483153 1.29433955]) to ([0.5314683 1.42377351])\nIncrease radii from ([0.5314683 1.42377351]) to ([0.58461513 1.56615086])\nIncrease radii from ([0.58461513 1.56615086]) to ([0.64307664 1.72276594])\nIncrease radii from ([0.25 0.67559929]) to ([0.275 0.74315922])\nIncrease radii from ([0.275 0.74315922]) to ([0.3025 0.81747515])\nIncrease radii from ([0.3025 0.81747515]) to ([0.33275 0.89922266])\nIncrease radii from ([0.33275 0.89922266]) to ([0.366025 0.98914493])\nIncrease radii from ([0.366025 0.98914493]) to ([0.4026275 1.08805942])\nIncrease radii from ([0.4026275 1.08805942]) to ([0.44289025 1.19686536])\nIncrease radii from ([0.44289025 1.19686536]) to ([0.48717928 1.3165519 ])\nIncrease radii from ([0.48717928 1.3165519 ]) to ([0.5358972 1.44820709])\nIncrease radii from ([0.5358972 1.44820709]) to ([0.58948692 1.5930278 ])\nIncrease radii from ([0.58948692 1.5930278 ]) to ([0.64843562 1.75233058])\nIncrease radii from ([0.25 1.33262585]) to ([0.275 1.46588844])\nIncrease radii from ([0.275 1.46588844]) to ([0.3025 1.61247728])\nIncrease radii from ([0.3025 1.61247728]) to ([0.33275 1.77372501])\nIncrease radii from ([0.25 0.93200425]) to ([0.275 1.02520467])\nIncrease radii from ([0.275 1.02520467]) to ([0.3025 1.12772514])\nIncrease radii from ([0.25 0.65645441]) to ([0.275 0.72209985])\nIncrease radii from ([0.275 0.72209985]) to ([0.3025 0.79430983])\nIncrease radii from ([0.3025 0.79430983]) to ([0.33275 0.87374082])\nIncrease radii from ([0.33275 0.87374082]) to ([0.366025 0.9611149])\nIncrease radii from ([0.366025 0.9611149]) to ([0.4026275 1.05722639])\nIncrease radii from ([0.4026275 1.05722639]) to ([0.44289025 1.16294903])\nIncrease radii from ([0.44289025 1.16294903]) to ([0.48717928 1.27924393])\nIncrease radii from ([0.48717928 1.27924393]) to ([0.5358972 1.40716832])\nIncrease radii from ([0.5358972 1.40716832]) to ([0.58948692 1.54788516])\nIncrease radii from ([0.58948692 1.54788516]) to ([0.64843562 1.70267367])\nIncrease radii from ([0.25 1.24867165]) to ([0.275 1.37353881])\nIncrease radii from ([0.275 1.37353881]) to ([0.3025 1.51089269])\nIncrease radii from ([0.3025 1.51089269]) to ([0.33275 1.66198196])\nIncrease radii from ([0.25 0.66973586]) to ([0.275 0.73670944])\nIncrease radii from ([0.275 0.73670944]) to ([0.3025 0.81038039])\nIncrease radii from ([0.3025 0.81038039]) to ([0.33275 0.89141842])\nIncrease radii from ([0.33275 0.89141842]) to ([0.366025 0.98056027])\nIncrease radii from ([0.366025 0.98056027]) to ([0.4026275 1.07861629])\nIncrease radii from ([0.4026275 1.07861629]) to ([0.44289025 1.18647792])\nIncrease radii from ([0.44289025 1.18647792]) to ([0.48717928 1.30512572])\nIncrease radii from ([0.48717928 1.30512572]) to ([0.5358972 1.43563829])\nIncrease radii from ([0.5358972 1.43563829]) to ([0.58948692 1.57920212])\nIncrease radii from ([0.58948692 1.57920212]) to ([0.64843562 1.73712233])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1708.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.3 1.60351614]) to ([0.33 1.76386776])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.25 1.33626345]) to ([0.275 1.4698898])\nIncrease radii from ([0.275 1.4698898]) to ([0.3025 1.61687878])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.92975661]) to ([0.275 1.02273227])\nIncrease radii from ([0.275 1.02273227]) to ([0.3025 1.1250055])\nIncrease radii from ([0.44410797 0.76921758]) to ([0.48851877 0.84613933])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1709.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.49346062]) to ([0.605 1.64280668])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.35769147]) to ([0.55 1.49346062])\nIncrease radii from ([0.55 1.49346062]) to ([0.605 1.64280668])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.45 1.22192232]) to ([0.495 1.34411455])\nIncrease radii from ([0.495 1.34411455]) to ([0.5445 1.47852601])\nIncrease radii from ([0.5445 1.47852601]) to ([0.59895 1.62637861])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.4 1.08615318]) to ([0.44 1.19476849])\nIncrease radii from ([0.44 1.19476849]) to ([0.484 1.31424534])\nIncrease radii from ([0.484 1.31424534]) to ([0.5324 1.44566988])\nIncrease radii from ([0.5324 1.44566988]) to ([0.58564 1.59023686])\nIncrease radii from ([0.58564 1.59023686]) to ([0.644204 1.74926055])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.35 0.95038403]) to ([0.385 1.04542243])\nIncrease radii from ([0.385 1.04542243]) to ([0.4235 1.14996467])\nIncrease radii from ([0.4235 1.14996467]) to ([0.46585 1.26496114])\nIncrease radii from ([0.46585 1.26496114]) to ([0.512435 1.39145726])\nIncrease radii from ([0.512435 1.39145726]) to ([0.5636785 1.53060298])\nIncrease radii from ([0.5636785 1.53060298]) to ([0.62004635 1.68366328])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.3 0.81461488]) to ([0.33 0.89607637])\nIncrease radii from ([0.33 0.89607637]) to ([0.363 0.98568401])\nIncrease radii from ([0.363 0.98568401]) to ([0.3993 1.08425241])\nIncrease radii from ([0.3993 1.08425241]) to ([0.43923 1.19267765])\nIncrease radii from ([0.43923 1.19267765]) to ([0.483153 1.31194541])\nIncrease radii from ([0.483153 1.31194541]) to ([0.5314683 1.44313995])\nIncrease radii from ([0.5314683 1.44313995]) to ([0.58461513 1.58745395])\nIncrease radii from ([0.58461513 1.58745395]) to ([0.64307664 1.74619934])\nIncrease radii from ([0.3 1.60351614]) to ([0.33 1.76386776])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.25 0.67884573]) to ([0.275 0.74673031])\nIncrease radii from ([0.275 0.74673031]) to ([0.3025 0.82140334])\nIncrease radii from ([0.3025 0.82140334]) to ([0.33275 0.90354367])\nIncrease radii from ([0.33275 0.90354367]) to ([0.366025 0.99389804])\nIncrease radii from ([0.366025 0.99389804]) to ([0.4026275 1.09328784])\nIncrease radii from ([0.4026275 1.09328784]) to ([0.44289025 1.20261663])\nIncrease radii from ([0.44289025 1.20261663]) to ([0.48717928 1.32287829])\nIncrease radii from ([0.48717928 1.32287829]) to ([0.5358972 1.45516612])\nIncrease radii from ([0.5358972 1.45516612]) to ([0.58948692 1.60068273])\nIncrease radii from ([0.58948692 1.60068273]) to ([0.64843562 1.76075101])\nIncrease radii from ([0.25 1.33626345]) to ([0.275 1.4698898])\nIncrease radii from ([0.275 1.4698898]) to ([0.3025 1.61687878])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.92975661]) to ([0.275 1.02273227])\nIncrease radii from ([0.275 1.02273227]) to ([0.3025 1.1250055])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1710.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.56744827]) to ([0.66 1.7241931])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.43682758]) to ([0.605 1.58051034])\nIncrease radii from ([0.605 1.58051034]) to ([0.6655 1.73856137])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.30620689]) to ([0.55 1.43682758])\nIncrease radii from ([0.55 1.43682758]) to ([0.605 1.58051034])\nIncrease radii from ([0.605 1.58051034]) to ([0.6655 1.73856137])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.1755862]) to ([0.495 1.29314482])\nIncrease radii from ([0.495 1.29314482]) to ([0.5445 1.42245931])\nIncrease radii from ([0.5445 1.42245931]) to ([0.59895 1.56470524])\nIncrease radii from ([0.59895 1.56470524]) to ([0.658845 1.72117576])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.04496551]) to ([0.44 1.14946206])\nIncrease radii from ([0.44 1.14946206]) to ([0.484 1.26440827])\nIncrease radii from ([0.484 1.26440827]) to ([0.5324 1.3908491])\nIncrease radii from ([0.5324 1.3908491]) to ([0.58564 1.52993401])\nIncrease radii from ([0.58564 1.52993401]) to ([0.644204 1.68292741])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.91434482]) to ([0.385 1.00577931])\nIncrease radii from ([0.385 1.00577931]) to ([0.4235 1.10635724])\nIncrease radii from ([0.4235 1.10635724]) to ([0.46585 1.21699296])\nIncrease radii from ([0.46585 1.21699296]) to ([0.512435 1.33869226])\nIncrease radii from ([0.512435 1.33869226]) to ([0.5636785 1.47256148])\nIncrease radii from ([0.5636785 1.47256148]) to ([0.62004635 1.61981763])\nIncrease radii from ([0.3 1.60351614]) to ([0.33 1.76386776])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.78372414]) to ([0.33 0.86209655])\nIncrease radii from ([0.33 0.86209655]) to ([0.363 0.9483062])\nIncrease radii from ([0.363 0.9483062]) to ([0.3993 1.04313682])\nIncrease radii from ([0.3993 1.04313682]) to ([0.43923 1.14745051])\nIncrease radii from ([0.43923 1.14745051]) to ([0.483153 1.26219556])\nIncrease radii from ([0.483153 1.26219556]) to ([0.5314683 1.38841511])\nIncrease radii from ([0.5314683 1.38841511]) to ([0.58461513 1.52725662])\nIncrease radii from ([0.58461513 1.52725662]) to ([0.64307664 1.67998229])\nIncrease radii from ([0.25 1.33626345]) to ([0.275 1.4698898])\nIncrease radii from ([0.275 1.4698898]) to ([0.3025 1.61687878])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.92975661]) to ([0.275 1.02273227])\nIncrease radii from ([0.275 1.02273227]) to ([0.3025 1.1250055])\nIncrease radii from ([0.25 0.65310345]) to ([0.275 0.71841379])\nIncrease radii from ([0.275 0.71841379]) to ([0.3025 0.79025517])\nIncrease radii from ([0.3025 0.79025517]) to ([0.33275 0.86928069])\nIncrease radii from ([0.33275 0.86928069]) to ([0.366025 0.95620876])\nIncrease radii from ([0.366025 0.95620876]) to ([0.4026275 1.05182963])\nIncrease radii from ([0.4026275 1.05182963]) to ([0.44289025 1.15701259])\nIncrease radii from ([0.44289025 1.15701259]) to ([0.48717928 1.27271385])\nIncrease radii from ([0.48717928 1.27271385]) to ([0.5358972 1.39998524])\nIncrease radii from ([0.5358972 1.39998524]) to ([0.58948692 1.53998376])\nIncrease radii from ([0.58948692 1.53998376]) to ([0.64843562 1.69398214])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1711.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.56744827]) to ([0.66 1.7241931])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.43682758]) to ([0.605 1.58051034])\nIncrease radii from ([0.605 1.58051034]) to ([0.6655 1.73856137])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.30620689]) to ([0.55 1.43682758])\nIncrease radii from ([0.55 1.43682758]) to ([0.605 1.58051034])\nIncrease radii from ([0.605 1.58051034]) to ([0.6655 1.73856137])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.1755862]) to ([0.495 1.29314482])\nIncrease radii from ([0.495 1.29314482]) to ([0.5445 1.42245931])\nIncrease radii from ([0.5445 1.42245931]) to ([0.59895 1.56470524])\nIncrease radii from ([0.59895 1.56470524]) to ([0.658845 1.72117576])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.04496551]) to ([0.44 1.14946206])\nIncrease radii from ([0.44 1.14946206]) to ([0.484 1.26440827])\nIncrease radii from ([0.484 1.26440827]) to ([0.5324 1.3908491])\nIncrease radii from ([0.5324 1.3908491]) to ([0.58564 1.52993401])\nIncrease radii from ([0.58564 1.52993401]) to ([0.644204 1.68292741])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.91434482]) to ([0.385 1.00577931])\nIncrease radii from ([0.385 1.00577931]) to ([0.4235 1.10635724])\nIncrease radii from ([0.4235 1.10635724]) to ([0.46585 1.21699296])\nIncrease radii from ([0.46585 1.21699296]) to ([0.512435 1.33869226])\nIncrease radii from ([0.512435 1.33869226]) to ([0.5636785 1.47256148])\nIncrease radii from ([0.5636785 1.47256148]) to ([0.62004635 1.61981763])\nIncrease radii from ([0.3 1.60351614]) to ([0.33 1.76386776])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.78372414]) to ([0.33 0.86209655])\nIncrease radii from ([0.33 0.86209655]) to ([0.363 0.9483062])\nIncrease radii from ([0.363 0.9483062]) to ([0.3993 1.04313682])\nIncrease radii from ([0.3993 1.04313682]) to ([0.43923 1.14745051])\nIncrease radii from ([0.43923 1.14745051]) to ([0.483153 1.26219556])\nIncrease radii from ([0.483153 1.26219556]) to ([0.5314683 1.38841511])\nIncrease radii from ([0.5314683 1.38841511]) to ([0.58461513 1.52725662])\nIncrease radii from ([0.58461513 1.52725662]) to ([0.64307664 1.67998229])\nIncrease radii from ([0.25 1.33626345]) to ([0.275 1.4698898])\nIncrease radii from ([0.275 1.4698898]) to ([0.3025 1.61687878])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.9307286]) to ([0.275 1.02380146])\nIncrease radii from ([0.275 1.02380146]) to ([0.3025 1.12618161])\nIncrease radii from ([0.25 0.65310345]) to ([0.275 0.71841379])\nIncrease radii from ([0.275 0.71841379]) to ([0.3025 0.79025517])\nIncrease radii from ([0.3025 0.79025517]) to ([0.33275 0.86928069])\nIncrease radii from ([0.33275 0.86928069]) to ([0.366025 0.95620876])\nIncrease radii from ([0.366025 0.95620876]) to ([0.4026275 1.05182963])\nIncrease radii from ([0.4026275 1.05182963]) to ([0.44289025 1.15701259])\nIncrease radii from ([0.44289025 1.15701259]) to ([0.48717928 1.27271385])\nIncrease radii from ([0.48717928 1.27271385]) to ([0.5358972 1.39998524])\nIncrease radii from ([0.5358972 1.39998524]) to ([0.58948692 1.53998376])\nIncrease radii from ([0.58948692 1.53998376]) to ([0.64843562 1.69398214])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1712.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.56744827]) to ([0.66 1.7241931])\nIncrease radii from ([0.55 1.478279]) to ([0.605 1.6261069])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.50231752]) to ([0.605 1.65254927])\nIncrease radii from ([0.55 1.43682758]) to ([0.605 1.58051034])\nIncrease radii from ([0.605 1.58051034]) to ([0.6655 1.73856137])\nIncrease radii from ([0.5 1.34389]) to ([0.55 1.478279])\nIncrease radii from ([0.55 1.478279]) to ([0.605 1.6261069])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.3657432]) to ([0.55 1.50231752])\nIncrease radii from ([0.55 1.50231752]) to ([0.605 1.65254927])\nIncrease radii from ([0.5 1.30620689]) to ([0.55 1.43682758])\nIncrease radii from ([0.55 1.43682758]) to ([0.605 1.58051034])\nIncrease radii from ([0.605 1.58051034]) to ([0.6655 1.73856137])\nIncrease radii from ([0.45 1.209501]) to ([0.495 1.3304511])\nIncrease radii from ([0.495 1.3304511]) to ([0.5445 1.46349621])\nIncrease radii from ([0.5445 1.46349621]) to ([0.59895 1.60984584])\nIncrease radii from ([0.59895 1.60984584]) to ([0.658845 1.77083042])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.22916888]) to ([0.495 1.35208576])\nIncrease radii from ([0.495 1.35208576]) to ([0.5445 1.48729434])\nIncrease radii from ([0.5445 1.48729434]) to ([0.59895 1.63602378])\nIncrease radii from ([0.45 1.1755862]) to ([0.495 1.29314482])\nIncrease radii from ([0.495 1.29314482]) to ([0.5445 1.42245931])\nIncrease radii from ([0.5445 1.42245931]) to ([0.59895 1.56470524])\nIncrease radii from ([0.59895 1.56470524]) to ([0.658845 1.72117576])\nIncrease radii from ([0.4 1.075112]) to ([0.44 1.1826232])\nIncrease radii from ([0.44 1.1826232]) to ([0.484 1.30088552])\nIncrease radii from ([0.484 1.30088552]) to ([0.5324 1.43097408])\nIncrease radii from ([0.5324 1.43097408]) to ([0.58564 1.57407148])\nIncrease radii from ([0.58564 1.57407148]) to ([0.644204 1.73147863])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.09259456]) to ([0.44 1.20185401])\nIncrease radii from ([0.44 1.20185401]) to ([0.484 1.32203941])\nIncrease radii from ([0.484 1.32203941]) to ([0.5324 1.45424336])\nIncrease radii from ([0.5324 1.45424336]) to ([0.58564 1.59966769])\nIncrease radii from ([0.58564 1.59966769]) to ([0.644204 1.75963446])\nIncrease radii from ([0.4 1.04496551]) to ([0.44 1.14946206])\nIncrease radii from ([0.44 1.14946206]) to ([0.484 1.26440827])\nIncrease radii from ([0.484 1.26440827]) to ([0.5324 1.3908491])\nIncrease radii from ([0.5324 1.3908491]) to ([0.58564 1.52993401])\nIncrease radii from ([0.58564 1.52993401]) to ([0.644204 1.68292741])\nIncrease radii from ([0.35 0.940723]) to ([0.385 1.0347953])\nIncrease radii from ([0.385 1.0347953]) to ([0.4235 1.13827483])\nIncrease radii from ([0.4235 1.13827483]) to ([0.46585 1.25210232])\nIncrease radii from ([0.46585 1.25210232]) to ([0.512435 1.37731255])\nIncrease radii from ([0.512435 1.37731255]) to ([0.5636785 1.5150438])\nIncrease radii from ([0.5636785 1.5150438]) to ([0.62004635 1.66654818])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.95602024]) to ([0.385 1.05162226])\nIncrease radii from ([0.385 1.05162226]) to ([0.4235 1.15678449])\nIncrease radii from ([0.4235 1.15678449]) to ([0.46585 1.27246294])\nIncrease radii from ([0.46585 1.27246294]) to ([0.512435 1.39970923])\nIncrease radii from ([0.512435 1.39970923]) to ([0.5636785 1.53968015])\nIncrease radii from ([0.5636785 1.53968015]) to ([0.62004635 1.69364817])\nIncrease radii from ([0.35 0.91434482]) to ([0.385 1.00577931])\nIncrease radii from ([0.385 1.00577931]) to ([0.4235 1.10635724])\nIncrease radii from ([0.4235 1.10635724]) to ([0.46585 1.21699296])\nIncrease radii from ([0.46585 1.21699296]) to ([0.512435 1.33869226])\nIncrease radii from ([0.512435 1.33869226]) to ([0.5636785 1.47256148])\nIncrease radii from ([0.5636785 1.47256148]) to ([0.62004635 1.61981763])\nIncrease radii from ([0.3 0.806334]) to ([0.33 0.8869674])\nIncrease radii from ([0.33 0.8869674]) to ([0.363 0.97566414])\nIncrease radii from ([0.363 0.97566414]) to ([0.3993 1.07323056])\nIncrease radii from ([0.3993 1.07323056]) to ([0.43923 1.18055361])\nIncrease radii from ([0.43923 1.18055361]) to ([0.483153 1.29860897])\nIncrease radii from ([0.483153 1.29860897]) to ([0.5314683 1.42846987])\nIncrease radii from ([0.5314683 1.42846987]) to ([0.58461513 1.57131686])\nIncrease radii from ([0.58461513 1.57131686]) to ([0.64307664 1.72844854])\nIncrease radii from ([0.3 1.60351614]) to ([0.33 1.76386776])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.81944592]) to ([0.33 0.90139051])\nIncrease radii from ([0.33 0.90139051]) to ([0.363 0.99152956])\nIncrease radii from ([0.363 0.99152956]) to ([0.3993 1.09068252])\nIncrease radii from ([0.3993 1.09068252]) to ([0.43923 1.19975077])\nIncrease radii from ([0.43923 1.19975077]) to ([0.483153 1.31972585])\nIncrease radii from ([0.483153 1.31972585]) to ([0.5314683 1.45169843])\nIncrease radii from ([0.5314683 1.45169843]) to ([0.58461513 1.59686827])\nIncrease radii from ([0.58461513 1.59686827]) to ([0.64307664 1.7565551 ])\nIncrease radii from ([0.3 0.78372414]) to ([0.33 0.86209655])\nIncrease radii from ([0.33 0.86209655]) to ([0.363 0.9483062])\nIncrease radii from ([0.363 0.9483062]) to ([0.3993 1.04313682])\nIncrease radii from ([0.3993 1.04313682]) to ([0.43923 1.14745051])\nIncrease radii from ([0.43923 1.14745051]) to ([0.483153 1.26219556])\nIncrease radii from ([0.483153 1.26219556]) to ([0.5314683 1.38841511])\nIncrease radii from ([0.5314683 1.38841511]) to ([0.58461513 1.52725662])\nIncrease radii from ([0.58461513 1.52725662]) to ([0.64307664 1.67998229])\nIncrease radii from ([0.25 0.671945]) to ([0.275 0.7391395])\nIncrease radii from ([0.275 0.7391395]) to ([0.3025 0.81305345])\nIncrease radii from ([0.3025 0.81305345]) to ([0.33275 0.8943588])\nIncrease radii from ([0.33275 0.8943588]) to ([0.366025 0.98379468])\nIncrease radii from ([0.366025 0.98379468]) to ([0.4026275 1.08217415])\nIncrease radii from ([0.4026275 1.08217415]) to ([0.44289025 1.19039156])\nIncrease radii from ([0.44289025 1.19039156]) to ([0.48717928 1.30943072])\nIncrease radii from ([0.48717928 1.30943072]) to ([0.5358972 1.44037379])\nIncrease radii from ([0.5358972 1.44037379]) to ([0.58948692 1.58441117])\nIncrease radii from ([0.58948692 1.58441117]) to ([0.64843562 1.74285228])\nIncrease radii from ([0.25 1.33626345]) to ([0.275 1.4698898])\nIncrease radii from ([0.275 1.4698898]) to ([0.3025 1.61687878])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.6828716]) to ([0.275 0.75115876])\nIncrease radii from ([0.275 0.75115876]) to ([0.3025 0.82627463])\nIncrease radii from ([0.3025 0.82627463]) to ([0.33275 0.9089021])\nIncrease radii from ([0.33275 0.9089021]) to ([0.366025 0.99979231])\nIncrease radii from ([0.366025 0.99979231]) to ([0.4026275 1.09977154])\nIncrease radii from ([0.4026275 1.09977154]) to ([0.44289025 1.20974869])\nIncrease radii from ([0.44289025 1.20974869]) to ([0.48717928 1.33072356])\nIncrease radii from ([0.48717928 1.33072356]) to ([0.5358972 1.46379592])\nIncrease radii from ([0.5358972 1.46379592]) to ([0.58948692 1.61017551])\nIncrease radii from ([0.58948692 1.61017551]) to ([0.64843562 1.77119306])\nIncrease radii from ([0.4437876 0.76866267]) to ([0.48816636 0.84552894])\nIncrease radii from ([0.25 0.65310345]) to ([0.275 0.71841379])\nIncrease radii from ([0.275 0.71841379]) to ([0.3025 0.79025517])\nIncrease radii from ([0.3025 0.79025517]) to ([0.33275 0.86928069])\nIncrease radii from ([0.33275 0.86928069]) to ([0.366025 0.95620876])\nIncrease radii from ([0.366025 0.95620876]) to ([0.4026275 1.05182963])\nIncrease radii from ([0.4026275 1.05182963]) to ([0.44289025 1.15701259])\nIncrease radii from ([0.44289025 1.15701259]) to ([0.48717928 1.27271385])\nIncrease radii from ([0.48717928 1.27271385]) to ([0.5358972 1.39998524])\nIncrease radii from ([0.5358972 1.39998524]) to ([0.58948692 1.53998376])\nIncrease radii from ([0.58948692 1.53998376]) to ([0.64843562 1.69398214])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1713.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.57797962]) to ([0.66 1.73577759])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.50185815]) to ([0.605 1.65204396])\nIncrease radii from ([0.55 1.44648132]) to ([0.605 1.59112945])\nIncrease radii from ([0.605 1.59112945]) to ([0.6655 1.7502424])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.36532559]) to ([0.55 1.50185815])\nIncrease radii from ([0.55 1.50185815]) to ([0.605 1.65204396])\nIncrease radii from ([0.5 1.31498302]) to ([0.55 1.44648132])\nIncrease radii from ([0.55 1.44648132]) to ([0.605 1.59112945])\nIncrease radii from ([0.605 1.59112945]) to ([0.6655 1.7502424])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.22879303]) to ([0.495 1.35167233])\nIncrease radii from ([0.495 1.35167233]) to ([0.5445 1.48683957])\nIncrease radii from ([0.5445 1.48683957]) to ([0.59895 1.63552352])\nIncrease radii from ([0.45 1.18348472]) to ([0.495 1.30183319])\nIncrease radii from ([0.495 1.30183319]) to ([0.5445 1.43201651])\nIncrease radii from ([0.5445 1.43201651]) to ([0.59895 1.57521816])\nIncrease radii from ([0.59895 1.57521816]) to ([0.658845 1.73273997])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.09226047]) to ([0.44 1.20148652])\nIncrease radii from ([0.44 1.20148652]) to ([0.484 1.32163517])\nIncrease radii from ([0.484 1.32163517]) to ([0.5324 1.45379869])\nIncrease radii from ([0.5324 1.45379869]) to ([0.58564 1.59917855])\nIncrease radii from ([0.58564 1.59917855]) to ([0.644204 1.75909641])\nIncrease radii from ([0.4 1.05198642]) to ([0.44 1.15718506])\nIncrease radii from ([0.44 1.15718506]) to ([0.484 1.27290356])\nIncrease radii from ([0.484 1.27290356]) to ([0.5324 1.40019392])\nIncrease radii from ([0.5324 1.40019392]) to ([0.58564 1.54021331])\nIncrease radii from ([0.58564 1.54021331]) to ([0.644204 1.69423464])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.95572791]) to ([0.385 1.0513007])\nIncrease radii from ([0.385 1.0513007]) to ([0.4235 1.15643077])\nIncrease radii from ([0.4235 1.15643077]) to ([0.46585 1.27207385])\nIncrease radii from ([0.46585 1.27207385]) to ([0.512435 1.39928124])\nIncrease radii from ([0.512435 1.39928124]) to ([0.5636785 1.53920936])\nIncrease radii from ([0.5636785 1.53920936]) to ([0.62004635 1.69313029])\nIncrease radii from ([0.35 0.92048811]) to ([0.385 1.01253692])\nIncrease radii from ([0.385 1.01253692]) to ([0.4235 1.11379062])\nIncrease radii from ([0.4235 1.11379062]) to ([0.46585 1.22516968])\nIncrease radii from ([0.46585 1.22516968]) to ([0.512435 1.34768665])\nIncrease radii from ([0.512435 1.34768665]) to ([0.5636785 1.48245531])\nIncrease radii from ([0.5636785 1.48245531]) to ([0.62004635 1.63070084])\nIncrease radii from ([0.3 1.60351614]) to ([0.33 1.76386776])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.81919535]) to ([0.33 0.90111489])\nIncrease radii from ([0.33 0.90111489]) to ([0.363 0.99122638])\nIncrease radii from ([0.363 0.99122638]) to ([0.3993 1.09034901])\nIncrease radii from ([0.3993 1.09034901]) to ([0.43923 1.19938392])\nIncrease radii from ([0.43923 1.19938392]) to ([0.483153 1.31932231])\nIncrease radii from ([0.483153 1.31932231]) to ([0.5314683 1.45125454])\nIncrease radii from ([0.5314683 1.45125454]) to ([0.58461513 1.59637999])\nIncrease radii from ([0.58461513 1.59637999]) to ([0.64307664 1.75601799])\nIncrease radii from ([0.3 0.78898981]) to ([0.33 0.86788879])\nIncrease radii from ([0.33 0.86788879]) to ([0.363 0.95467767])\nIncrease radii from ([0.363 0.95467767]) to ([0.3993 1.05014544])\nIncrease radii from ([0.3993 1.05014544]) to ([0.43923 1.15515998])\nIncrease radii from ([0.43923 1.15515998]) to ([0.483153 1.27067598])\nIncrease radii from ([0.483153 1.27067598]) to ([0.5314683 1.39774358])\nIncrease radii from ([0.5314683 1.39774358]) to ([0.58461513 1.53751794])\nIncrease radii from ([0.58461513 1.53751794]) to ([0.64307664 1.69126973])\nIncrease radii from ([0.3 1.13210508]) to ([0.33 1.24531559])\nIncrease radii from ([0.33 1.24531559]) to ([0.363 1.36984715])\nIncrease radii from ([0.25 1.33626345]) to ([0.275 1.4698898])\nIncrease radii from ([0.275 1.4698898]) to ([0.3025 1.61687878])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.68266279]) to ([0.275 0.75092907])\nIncrease radii from ([0.275 0.75092907]) to ([0.3025 0.82602198])\nIncrease radii from ([0.3025 0.82602198]) to ([0.33275 0.90862418])\nIncrease radii from ([0.33275 0.90862418]) to ([0.366025 0.9994866])\nIncrease radii from ([0.366025 0.9994866]) to ([0.4026275 1.09943526])\nIncrease radii from ([0.4026275 1.09943526]) to ([0.44289025 1.20937878])\nIncrease radii from ([0.44289025 1.20937878]) to ([0.48717928 1.33031666])\nIncrease radii from ([0.48717928 1.33031666]) to ([0.5358972 1.46334833])\nIncrease radii from ([0.5358972 1.46334833]) to ([0.58948692 1.60968316])\nIncrease radii from ([0.58948692 1.60968316]) to ([0.64843562 1.77065147])\nIncrease radii from ([0.43956461 0.76134824]) to ([0.48352107 0.83748306])\nIncrease radii from ([0.25 0.65749151]) to ([0.275 0.72324066])\nIncrease radii from ([0.275 0.72324066]) to ([0.3025 0.79556473])\nIncrease radii from ([0.3025 0.79556473]) to ([0.33275 0.8751212])\nIncrease radii from ([0.33275 0.8751212]) to ([0.366025 0.96263332])\nIncrease radii from ([0.366025 0.96263332]) to ([0.4026275 1.05889665])\nIncrease radii from ([0.4026275 1.05889665]) to ([0.44289025 1.16478632])\nIncrease radii from ([0.44289025 1.16478632]) to ([0.48717928 1.28126495])\nIncrease radii from ([0.48717928 1.28126495]) to ([0.5358972 1.40939144])\nIncrease radii from ([0.5358972 1.40939144]) to ([0.58948692 1.55033059])\nIncrease radii from ([0.58948692 1.55033059]) to ([0.64843562 1.70536365])\nIncrease radii from ([0.25 0.9434209]) to ([0.275 1.03776299])\nIncrease radii from ([0.275 1.03776299]) to ([0.3025 1.14153929])\nIncrease radii from ([0.3025 1.14153929]) to ([0.33275 1.25569322])\nIncrease radii from ([0.33275 1.25569322]) to ([0.366025 1.38126254])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1714.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.4928933]) to ([0.605 1.64218263])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.35717573]) to ([0.55 1.4928933])\nIncrease radii from ([0.55 1.4928933]) to ([0.605 1.64218263])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.22145816]) to ([0.495 1.34360397])\nIncrease radii from ([0.495 1.34360397]) to ([0.5445 1.47796437])\nIncrease radii from ([0.5445 1.47796437]) to ([0.59895 1.62576081])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.08574058]) to ([0.44 1.19431464])\nIncrease radii from ([0.44 1.19431464]) to ([0.484 1.31374611])\nIncrease radii from ([0.484 1.31374611]) to ([0.5324 1.44512072])\nIncrease radii from ([0.5324 1.44512072]) to ([0.58564 1.58963279])\nIncrease radii from ([0.58564 1.58963279]) to ([0.644204 1.74859607])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.95002301]) to ([0.385 1.04502531])\nIncrease radii from ([0.385 1.04502531]) to ([0.4235 1.14952784])\nIncrease radii from ([0.4235 1.14952784]) to ([0.46585 1.26448063])\nIncrease radii from ([0.46585 1.26448063]) to ([0.512435 1.39092869])\nIncrease radii from ([0.512435 1.39092869]) to ([0.5636785 1.53002156])\nIncrease radii from ([0.5636785 1.53002156]) to ([0.62004635 1.68302371])\nIncrease radii from ([0.3 1.60351614]) to ([0.33 1.76386776])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.81430544]) to ([0.33 0.89573598])\nIncrease radii from ([0.33 0.89573598]) to ([0.363 0.98530958])\nIncrease radii from ([0.363 0.98530958]) to ([0.3993 1.08384054])\nIncrease radii from ([0.3993 1.08384054]) to ([0.43923 1.19222459])\nIncrease radii from ([0.43923 1.19222459]) to ([0.483153 1.31144705])\nIncrease radii from ([0.483153 1.31144705]) to ([0.5314683 1.44259176])\nIncrease radii from ([0.5314683 1.44259176]) to ([0.58461513 1.58685093])\nIncrease radii from ([0.58461513 1.58685093]) to ([0.64307664 1.74553602])\nIncrease radii from ([0.25 1.33626345]) to ([0.275 1.4698898])\nIncrease radii from ([0.275 1.4698898]) to ([0.3025 1.61687878])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.67858786]) to ([0.275 0.74644665])\nIncrease radii from ([0.275 0.74644665]) to ([0.3025 0.82109132])\nIncrease radii from ([0.3025 0.82109132]) to ([0.33275 0.90320045])\nIncrease radii from ([0.33275 0.90320045]) to ([0.366025 0.99352049])\nIncrease radii from ([0.366025 0.99352049]) to ([0.4026275 1.09287254])\nIncrease radii from ([0.4026275 1.09287254]) to ([0.44289025 1.2021598 ])\nIncrease radii from ([0.44289025 1.2021598 ]) to ([0.48717928 1.32237578])\nIncrease radii from ([0.48717928 1.32237578]) to ([0.5358972 1.45461335])\nIncrease radii from ([0.5358972 1.45461335]) to ([0.58948692 1.60007469])\nIncrease radii from ([0.58948692 1.60007469]) to ([0.64843562 1.76008216])\nIncrease radii from ([0.25 0.93791081]) to ([0.275 1.03170189])\nIncrease radii from ([0.275 1.03170189]) to ([0.3025 1.13487207])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1715.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.59252706]) to ([0.66 1.75177976])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.48010898]) to ([0.605 1.62811987])\nIncrease radii from ([0.55 1.45981647]) to ([0.605 1.60579811])\nIncrease radii from ([0.605 1.60579811]) to ([0.6655 1.76637793])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.34555361]) to ([0.55 1.48010898])\nIncrease radii from ([0.55 1.48010898]) to ([0.605 1.62811987])\nIncrease radii from ([0.5 1.32710588]) to ([0.55 1.45981647])\nIncrease radii from ([0.55 1.45981647]) to ([0.605 1.60579811])\nIncrease radii from ([0.605 1.60579811]) to ([0.6655 1.76637793])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.21099825]) to ([0.495 1.33209808])\nIncrease radii from ([0.495 1.33209808]) to ([0.5445 1.46530789])\nIncrease radii from ([0.5445 1.46530789]) to ([0.59895 1.61183868])\nIncrease radii from ([0.59895 1.61183868]) to ([0.658845 1.77302254])\nIncrease radii from ([0.45 1.19439529]) to ([0.495 1.31383482])\nIncrease radii from ([0.495 1.31383482]) to ([0.5445 1.4452183])\nIncrease radii from ([0.5445 1.4452183]) to ([0.59895 1.58974013])\nIncrease radii from ([0.59895 1.58974013]) to ([0.658845 1.74871415])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.07644289]) to ([0.44 1.18408718])\nIncrease radii from ([0.44 1.18408718]) to ([0.484 1.3024959])\nIncrease radii from ([0.484 1.3024959]) to ([0.5324 1.43274549])\nIncrease radii from ([0.5324 1.43274549]) to ([0.58564 1.57602004])\nIncrease radii from ([0.58564 1.57602004]) to ([0.644204 1.73362204])\nIncrease radii from ([0.4 1.0616847]) to ([0.44 1.16785317])\nIncrease radii from ([0.44 1.16785317]) to ([0.484 1.28463849])\nIncrease radii from ([0.484 1.28463849]) to ([0.5324 1.41310234])\nIncrease radii from ([0.5324 1.41310234]) to ([0.58564 1.55441257])\nIncrease radii from ([0.58564 1.55441257]) to ([0.644204 1.70985383])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.94188753]) to ([0.385 1.03607628])\nIncrease radii from ([0.385 1.03607628]) to ([0.4235 1.13968391])\nIncrease radii from ([0.4235 1.13968391]) to ([0.46585 1.2536523])\nIncrease radii from ([0.46585 1.2536523]) to ([0.512435 1.37901753])\nIncrease radii from ([0.512435 1.37901753]) to ([0.5636785 1.51691929])\nIncrease radii from ([0.5636785 1.51691929]) to ([0.62004635 1.66861122])\nIncrease radii from ([0.35 0.92897412]) to ([0.385 1.02187153])\nIncrease radii from ([0.385 1.02187153]) to ([0.4235 1.12405868])\nIncrease radii from ([0.4235 1.12405868]) to ([0.46585 1.23646455])\nIncrease radii from ([0.46585 1.23646455]) to ([0.512435 1.360111])\nIncrease radii from ([0.512435 1.360111]) to ([0.5636785 1.4961221])\nIncrease radii from ([0.5636785 1.4961221]) to ([0.62004635 1.64573431])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.80733217]) to ([0.33 0.88806539])\nIncrease radii from ([0.33 0.88806539]) to ([0.363 0.97687192])\nIncrease radii from ([0.363 0.97687192]) to ([0.3993 1.07455912])\nIncrease radii from ([0.3993 1.07455912]) to ([0.43923 1.18201503])\nIncrease radii from ([0.43923 1.18201503]) to ([0.483153 1.30021653])\nIncrease radii from ([0.483153 1.30021653]) to ([0.5314683 1.43023818])\nIncrease radii from ([0.5314683 1.43023818]) to ([0.58461513 1.573262 ])\nIncrease radii from ([0.58461513 1.573262 ]) to ([0.64307664 1.7305882 ])\nIncrease radii from ([0.3 0.79626353]) to ([0.33 0.87588988])\nIncrease radii from ([0.33 0.87588988]) to ([0.363 0.96347887])\nIncrease radii from ([0.363 0.96347887]) to ([0.3993 1.05982676])\nIncrease radii from ([0.3993 1.05982676]) to ([0.43923 1.16580943])\nIncrease radii from ([0.43923 1.16580943]) to ([0.483153 1.28239037])\nIncrease radii from ([0.483153 1.28239037]) to ([0.5314683 1.41062941])\nIncrease radii from ([0.5314683 1.41062941]) to ([0.58461513 1.55169235])\nIncrease radii from ([0.58461513 1.55169235]) to ([0.64307664 1.70686159])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.67277681]) to ([0.275 0.74005449])\nIncrease radii from ([0.275 0.74005449]) to ([0.3025 0.81405994])\nIncrease radii from ([0.3025 0.81405994]) to ([0.33275 0.89546593])\nIncrease radii from ([0.33275 0.89546593]) to ([0.366025 0.98501252])\nIncrease radii from ([0.366025 0.98501252]) to ([0.4026275 1.08351378])\nIncrease radii from ([0.4026275 1.08351378]) to ([0.44289025 1.19186515])\nIncrease radii from ([0.44289025 1.19186515]) to ([0.48717928 1.31105167])\nIncrease radii from ([0.48717928 1.31105167]) to ([0.5358972 1.44215684])\nIncrease radii from ([0.5358972 1.44215684]) to ([0.58948692 1.58637252])\nIncrease radii from ([0.58948692 1.58637252]) to ([0.64843562 1.74500977])\nIncrease radii from ([0.25 0.66355294]) to ([0.275 0.72990823])\nIncrease radii from ([0.275 0.72990823]) to ([0.3025 0.80289906])\nIncrease radii from ([0.3025 0.80289906]) to ([0.33275 0.88318896])\nIncrease radii from ([0.33275 0.88318896]) to ([0.366025 0.97150786])\nIncrease radii from ([0.366025 0.97150786]) to ([0.4026275 1.06865865])\nIncrease radii from ([0.4026275 1.06865865]) to ([0.44289025 1.17552451])\nIncrease radii from ([0.44289025 1.17552451]) to ([0.48717928 1.29307696])\nIncrease radii from ([0.48717928 1.29307696]) to ([0.5358972 1.42238466])\nIncrease radii from ([0.5358972 1.42238466]) to ([0.58948692 1.56462312])\nIncrease radii from ([0.58948692 1.56462312]) to ([0.64843562 1.72108543])\nIncrease radii from ([0.25 0.93791081]) to ([0.275 1.03170189])\nIncrease radii from ([0.275 1.03170189]) to ([0.3025 1.13487207])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1716.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.5674575]) to ([0.66 1.72420325])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.48010898]) to ([0.605 1.62811987])\nIncrease radii from ([0.55 1.43683604]) to ([0.605 1.58051964])\nIncrease radii from ([0.605 1.58051964]) to ([0.6655 1.73857161])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.34555361]) to ([0.55 1.48010898])\nIncrease radii from ([0.55 1.48010898]) to ([0.605 1.62811987])\nIncrease radii from ([0.5 1.30621458]) to ([0.55 1.43683604])\nIncrease radii from ([0.55 1.43683604]) to ([0.605 1.58051964])\nIncrease radii from ([0.605 1.58051964]) to ([0.6655 1.73857161])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.21099825]) to ([0.495 1.33209808])\nIncrease radii from ([0.495 1.33209808]) to ([0.5445 1.46530789])\nIncrease radii from ([0.5445 1.46530789]) to ([0.59895 1.61183868])\nIncrease radii from ([0.59895 1.61183868]) to ([0.658845 1.77302254])\nIncrease radii from ([0.45 1.17559312]) to ([0.495 1.29315243])\nIncrease radii from ([0.495 1.29315243]) to ([0.5445 1.42246768])\nIncrease radii from ([0.5445 1.42246768]) to ([0.59895 1.56471445])\nIncrease radii from ([0.59895 1.56471445]) to ([0.658845 1.72118589])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.07644289]) to ([0.44 1.18408718])\nIncrease radii from ([0.44 1.18408718]) to ([0.484 1.3024959])\nIncrease radii from ([0.484 1.3024959]) to ([0.5324 1.43274549])\nIncrease radii from ([0.5324 1.43274549]) to ([0.58564 1.57602004])\nIncrease radii from ([0.58564 1.57602004]) to ([0.644204 1.73362204])\nIncrease radii from ([0.4 1.04497166]) to ([0.44 1.14946883])\nIncrease radii from ([0.44 1.14946883]) to ([0.484 1.26441571])\nIncrease radii from ([0.484 1.26441571]) to ([0.5324 1.39085729])\nIncrease radii from ([0.5324 1.39085729]) to ([0.58564 1.52994301])\nIncrease radii from ([0.58564 1.52994301]) to ([0.644204 1.68293732])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.94188753]) to ([0.385 1.03607628])\nIncrease radii from ([0.385 1.03607628]) to ([0.4235 1.13968391])\nIncrease radii from ([0.4235 1.13968391]) to ([0.46585 1.2536523])\nIncrease radii from ([0.46585 1.2536523]) to ([0.512435 1.37901753])\nIncrease radii from ([0.512435 1.37901753]) to ([0.5636785 1.51691929])\nIncrease radii from ([0.5636785 1.51691929]) to ([0.62004635 1.66861122])\nIncrease radii from ([0.35 0.91435021]) to ([0.385 1.00578523])\nIncrease radii from ([0.385 1.00578523]) to ([0.4235 1.10636375])\nIncrease radii from ([0.4235 1.10636375]) to ([0.46585 1.21700012])\nIncrease radii from ([0.46585 1.21700012]) to ([0.512435 1.33870014])\nIncrease radii from ([0.512435 1.33870014]) to ([0.5636785 1.47257015])\nIncrease radii from ([0.5636785 1.47257015]) to ([0.62004635 1.61982717])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 1.13411196]) to ([0.33 1.24752316])\nIncrease radii from ([0.33 1.24752316]) to ([0.363 1.37227548])\nIncrease radii from ([0.3 0.80733217]) to ([0.33 0.88806539])\nIncrease radii from ([0.33 0.88806539]) to ([0.363 0.97687192])\nIncrease radii from ([0.363 0.97687192]) to ([0.3993 1.07455912])\nIncrease radii from ([0.3993 1.07455912]) to ([0.43923 1.18201503])\nIncrease radii from ([0.43923 1.18201503]) to ([0.483153 1.30021653])\nIncrease radii from ([0.483153 1.30021653]) to ([0.5314683 1.43023818])\nIncrease radii from ([0.5314683 1.43023818]) to ([0.58461513 1.573262 ])\nIncrease radii from ([0.58461513 1.573262 ]) to ([0.64307664 1.7305882 ])\nIncrease radii from ([0.3 0.78372875]) to ([0.33 0.86210162])\nIncrease radii from ([0.33 0.86210162]) to ([0.363 0.94831179])\nIncrease radii from ([0.363 0.94831179]) to ([0.3993 1.04314296])\nIncrease radii from ([0.3993 1.04314296]) to ([0.43923 1.14745726])\nIncrease radii from ([0.43923 1.14745726]) to ([0.483153 1.26220299])\nIncrease radii from ([0.483153 1.26220299]) to ([0.5314683 1.38842329])\nIncrease radii from ([0.5314683 1.38842329]) to ([0.58461513 1.52726561])\nIncrease radii from ([0.58461513 1.52726561]) to ([0.64307664 1.67999218])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.9450933]) to ([0.275 1.03960263])\nIncrease radii from ([0.275 1.03960263]) to ([0.3025 1.1435629])\nIncrease radii from ([0.3025 1.1435629]) to ([0.33275 1.25791919])\nIncrease radii from ([0.33275 1.25791919]) to ([0.366025 1.38371111])\nIncrease radii from ([0.25 0.67277681]) to ([0.275 0.74005449])\nIncrease radii from ([0.275 0.74005449]) to ([0.3025 0.81405994])\nIncrease radii from ([0.3025 0.81405994]) to ([0.33275 0.89546593])\nIncrease radii from ([0.33275 0.89546593]) to ([0.366025 0.98501252])\nIncrease radii from ([0.366025 0.98501252]) to ([0.4026275 1.08351378])\nIncrease radii from ([0.4026275 1.08351378]) to ([0.44289025 1.19186515])\nIncrease radii from ([0.44289025 1.19186515]) to ([0.48717928 1.31105167])\nIncrease radii from ([0.48717928 1.31105167]) to ([0.5358972 1.44215684])\nIncrease radii from ([0.5358972 1.44215684]) to ([0.58948692 1.58637252])\nIncrease radii from ([0.58948692 1.58637252]) to ([0.64843562 1.74500977])\nIncrease radii from ([0.25 0.65310729]) to ([0.275 0.71841802])\nIncrease radii from ([0.275 0.71841802]) to ([0.3025 0.79025982])\nIncrease radii from ([0.3025 0.79025982]) to ([0.33275 0.8692858])\nIncrease radii from ([0.33275 0.8692858]) to ([0.366025 0.95621438])\nIncrease radii from ([0.366025 0.95621438]) to ([0.4026275 1.05183582])\nIncrease radii from ([0.4026275 1.05183582]) to ([0.44289025 1.1570194 ])\nIncrease radii from ([0.44289025 1.1570194 ]) to ([0.48717928 1.27272134])\nIncrease radii from ([0.48717928 1.27272134]) to ([0.5358972 1.39999348])\nIncrease radii from ([0.5358972 1.39999348]) to ([0.58948692 1.53999283])\nIncrease radii from ([0.58948692 1.53999283]) to ([0.64843562 1.69399211])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1717.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.51020186]) to ([0.605 1.66122205])\nIncrease radii from ([0.55 1.51446898]) to ([0.605 1.66591588])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.37291079]) to ([0.55 1.51020186])\nIncrease radii from ([0.55 1.51020186]) to ([0.605 1.66122205])\nIncrease radii from ([0.5 1.37678998]) to ([0.55 1.51446898])\nIncrease radii from ([0.55 1.51446898]) to ([0.605 1.66591588])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.23561971]) to ([0.495 1.35918168])\nIncrease radii from ([0.495 1.35918168]) to ([0.5445 1.49509985])\nIncrease radii from ([0.5445 1.49509985]) to ([0.59895 1.64460983])\nIncrease radii from ([0.45 1.23911098]) to ([0.495 1.36302208])\nIncrease radii from ([0.495 1.36302208]) to ([0.5445 1.49932429])\nIncrease radii from ([0.5445 1.49932429]) to ([0.59895 1.64925672])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.09832863]) to ([0.44 1.20816149])\nIncrease radii from ([0.44 1.20816149]) to ([0.484 1.32897764])\nIncrease radii from ([0.484 1.32897764]) to ([0.5324 1.4618754])\nIncrease radii from ([0.5324 1.4618754]) to ([0.58564 1.60806295])\nIncrease radii from ([0.58564 1.60806295]) to ([0.644204 1.76886924])\nIncrease radii from ([0.4 1.10143199]) to ([0.44 1.21157518])\nIncrease radii from ([0.44 1.21157518]) to ([0.484 1.3327327])\nIncrease radii from ([0.484 1.3327327]) to ([0.5324 1.46600597])\nIncrease radii from ([0.5324 1.46600597]) to ([0.58564 1.61260657])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.96103755]) to ([0.385 1.0571413])\nIncrease radii from ([0.385 1.0571413]) to ([0.4235 1.16285544])\nIncrease radii from ([0.4235 1.16285544]) to ([0.46585 1.27914098])\nIncrease radii from ([0.46585 1.27914098]) to ([0.512435 1.40705508])\nIncrease radii from ([0.512435 1.40705508]) to ([0.5636785 1.54776058])\nIncrease radii from ([0.5636785 1.54776058]) to ([0.62004635 1.70253664])\nIncrease radii from ([0.35 0.96375299]) to ([0.385 1.06012829])\nIncrease radii from ([0.385 1.06012829]) to ([0.4235 1.16614111])\nIncrease radii from ([0.4235 1.16614111]) to ([0.46585 1.28275523])\nIncrease radii from ([0.46585 1.28275523]) to ([0.512435 1.41103075])\nIncrease radii from ([0.512435 1.41103075]) to ([0.5636785 1.55213382])\nIncrease radii from ([0.5636785 1.55213382]) to ([0.62004635 1.70734721])\nIncrease radii from ([0.93327605 0.93327605]) to ([1.02660366 1.02660366])\nIncrease radii from ([0.3 1.60543876]) to ([0.33 1.76598264])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.82374647]) to ([0.33 0.90612112])\nIncrease radii from ([0.33 0.90612112]) to ([0.363 0.99673323])\nIncrease radii from ([0.363 0.99673323]) to ([0.3993 1.09640655])\nIncrease radii from ([0.3993 1.09640655]) to ([0.43923 1.20604721])\nIncrease radii from ([0.43923 1.20604721]) to ([0.483153 1.32665193])\nIncrease radii from ([0.483153 1.32665193]) to ([0.5314683 1.45931712])\nIncrease radii from ([0.5314683 1.45931712]) to ([0.58461513 1.60524883])\nIncrease radii from ([0.58461513 1.60524883]) to ([0.64307664 1.76577372])\nIncrease radii from ([0.3 0.82607399]) to ([0.33 0.90868139])\nIncrease radii from ([0.33 0.90868139]) to ([0.363 0.99954953])\nIncrease radii from ([0.363 0.99954953]) to ([0.3993 1.09950448])\nIncrease radii from ([0.3993 1.09950448]) to ([0.43923 1.20945493])\nIncrease radii from ([0.43923 1.20945493]) to ([0.483153 1.33040042])\nIncrease radii from ([0.483153 1.33040042]) to ([0.5314683 1.46344046])\nIncrease radii from ([0.5314683 1.46344046]) to ([0.58461513 1.60978451])\nIncrease radii from ([0.58461513 1.60978451]) to ([0.64307664 1.77076296])\nIncrease radii from ([0.7999509 0.7999509]) to ([0.87994599 0.87994599])\nIncrease radii from ([0.87994599 0.87994599]) to ([0.96794059 0.96794059])\nIncrease radii from ([0.25 1.33786564]) to ([0.275 1.4716522])\nIncrease radii from ([0.275 1.4716522]) to ([0.3025 1.61881742])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.68645539]) to ([0.275 0.75510093])\nIncrease radii from ([0.275 0.75510093]) to ([0.3025 0.83061103])\nIncrease radii from ([0.3025 0.83061103]) to ([0.33275 0.91367213])\nIncrease radii from ([0.33275 0.91367213]) to ([0.366025 1.00503934])\nIncrease radii from ([0.366025 1.00503934]) to ([0.4026275 1.10554327])\nIncrease radii from ([0.4026275 1.10554327]) to ([0.44289025 1.2160976 ])\nIncrease radii from ([0.44289025 1.2160976 ]) to ([0.48717928 1.33770736])\nIncrease radii from ([0.48717928 1.33770736]) to ([0.5358972 1.4714781])\nIncrease radii from ([0.5358972 1.4714781]) to ([0.58948692 1.61862591])\nIncrease radii from ([0.25 0.68839499]) to ([0.275 0.75723449])\nIncrease radii from ([0.275 0.75723449]) to ([0.3025 0.83295794])\nIncrease radii from ([0.3025 0.83295794]) to ([0.33275 0.91625373])\nIncrease radii from ([0.33275 0.91625373]) to ([0.366025 1.00787911])\nIncrease radii from ([0.366025 1.00787911]) to ([0.4026275 1.10866702])\nIncrease radii from ([0.4026275 1.10866702]) to ([0.44289025 1.21953372])\nIncrease radii from ([0.44289025 1.21953372]) to ([0.48717928 1.34148709])\nIncrease radii from ([0.48717928 1.34148709]) to ([0.5358972 1.4756358])\nIncrease radii from ([0.5358972 1.4756358]) to ([0.58948692 1.62319938])\nIncrease radii from ([0.66662575 0.66662575]) to ([0.73328833 0.73328833])\nIncrease radii from ([0.73328833 0.73328833]) to ([0.80661716 0.80661716])\nIncrease radii from ([0.80661716 0.80661716]) to ([0.88727888 0.88727888])\nIncrease radii from ([0.88727888 0.88727888]) to ([0.97600676 0.97600676])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1718.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.51446898]) to ([0.605 1.66591588])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.37678998]) to ([0.55 1.51446898])\nIncrease radii from ([0.55 1.51446898]) to ([0.605 1.66591588])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.23911098]) to ([0.495 1.36302208])\nIncrease radii from ([0.495 1.36302208]) to ([0.5445 1.49932429])\nIncrease radii from ([0.5445 1.49932429]) to ([0.59895 1.64925672])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.10143199]) to ([0.44 1.21157518])\nIncrease radii from ([0.44 1.21157518]) to ([0.484 1.3327327])\nIncrease radii from ([0.484 1.3327327]) to ([0.5324 1.46600597])\nIncrease radii from ([0.5324 1.46600597]) to ([0.58564 1.61260657])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.96375299]) to ([0.385 1.06012829])\nIncrease radii from ([0.385 1.06012829]) to ([0.4235 1.16614111])\nIncrease radii from ([0.4235 1.16614111]) to ([0.46585 1.28275523])\nIncrease radii from ([0.46585 1.28275523]) to ([0.512435 1.41103075])\nIncrease radii from ([0.512435 1.41103075]) to ([0.5636785 1.55213382])\nIncrease radii from ([0.5636785 1.55213382]) to ([0.62004635 1.70734721])\nIncrease radii from ([0.93327605 0.93327605]) to ([1.02660366 1.02660366])\nIncrease radii from ([0.3 1.60543876]) to ([0.33 1.76598264])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.82607399]) to ([0.33 0.90868139])\nIncrease radii from ([0.33 0.90868139]) to ([0.363 0.99954953])\nIncrease radii from ([0.363 0.99954953]) to ([0.3993 1.09950448])\nIncrease radii from ([0.3993 1.09950448]) to ([0.43923 1.20945493])\nIncrease radii from ([0.43923 1.20945493]) to ([0.483153 1.33040042])\nIncrease radii from ([0.483153 1.33040042]) to ([0.5314683 1.46344046])\nIncrease radii from ([0.5314683 1.46344046]) to ([0.58461513 1.60978451])\nIncrease radii from ([0.58461513 1.60978451]) to ([0.64307664 1.77076296])\nIncrease radii from ([0.7999509 0.7999509]) to ([0.87994599 0.87994599])\nIncrease radii from ([0.87994599 0.87994599]) to ([0.96794059 0.96794059])\nIncrease radii from ([0.25 1.33786564]) to ([0.275 1.4716522])\nIncrease radii from ([0.275 1.4716522]) to ([0.3025 1.61881742])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.68839499]) to ([0.275 0.75723449])\nIncrease radii from ([0.275 0.75723449]) to ([0.3025 0.83295794])\nIncrease radii from ([0.3025 0.83295794]) to ([0.33275 0.91625373])\nIncrease radii from ([0.33275 0.91625373]) to ([0.366025 1.00787911])\nIncrease radii from ([0.366025 1.00787911]) to ([0.4026275 1.10866702])\nIncrease radii from ([0.4026275 1.10866702]) to ([0.44289025 1.21953372])\nIncrease radii from ([0.44289025 1.21953372]) to ([0.48717928 1.34148709])\nIncrease radii from ([0.48717928 1.34148709]) to ([0.5358972 1.4756358])\nIncrease radii from ([0.5358972 1.4756358]) to ([0.58948692 1.62319938])\nIncrease radii from ([0.25 0.93924408]) to ([0.275 1.03316849])\nIncrease radii from ([0.275 1.03316849]) to ([0.3025 1.13648534])\nIncrease radii from ([0.25 0.92405129]) to ([0.275 1.01645642])\nIncrease radii from ([0.275 1.01645642]) to ([0.3025 1.11810206])\nIncrease radii from ([0.66662575 0.66662575]) to ([0.73328833 0.73328833])\nIncrease radii from ([0.73328833 0.73328833]) to ([0.80661716 0.80661716])\nIncrease radii from ([0.80661716 0.80661716]) to ([0.88727888 0.88727888])\nIncrease radii from ([0.88727888 0.88727888]) to ([0.97600676 0.97600676])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1719.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.51446898]) to ([0.605 1.66591588])\nIncrease radii from ([0.55 1.48888119]) to ([0.605 1.63776931])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.37678998]) to ([0.55 1.51446898])\nIncrease radii from ([0.55 1.51446898]) to ([0.605 1.66591588])\nIncrease radii from ([0.5 1.35352836]) to ([0.55 1.48888119])\nIncrease radii from ([0.55 1.48888119]) to ([0.605 1.63776931])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.23911098]) to ([0.495 1.36302208])\nIncrease radii from ([0.495 1.36302208]) to ([0.5445 1.49932429])\nIncrease radii from ([0.5445 1.49932429]) to ([0.59895 1.64925672])\nIncrease radii from ([0.45 1.21817552]) to ([0.495 1.33999307])\nIncrease radii from ([0.495 1.33999307]) to ([0.5445 1.47399238])\nIncrease radii from ([0.5445 1.47399238]) to ([0.59895 1.62139162])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.10143199]) to ([0.44 1.21157518])\nIncrease radii from ([0.44 1.21157518]) to ([0.484 1.3327327])\nIncrease radii from ([0.484 1.3327327]) to ([0.5324 1.46600597])\nIncrease radii from ([0.5324 1.46600597]) to ([0.58564 1.61260657])\nIncrease radii from ([0.4 1.08282269]) to ([0.44 1.19110495])\nIncrease radii from ([0.44 1.19110495]) to ([0.484 1.31021545])\nIncrease radii from ([0.484 1.31021545]) to ([0.5324 1.44123699])\nIncrease radii from ([0.5324 1.44123699]) to ([0.58564 1.58536069])\nIncrease radii from ([0.58564 1.58536069]) to ([0.644204 1.74389676])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.96375299]) to ([0.385 1.06012829])\nIncrease radii from ([0.385 1.06012829]) to ([0.4235 1.16614111])\nIncrease radii from ([0.4235 1.16614111]) to ([0.46585 1.28275523])\nIncrease radii from ([0.46585 1.28275523]) to ([0.512435 1.41103075])\nIncrease radii from ([0.512435 1.41103075]) to ([0.5636785 1.55213382])\nIncrease radii from ([0.5636785 1.55213382]) to ([0.62004635 1.70734721])\nIncrease radii from ([0.35 0.94746985]) to ([0.385 1.04221683])\nIncrease radii from ([0.385 1.04221683]) to ([0.4235 1.14643852])\nIncrease radii from ([0.4235 1.14643852]) to ([0.46585 1.26108237])\nIncrease radii from ([0.46585 1.26108237]) to ([0.512435 1.38719061])\nIncrease radii from ([0.512435 1.38719061]) to ([0.5636785 1.52590967])\nIncrease radii from ([0.5636785 1.52590967]) to ([0.62004635 1.67850063])\nIncrease radii from ([0.3 1.60543876]) to ([0.33 1.76598264])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.82607399]) to ([0.33 0.90868139])\nIncrease radii from ([0.33 0.90868139]) to ([0.363 0.99954953])\nIncrease radii from ([0.363 0.99954953]) to ([0.3993 1.09950448])\nIncrease radii from ([0.3993 1.09950448]) to ([0.43923 1.20945493])\nIncrease radii from ([0.43923 1.20945493]) to ([0.483153 1.33040042])\nIncrease radii from ([0.483153 1.33040042]) to ([0.5314683 1.46344046])\nIncrease radii from ([0.5314683 1.46344046]) to ([0.58461513 1.60978451])\nIncrease radii from ([0.58461513 1.60978451]) to ([0.64307664 1.77076296])\nIncrease radii from ([0.3 0.81211701]) to ([0.33 0.89332872])\nIncrease radii from ([0.33 0.89332872]) to ([0.363 0.98266159])\nIncrease radii from ([0.363 0.98266159]) to ([0.3993 1.08092775])\nIncrease radii from ([0.3993 1.08092775]) to ([0.43923 1.18902052])\nIncrease radii from ([0.43923 1.18902052]) to ([0.483153 1.30792257])\nIncrease radii from ([0.483153 1.30792257]) to ([0.5314683 1.43871483])\nIncrease radii from ([0.5314683 1.43871483]) to ([0.58461513 1.58258631])\nIncrease radii from ([0.58461513 1.58258631]) to ([0.64307664 1.74084494])\nIncrease radii from ([0.25 1.33786564]) to ([0.275 1.4716522])\nIncrease radii from ([0.275 1.4716522]) to ([0.3025 1.61881742])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.68839499]) to ([0.275 0.75723449])\nIncrease radii from ([0.275 0.75723449]) to ([0.3025 0.83295794])\nIncrease radii from ([0.3025 0.83295794]) to ([0.33275 0.91625373])\nIncrease radii from ([0.33275 0.91625373]) to ([0.366025 1.00787911])\nIncrease radii from ([0.366025 1.00787911]) to ([0.4026275 1.10866702])\nIncrease radii from ([0.4026275 1.10866702]) to ([0.44289025 1.21953372])\nIncrease radii from ([0.44289025 1.21953372]) to ([0.48717928 1.34148709])\nIncrease radii from ([0.48717928 1.34148709]) to ([0.5358972 1.4756358])\nIncrease radii from ([0.5358972 1.4756358]) to ([0.58948692 1.62319938])\nIncrease radii from ([0.25 0.95472815]) to ([0.275 1.05020097])\nIncrease radii from ([0.275 1.05020097]) to ([0.3025 1.15522107])\nIncrease radii from ([0.25 0.67676418]) to ([0.275 0.7444406])\nIncrease radii from ([0.275 0.7444406]) to ([0.3025 0.81888466])\nIncrease radii from ([0.3025 0.81888466]) to ([0.33275 0.90077312])\nIncrease radii from ([0.33275 0.90077312]) to ([0.366025 0.99085043])\nIncrease radii from ([0.366025 0.99085043]) to ([0.4026275 1.08993548])\nIncrease radii from ([0.4026275 1.08993548]) to ([0.44289025 1.19892902])\nIncrease radii from ([0.44289025 1.19892902]) to ([0.48717928 1.31882193])\nIncrease radii from ([0.48717928 1.31882193]) to ([0.5358972 1.45070412])\nIncrease radii from ([0.5358972 1.45070412]) to ([0.58948692 1.59577453])\nIncrease radii from ([0.58948692 1.59577453]) to ([0.64843562 1.75535198])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1720.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.58854938]) to ([0.66 1.74740432])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.51446898]) to ([0.605 1.66591588])\nIncrease radii from ([0.55 1.45617027]) to ([0.605 1.60178729])\nIncrease radii from ([0.605 1.60178729]) to ([0.6655 1.76196602])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.37678998]) to ([0.55 1.51446898])\nIncrease radii from ([0.55 1.51446898]) to ([0.605 1.66591588])\nIncrease radii from ([0.5 1.32379115]) to ([0.55 1.45617027])\nIncrease radii from ([0.55 1.45617027]) to ([0.605 1.60178729])\nIncrease radii from ([0.605 1.60178729]) to ([0.6655 1.76196602])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.23911098]) to ([0.495 1.36302208])\nIncrease radii from ([0.495 1.36302208]) to ([0.5445 1.49932429])\nIncrease radii from ([0.5445 1.49932429]) to ([0.59895 1.64925672])\nIncrease radii from ([0.45 1.19141204]) to ([0.495 1.31055324])\nIncrease radii from ([0.495 1.31055324]) to ([0.5445 1.44160856])\nIncrease radii from ([0.5445 1.44160856]) to ([0.59895 1.58576942])\nIncrease radii from ([0.59895 1.58576942]) to ([0.658845 1.74434636])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.10143199]) to ([0.44 1.21157518])\nIncrease radii from ([0.44 1.21157518]) to ([0.484 1.3327327])\nIncrease radii from ([0.484 1.3327327]) to ([0.5324 1.46600597])\nIncrease radii from ([0.5324 1.46600597]) to ([0.58564 1.61260657])\nIncrease radii from ([0.4 1.05903292]) to ([0.44 1.16493621])\nIncrease radii from ([0.44 1.16493621]) to ([0.484 1.28142984])\nIncrease radii from ([0.484 1.28142984]) to ([0.5324 1.40957282])\nIncrease radii from ([0.5324 1.40957282]) to ([0.58564 1.5505301])\nIncrease radii from ([0.58564 1.5505301]) to ([0.644204 1.70558311])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.96375299]) to ([0.385 1.06012829])\nIncrease radii from ([0.385 1.06012829]) to ([0.4235 1.16614111])\nIncrease radii from ([0.4235 1.16614111]) to ([0.46585 1.28275523])\nIncrease radii from ([0.46585 1.28275523]) to ([0.512435 1.41103075])\nIncrease radii from ([0.512435 1.41103075]) to ([0.5636785 1.55213382])\nIncrease radii from ([0.5636785 1.55213382]) to ([0.62004635 1.70734721])\nIncrease radii from ([0.35 0.92665381]) to ([0.385 1.01931919])\nIncrease radii from ([0.385 1.01931919]) to ([0.4235 1.12125111])\nIncrease radii from ([0.4235 1.12125111]) to ([0.46585 1.23337622])\nIncrease radii from ([0.46585 1.23337622]) to ([0.512435 1.35671384])\nIncrease radii from ([0.512435 1.35671384]) to ([0.5636785 1.49238522])\nIncrease radii from ([0.5636785 1.49238522]) to ([0.62004635 1.64162374])\nIncrease radii from ([0.3 1.60543876]) to ([0.33 1.76598264])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.82607399]) to ([0.33 0.90868139])\nIncrease radii from ([0.33 0.90868139]) to ([0.363 0.99954953])\nIncrease radii from ([0.363 0.99954953]) to ([0.3993 1.09950448])\nIncrease radii from ([0.3993 1.09950448]) to ([0.43923 1.20945493])\nIncrease radii from ([0.43923 1.20945493]) to ([0.483153 1.33040042])\nIncrease radii from ([0.483153 1.33040042]) to ([0.5314683 1.46344046])\nIncrease radii from ([0.5314683 1.46344046]) to ([0.58461513 1.60978451])\nIncrease radii from ([0.58461513 1.60978451]) to ([0.64307664 1.77076296])\nIncrease radii from ([0.3 0.79427469]) to ([0.33 0.87370216])\nIncrease radii from ([0.33 0.87370216]) to ([0.363 0.96107238])\nIncrease radii from ([0.363 0.96107238]) to ([0.3993 1.05717961])\nIncrease radii from ([0.3993 1.05717961]) to ([0.43923 1.16289758])\nIncrease radii from ([0.43923 1.16289758]) to ([0.483153 1.27918733])\nIncrease radii from ([0.483153 1.27918733]) to ([0.5314683 1.40710607])\nIncrease radii from ([0.5314683 1.40710607]) to ([0.58461513 1.54781667])\nIncrease radii from ([0.58461513 1.54781667]) to ([0.64307664 1.70259834])\nIncrease radii from ([0.25 1.33786564]) to ([0.275 1.4716522])\nIncrease radii from ([0.275 1.4716522]) to ([0.3025 1.61881742])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.68839499]) to ([0.275 0.75723449])\nIncrease radii from ([0.275 0.75723449]) to ([0.3025 0.83295794])\nIncrease radii from ([0.3025 0.83295794]) to ([0.33275 0.91625373])\nIncrease radii from ([0.33275 0.91625373]) to ([0.366025 1.00787911])\nIncrease radii from ([0.366025 1.00787911]) to ([0.4026275 1.10866702])\nIncrease radii from ([0.4026275 1.10866702]) to ([0.44289025 1.21953372])\nIncrease radii from ([0.44289025 1.21953372]) to ([0.48717928 1.34148709])\nIncrease radii from ([0.48717928 1.34148709]) to ([0.5358972 1.4756358])\nIncrease radii from ([0.5358972 1.4756358]) to ([0.58948692 1.62319938])\nIncrease radii from ([0.25 0.66189558]) to ([0.275 0.72808513])\nIncrease radii from ([0.275 0.72808513]) to ([0.3025 0.80089365])\nIncrease radii from ([0.3025 0.80089365]) to ([0.33275 0.88098301])\nIncrease radii from ([0.33275 0.88098301]) to ([0.366025 0.96908131])\nIncrease radii from ([0.366025 0.96908131]) to ([0.4026275 1.06598944])\nIncrease radii from ([0.4026275 1.06598944]) to ([0.44289025 1.17258839])\nIncrease radii from ([0.44289025 1.17258839]) to ([0.48717928 1.28984723])\nIncrease radii from ([0.48717928 1.28984723]) to ([0.5358972 1.41883195])\nIncrease radii from ([0.5358972 1.41883195]) to ([0.58948692 1.56071515])\nIncrease radii from ([0.58948692 1.56071515]) to ([0.64843562 1.71678666])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1721.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.57121246]) to ([0.66 1.7283337])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.44027809]) to ([0.605 1.5843059])\nIncrease radii from ([0.605 1.5843059]) to ([0.6655 1.74273648])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.30934372]) to ([0.55 1.44027809])\nIncrease radii from ([0.55 1.44027809]) to ([0.605 1.5843059])\nIncrease radii from ([0.605 1.5843059]) to ([0.6655 1.74273648])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.17840934]) to ([0.495 1.29625028])\nIncrease radii from ([0.495 1.29625028]) to ([0.5445 1.42587531])\nIncrease radii from ([0.5445 1.42587531]) to ([0.59895 1.56846284])\nIncrease radii from ([0.59895 1.56846284]) to ([0.658845 1.72530912])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.04747497]) to ([0.44 1.15222247])\nIncrease radii from ([0.44 1.15222247]) to ([0.484 1.26744472])\nIncrease radii from ([0.484 1.26744472]) to ([0.5324 1.39418919])\nIncrease radii from ([0.5324 1.39418919]) to ([0.58564 1.53360811])\nIncrease radii from ([0.58564 1.53360811]) to ([0.644204 1.68696892])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.91659785 0.91659785]) to ([1.00825764 1.00825764])\nIncrease radii from ([0.35 0.9165406]) to ([0.385 1.00819466])\nIncrease radii from ([0.385 1.00819466]) to ([0.4235 1.10901413])\nIncrease radii from ([0.4235 1.10901413]) to ([0.46585 1.21991554])\nIncrease radii from ([0.46585 1.21991554]) to ([0.512435 1.34190709])\nIncrease radii from ([0.512435 1.34190709]) to ([0.5636785 1.4760978])\nIncrease radii from ([0.5636785 1.4760978]) to ([0.62004635 1.62370758])\nIncrease radii from ([0.3 1.60543876]) to ([0.33 1.76598264])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.7856553 0.7856553]) to ([0.86422083 0.86422083])\nIncrease radii from ([0.86422083 0.86422083]) to ([0.95064292 0.95064292])\nIncrease radii from ([0.95064292 0.95064292]) to ([1.04570721 1.04570721])\nIncrease radii from ([0.3 0.78560623]) to ([0.33 0.86416685])\nIncrease radii from ([0.33 0.86416685]) to ([0.363 0.95058354])\nIncrease radii from ([0.363 0.95058354]) to ([0.3993 1.04564189])\nIncrease radii from ([0.3993 1.04564189]) to ([0.43923 1.15020608])\nIncrease radii from ([0.43923 1.15020608]) to ([0.483153 1.26522669])\nIncrease radii from ([0.483153 1.26522669]) to ([0.5314683 1.39174936])\nIncrease radii from ([0.5314683 1.39174936]) to ([0.58461513 1.53092429])\nIncrease radii from ([0.58461513 1.53092429]) to ([0.64307664 1.68401672])\nIncrease radii from ([0.3 1.17216167]) to ([0.33 1.28937783])\nIncrease radii from ([0.33 1.28937783]) to ([0.363 1.41831562])\nIncrease radii from ([0.25 1.33786564]) to ([0.275 1.4716522])\nIncrease radii from ([0.275 1.4716522]) to ([0.3025 1.61881742])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.65471275 0.65471275]) to ([0.72018403 0.72018403])\nIncrease radii from ([0.72018403 0.72018403]) to ([0.79220243 0.79220243])\nIncrease radii from ([0.79220243 0.79220243]) to ([0.87142267 0.87142267])\nIncrease radii from ([0.87142267 0.87142267]) to ([0.95856494 0.95856494])\nIncrease radii from ([0.95856494 0.95856494]) to ([1.05442144 1.05442144])\nIncrease radii from ([0.25 0.65467186]) to ([0.275 0.72013904])\nIncrease radii from ([0.275 0.72013904]) to ([0.3025 0.79215295])\nIncrease radii from ([0.3025 0.79215295]) to ([0.33275 0.87136824])\nIncrease radii from ([0.33275 0.87136824]) to ([0.366025 0.95850507])\nIncrease radii from ([0.366025 0.95850507]) to ([0.4026275 1.05435557])\nIncrease radii from ([0.4026275 1.05435557]) to ([0.44289025 1.15979113])\nIncrease radii from ([0.44289025 1.15979113]) to ([0.48717928 1.27577024])\nIncrease radii from ([0.48717928 1.27577024]) to ([0.5358972 1.40334727])\nIncrease radii from ([0.5358972 1.40334727]) to ([0.58948692 1.54368199])\nIncrease radii from ([0.58948692 1.54368199]) to ([0.64843562 1.69805019])\nIncrease radii from ([0.25 0.97680139]) to ([0.275 1.07448153])\nIncrease radii from ([0.275 1.07448153]) to ([0.3025 1.18192968])\nIncrease radii from ([0.3025 1.18192968]) to ([0.33275 1.30012265])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1722.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.59527987]) to ([0.66 1.75480786])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.46233988]) to ([0.605 1.60857387])\nIncrease radii from ([0.605 1.60857387]) to ([0.6655 1.76943126])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.32939989]) to ([0.55 1.46233988])\nIncrease radii from ([0.55 1.46233988]) to ([0.605 1.60857387])\nIncrease radii from ([0.605 1.60857387]) to ([0.6655 1.76943126])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.1964599]) to ([0.495 1.31610589])\nIncrease radii from ([0.495 1.31610589]) to ([0.5445 1.44771648])\nIncrease radii from ([0.5445 1.44771648]) to ([0.59895 1.59248813])\nIncrease radii from ([0.59895 1.59248813]) to ([0.658845 1.75173694])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.06351991]) to ([0.44 1.16987191])\nIncrease radii from ([0.44 1.16987191]) to ([0.484 1.2868591])\nIncrease radii from ([0.484 1.2868591]) to ([0.5324 1.41554501])\nIncrease radii from ([0.5324 1.41554501]) to ([0.58564 1.55709951])\nIncrease radii from ([0.58564 1.55709951]) to ([0.644204 1.71280946])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.91659785 0.91659785]) to ([1.00825764 1.00825764])\nIncrease radii from ([0.35 0.93057992]) to ([0.385 1.02363792])\nIncrease radii from ([0.385 1.02363792]) to ([0.4235 1.12600171])\nIncrease radii from ([0.4235 1.12600171]) to ([0.46585 1.23860188])\nIncrease radii from ([0.46585 1.23860188]) to ([0.512435 1.36246207])\nIncrease radii from ([0.512435 1.36246207]) to ([0.5636785 1.49870827])\nIncrease radii from ([0.5636785 1.49870827]) to ([0.62004635 1.6485791 ])\nIncrease radii from ([0.3 1.60543876]) to ([0.33 1.76598264])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.7856553 0.7856553]) to ([0.86422083 0.86422083])\nIncrease radii from ([0.86422083 0.86422083]) to ([0.95064292 0.95064292])\nIncrease radii from ([0.95064292 0.95064292]) to ([1.04570721 1.04570721])\nIncrease radii from ([0.3 0.79763994]) to ([0.33 0.87740393])\nIncrease radii from ([0.33 0.87740393]) to ([0.363 0.96514432])\nIncrease radii from ([0.363 0.96514432]) to ([0.3993 1.06165875])\nIncrease radii from ([0.3993 1.06165875]) to ([0.43923 1.16782463])\nIncrease radii from ([0.43923 1.16782463]) to ([0.483153 1.28460709])\nIncrease radii from ([0.483153 1.28460709]) to ([0.5314683 1.4130678])\nIncrease radii from ([0.5314683 1.4130678]) to ([0.58461513 1.55437458])\nIncrease radii from ([0.58461513 1.55437458]) to ([0.64307664 1.70981204])\nIncrease radii from ([0.3 1.17216167]) to ([0.33 1.28937783])\nIncrease radii from ([0.33 1.28937783]) to ([0.363 1.41831562])\nIncrease radii from ([0.25 1.33786564]) to ([0.275 1.4716522])\nIncrease radii from ([0.275 1.4716522]) to ([0.3025 1.61881742])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.65471275 0.65471275]) to ([0.72018403 0.72018403])\nIncrease radii from ([0.72018403 0.72018403]) to ([0.79220243 0.79220243])\nIncrease radii from ([0.79220243 0.79220243]) to ([0.87142267 0.87142267])\nIncrease radii from ([0.87142267 0.87142267]) to ([0.95856494 0.95856494])\nIncrease radii from ([0.95856494 0.95856494]) to ([1.05442144 1.05442144])\nIncrease radii from ([0.25 0.66469995]) to ([0.275 0.73116994])\nIncrease radii from ([0.275 0.73116994]) to ([0.3025 0.80428694])\nIncrease radii from ([0.3025 0.80428694]) to ([0.33275 0.88471563])\nIncrease radii from ([0.33275 0.88471563]) to ([0.366025 0.97318719])\nIncrease radii from ([0.366025 0.97318719]) to ([0.4026275 1.07050591])\nIncrease radii from ([0.4026275 1.07050591]) to ([0.44289025 1.1775565 ])\nIncrease radii from ([0.44289025 1.1775565 ]) to ([0.48717928 1.29531215])\nIncrease radii from ([0.48717928 1.29531215]) to ([0.5358972 1.42484337])\nIncrease radii from ([0.5358972 1.42484337]) to ([0.58948692 1.5673277 ])\nIncrease radii from ([0.58948692 1.5673277 ]) to ([0.64843562 1.72406047])\nIncrease radii from ([0.25 0.94159318]) to ([0.275 1.0357525])\nIncrease radii from ([0.275 1.0357525]) to ([0.3025 1.13932775])\nIncrease radii from ([0.25 0.97680139]) to ([0.275 1.07448153])\nIncrease radii from ([0.275 1.07448153]) to ([0.3025 1.18192968])\nIncrease radii from ([0.3025 1.18192968]) to ([0.33275 1.30012265])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1723.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.482217]) to ([0.605 1.6304387])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.34747]) to ([0.55 1.482217])\nIncrease radii from ([0.55 1.482217]) to ([0.605 1.6304387])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.212723]) to ([0.495 1.3339953])\nIncrease radii from ([0.495 1.3339953]) to ([0.5445 1.46739483])\nIncrease radii from ([0.5445 1.46739483]) to ([0.59895 1.61413432])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.077976]) to ([0.44 1.1857736])\nIncrease radii from ([0.44 1.1857736]) to ([0.484 1.30435096])\nIncrease radii from ([0.484 1.30435096]) to ([0.5324 1.43478606])\nIncrease radii from ([0.5324 1.43478606]) to ([0.58564 1.57826466])\nIncrease radii from ([0.58564 1.57826466]) to ([0.644204 1.73609113])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.91659785 0.91659785]) to ([1.00825764 1.00825764])\nIncrease radii from ([0.35 0.943229]) to ([0.385 1.0375519])\nIncrease radii from ([0.385 1.0375519]) to ([0.4235 1.14130709])\nIncrease radii from ([0.4235 1.14130709]) to ([0.46585 1.2554378])\nIncrease radii from ([0.46585 1.2554378]) to ([0.512435 1.38098158])\nIncrease radii from ([0.512435 1.38098158]) to ([0.5636785 1.51907974])\nIncrease radii from ([0.5636785 1.51907974]) to ([0.62004635 1.67098771])\nIncrease radii from ([0.3 1.60543876]) to ([0.33 1.76598264])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.7856553 0.7856553]) to ([0.86422083 0.86422083])\nIncrease radii from ([0.86422083 0.86422083]) to ([0.95064292 0.95064292])\nIncrease radii from ([0.95064292 0.95064292]) to ([1.04570721 1.04570721])\nIncrease radii from ([0.3 0.808482]) to ([0.33 0.8893302])\nIncrease radii from ([0.33 0.8893302]) to ([0.363 0.97826322])\nIncrease radii from ([0.363 0.97826322]) to ([0.3993 1.07608954])\nIncrease radii from ([0.3993 1.07608954]) to ([0.43923 1.1836985])\nIncrease radii from ([0.43923 1.1836985]) to ([0.483153 1.30206835])\nIncrease radii from ([0.483153 1.30206835]) to ([0.5314683 1.43227518])\nIncrease radii from ([0.5314683 1.43227518]) to ([0.58461513 1.5755027 ])\nIncrease radii from ([0.58461513 1.5755027 ]) to ([0.64307664 1.73305297])\nIncrease radii from ([0.25 1.33786564]) to ([0.275 1.4716522])\nIncrease radii from ([0.275 1.4716522]) to ([0.3025 1.61881742])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.65471275 0.65471275]) to ([0.72018403 0.72018403])\nIncrease radii from ([0.72018403 0.72018403]) to ([0.79220243 0.79220243])\nIncrease radii from ([0.79220243 0.79220243]) to ([0.87142267 0.87142267])\nIncrease radii from ([0.87142267 0.87142267]) to ([0.95856494 0.95856494])\nIncrease radii from ([0.95856494 0.95856494]) to ([1.05442144 1.05442144])\nIncrease radii from ([0.43997668 0.76206196]) to ([0.48397434 0.83826815])\nIncrease radii from ([0.25 0.92662222]) to ([0.275 1.01928444])\nIncrease radii from ([0.275 1.01928444]) to ([0.3025 1.12121288])\nIncrease radii from ([0.25 0.673735]) to ([0.275 0.7411085])\nIncrease radii from ([0.275 0.7411085]) to ([0.3025 0.81521935])\nIncrease radii from ([0.3025 0.81521935]) to ([0.33275 0.89674129])\nIncrease radii from ([0.33275 0.89674129]) to ([0.366025 0.98641541])\nIncrease radii from ([0.366025 0.98641541]) to ([0.4026275 1.08505696])\nIncrease radii from ([0.4026275 1.08505696]) to ([0.44289025 1.19356265])\nIncrease radii from ([0.44289025 1.19356265]) to ([0.48717928 1.31291892])\nIncrease radii from ([0.48717928 1.31291892]) to ([0.5358972 1.44421081])\nIncrease radii from ([0.5358972 1.44421081]) to ([0.58948692 1.58863189])\nIncrease radii from ([0.58948692 1.58863189]) to ([0.64843562 1.74749508])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1724.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.50390612]) to ([0.605 1.65429673])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.482217]) to ([0.605 1.6304387])\nIncrease radii from ([0.5 1.36718738]) to ([0.55 1.50390612])\nIncrease radii from ([0.55 1.50390612]) to ([0.605 1.65429673])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.34747]) to ([0.55 1.482217])\nIncrease radii from ([0.55 1.482217]) to ([0.605 1.6304387])\nIncrease radii from ([0.45 1.23046864]) to ([0.495 1.35351551])\nIncrease radii from ([0.495 1.35351551]) to ([0.5445 1.48886706])\nIncrease radii from ([0.5445 1.48886706]) to ([0.59895 1.63775377])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.212723]) to ([0.495 1.3339953])\nIncrease radii from ([0.495 1.3339953]) to ([0.5445 1.46739483])\nIncrease radii from ([0.5445 1.46739483]) to ([0.59895 1.61413432])\nIncrease radii from ([0.4 1.09374991]) to ([0.44 1.2031249])\nIncrease radii from ([0.44 1.2031249]) to ([0.484 1.32343739])\nIncrease radii from ([0.484 1.32343739]) to ([0.5324 1.45578113])\nIncrease radii from ([0.5324 1.45578113]) to ([0.58564 1.60135924])\nIncrease radii from ([0.58564 1.60135924]) to ([0.644204 1.76149516])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.077976]) to ([0.44 1.1857736])\nIncrease radii from ([0.44 1.1857736]) to ([0.484 1.30435096])\nIncrease radii from ([0.484 1.30435096]) to ([0.5324 1.43478606])\nIncrease radii from ([0.5324 1.43478606]) to ([0.58564 1.57826466])\nIncrease radii from ([0.58564 1.57826466]) to ([0.644204 1.73609113])\nIncrease radii from ([0.35 0.95703117]) to ([0.385 1.05273428])\nIncrease radii from ([0.385 1.05273428]) to ([0.4235 1.15800771])\nIncrease radii from ([0.4235 1.15800771]) to ([0.46585 1.27380848])\nIncrease radii from ([0.46585 1.27380848]) to ([0.512435 1.40118933])\nIncrease radii from ([0.512435 1.40118933]) to ([0.5636785 1.54130827])\nIncrease radii from ([0.5636785 1.54130827]) to ([0.62004635 1.69543909])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.943229]) to ([0.385 1.0375519])\nIncrease radii from ([0.385 1.0375519]) to ([0.4235 1.14130709])\nIncrease radii from ([0.4235 1.14130709]) to ([0.46585 1.2554378])\nIncrease radii from ([0.46585 1.2554378]) to ([0.512435 1.38098158])\nIncrease radii from ([0.512435 1.38098158]) to ([0.5636785 1.51907974])\nIncrease radii from ([0.5636785 1.51907974]) to ([0.62004635 1.67098771])\nIncrease radii from ([0.3 0.82031243]) to ([0.33 0.90234367])\nIncrease radii from ([0.33 0.90234367]) to ([0.363 0.99257804])\nIncrease radii from ([0.363 0.99257804]) to ([0.3993 1.09183584])\nIncrease radii from ([0.3993 1.09183584]) to ([0.43923 1.20101943])\nIncrease radii from ([0.43923 1.20101943]) to ([0.483153 1.32112137])\nIncrease radii from ([0.483153 1.32112137]) to ([0.5314683 1.45323351])\nIncrease radii from ([0.5314683 1.45323351]) to ([0.58461513 1.59855686])\nIncrease radii from ([0.58461513 1.59855686]) to ([0.64307664 1.75841255])\nIncrease radii from ([0.3 1.60543876]) to ([0.33 1.76598264])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.808482]) to ([0.33 0.8893302])\nIncrease radii from ([0.33 0.8893302]) to ([0.363 0.97826322])\nIncrease radii from ([0.363 0.97826322]) to ([0.3993 1.07608954])\nIncrease radii from ([0.3993 1.07608954]) to ([0.43923 1.1836985])\nIncrease radii from ([0.43923 1.1836985]) to ([0.483153 1.30206835])\nIncrease radii from ([0.483153 1.30206835]) to ([0.5314683 1.43227518])\nIncrease radii from ([0.5314683 1.43227518]) to ([0.58461513 1.5755027 ])\nIncrease radii from ([0.58461513 1.5755027 ]) to ([0.64307664 1.73305297])\nIncrease radii from ([0.25 0.68359369]) to ([0.275 0.75195306])\nIncrease radii from ([0.275 0.75195306]) to ([0.3025 0.82714837])\nIncrease radii from ([0.3025 0.82714837]) to ([0.33275 0.9098632])\nIncrease radii from ([0.33275 0.9098632]) to ([0.366025 1.00084952])\nIncrease radii from ([0.366025 1.00084952]) to ([0.4026275 1.10093448])\nIncrease radii from ([0.4026275 1.10093448]) to ([0.44289025 1.21102792])\nIncrease radii from ([0.44289025 1.21102792]) to ([0.48717928 1.33213072])\nIncrease radii from ([0.48717928 1.33213072]) to ([0.5358972 1.46534379])\nIncrease radii from ([0.5358972 1.46534379]) to ([0.58948692 1.61187817])\nIncrease radii from ([0.58948692 1.61187817]) to ([0.64843562 1.77306598])\nIncrease radii from ([0.25 1.33786564]) to ([0.275 1.4716522])\nIncrease radii from ([0.275 1.4716522]) to ([0.3025 1.61881742])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.92662222]) to ([0.275 1.01928444])\nIncrease radii from ([0.275 1.01928444]) to ([0.3025 1.12121288])\nIncrease radii from ([0.25 0.93644501]) to ([0.275 1.03008951])\nIncrease radii from ([0.275 1.03008951]) to ([0.3025 1.13309846])\nIncrease radii from ([0.25 0.673735]) to ([0.275 0.7411085])\nIncrease radii from ([0.275 0.7411085]) to ([0.3025 0.81521935])\nIncrease radii from ([0.3025 0.81521935]) to ([0.33275 0.89674129])\nIncrease radii from ([0.33275 0.89674129]) to ([0.366025 0.98641541])\nIncrease radii from ([0.366025 0.98641541]) to ([0.4026275 1.08505696])\nIncrease radii from ([0.4026275 1.08505696]) to ([0.44289025 1.19356265])\nIncrease radii from ([0.44289025 1.19356265]) to ([0.48717928 1.31291892])\nIncrease radii from ([0.48717928 1.31291892]) to ([0.5358972 1.44421081])\nIncrease radii from ([0.5358972 1.44421081]) to ([0.58948692 1.58863189])\nIncrease radii from ([0.58948692 1.58863189]) to ([0.64843562 1.74749508])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1725.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.3 1.60543876]) to ([0.33 1.76598264])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 1.14147586]) to ([0.33 1.25562344])\nIncrease radii from ([0.33 1.25562344]) to ([0.363 1.38118579])\nIncrease radii from ([0.25 1.33786564]) to ([0.275 1.4716522])\nIncrease radii from ([0.275 1.4716522]) to ([0.3025 1.61881742])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.95122988]) to ([0.275 1.04635287])\nIncrease radii from ([0.275 1.04635287]) to ([0.3025 1.15098816])\nIncrease radii from ([0.3025 1.15098816]) to ([0.33275 1.26608697])\nIncrease radii from ([0.33275 1.26608697]) to ([0.366025 1.39269567])\nIncrease radii from ([0.25 0.9311253]) to ([0.275 1.02423783])\nIncrease radii from ([0.275 1.02423783]) to ([0.3025 1.12666161])\nIncrease radii from ([0.25 0.93371552]) to ([0.275 1.02708708])\nIncrease radii from ([0.275 1.02708708]) to ([0.3025 1.12979578])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1726.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.48114052]) to ([0.605 1.62925458])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.34649138]) to ([0.55 1.48114052])\nIncrease radii from ([0.55 1.48114052]) to ([0.605 1.62925458])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.45 1.21184225]) to ([0.495 1.33302647])\nIncrease radii from ([0.495 1.33302647]) to ([0.5445 1.46632912])\nIncrease radii from ([0.5445 1.46632912]) to ([0.59895 1.61296203])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.4 1.07719311]) to ([0.44 1.18491242])\nIncrease radii from ([0.44 1.18491242]) to ([0.484 1.30340366])\nIncrease radii from ([0.484 1.30340366]) to ([0.5324 1.43374403])\nIncrease radii from ([0.5324 1.43374403]) to ([0.58564 1.57711843])\nIncrease radii from ([0.58564 1.57711843]) to ([0.644204 1.73483027])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.35 0.94254397]) to ([0.385 1.03679837])\nIncrease radii from ([0.385 1.03679837]) to ([0.4235 1.1404782])\nIncrease radii from ([0.4235 1.1404782]) to ([0.46585 1.25452602])\nIncrease radii from ([0.46585 1.25452602]) to ([0.512435 1.37997863])\nIncrease radii from ([0.512435 1.37997863]) to ([0.5636785 1.51797649])\nIncrease radii from ([0.5636785 1.51797649]) to ([0.62004635 1.66977414])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.3 0.80789483]) to ([0.33 0.88868431])\nIncrease radii from ([0.33 0.88868431]) to ([0.363 0.97755275])\nIncrease radii from ([0.363 0.97755275]) to ([0.3993 1.07530802])\nIncrease radii from ([0.3993 1.07530802]) to ([0.43923 1.18283882])\nIncrease radii from ([0.43923 1.18283882]) to ([0.483153 1.3011227])\nIncrease radii from ([0.483153 1.3011227]) to ([0.5314683 1.43123497])\nIncrease radii from ([0.5314683 1.43123497]) to ([0.58461513 1.57435847])\nIncrease radii from ([0.58461513 1.57435847]) to ([0.64307664 1.73179432])\nIncrease radii from ([0.3 1.60543876]) to ([0.33 1.76598264])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 1.14147586]) to ([0.33 1.25562344])\nIncrease radii from ([0.33 1.25562344]) to ([0.363 1.38118579])\nIncrease radii from ([0.25 0.67324569]) to ([0.275 0.74057026])\nIncrease radii from ([0.275 0.74057026]) to ([0.3025 0.81462729])\nIncrease radii from ([0.3025 0.81462729]) to ([0.33275 0.89609002])\nIncrease radii from ([0.33275 0.89609002]) to ([0.366025 0.98569902])\nIncrease radii from ([0.366025 0.98569902]) to ([0.4026275 1.08426892])\nIncrease radii from ([0.4026275 1.08426892]) to ([0.44289025 1.19269581])\nIncrease radii from ([0.44289025 1.19269581]) to ([0.48717928 1.31196539])\nIncrease radii from ([0.48717928 1.31196539]) to ([0.5358972 1.44316193])\nIncrease radii from ([0.5358972 1.44316193]) to ([0.58948692 1.58747813])\nIncrease radii from ([0.58948692 1.58747813]) to ([0.64843562 1.74622594])\nIncrease radii from ([0.25 1.33786564]) to ([0.275 1.4716522])\nIncrease radii from ([0.275 1.4716522]) to ([0.3025 1.61881742])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.95122988]) to ([0.275 1.04635287])\nIncrease radii from ([0.275 1.04635287]) to ([0.3025 1.15098816])\nIncrease radii from ([0.3025 1.15098816]) to ([0.33275 1.26608697])\nIncrease radii from ([0.33275 1.26608697]) to ([0.366025 1.39269567])\nIncrease radii from ([0.25 0.9311253]) to ([0.275 1.02423783])\nIncrease radii from ([0.275 1.02423783]) to ([0.3025 1.12666161])\nIncrease radii from ([0.25 0.93371552]) to ([0.275 1.02708708])\nIncrease radii from ([0.275 1.02708708]) to ([0.3025 1.12979578])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1727.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.58066001]) to ([0.66 1.73872601])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.49183998]) to ([0.605 1.64102397])\nIncrease radii from ([0.55 1.44893834]) to ([0.605 1.59383217])\nIncrease radii from ([0.605 1.59383217]) to ([0.6655 1.75321539])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.35621816]) to ([0.55 1.49183998])\nIncrease radii from ([0.55 1.49183998]) to ([0.605 1.64102397])\nIncrease radii from ([0.5 1.31721667]) to ([0.55 1.44893834])\nIncrease radii from ([0.55 1.44893834]) to ([0.605 1.59383217])\nIncrease radii from ([0.605 1.59383217]) to ([0.6655 1.75321539])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.22059634]) to ([0.495 1.34265598])\nIncrease radii from ([0.495 1.34265598]) to ([0.5445 1.47692158])\nIncrease radii from ([0.5445 1.47692158]) to ([0.59895 1.62461373])\nIncrease radii from ([0.45 1.18549501]) to ([0.495 1.30404451])\nIncrease radii from ([0.495 1.30404451]) to ([0.5445 1.43444896])\nIncrease radii from ([0.5445 1.43444896]) to ([0.59895 1.57789385])\nIncrease radii from ([0.59895 1.57789385]) to ([0.658845 1.73568324])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.08497453]) to ([0.44 1.19347198])\nIncrease radii from ([0.44 1.19347198]) to ([0.484 1.31281918])\nIncrease radii from ([0.484 1.31281918]) to ([0.5324 1.4441011])\nIncrease radii from ([0.5324 1.4441011]) to ([0.58564 1.58851121])\nIncrease radii from ([0.58564 1.58851121]) to ([0.644204 1.74736233])\nIncrease radii from ([0.4 1.05377334]) to ([0.44 1.15915067])\nIncrease radii from ([0.44 1.15915067]) to ([0.484 1.27506574])\nIncrease radii from ([0.484 1.27506574]) to ([0.5324 1.40257231])\nIncrease radii from ([0.5324 1.40257231]) to ([0.58564 1.54282955])\nIncrease radii from ([0.58564 1.54282955]) to ([0.644204 1.6971125])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.94935271]) to ([0.385 1.04428798])\nIncrease radii from ([0.385 1.04428798]) to ([0.4235 1.14871678])\nIncrease radii from ([0.4235 1.14871678]) to ([0.46585 1.26358846])\nIncrease radii from ([0.46585 1.26358846]) to ([0.512435 1.38994731])\nIncrease radii from ([0.512435 1.38994731]) to ([0.5636785 1.52894204])\nIncrease radii from ([0.5636785 1.52894204]) to ([0.62004635 1.68183624])\nIncrease radii from ([0.35 0.92205167]) to ([0.385 1.01425684])\nIncrease radii from ([0.385 1.01425684]) to ([0.4235 1.11568252])\nIncrease radii from ([0.4235 1.11568252]) to ([0.46585 1.22725077])\nIncrease radii from ([0.46585 1.22725077]) to ([0.512435 1.34997585])\nIncrease radii from ([0.512435 1.34997585]) to ([0.5636785 1.48497344])\nIncrease radii from ([0.5636785 1.48497344]) to ([0.62004635 1.63347078])\nIncrease radii from ([0.3 1.60543876]) to ([0.33 1.76598264])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 1.13586873]) to ([0.33 1.2494556])\nIncrease radii from ([0.33 1.2494556]) to ([0.363 1.37440116])\nIncrease radii from ([0.3 0.8137309]) to ([0.33 0.89510399])\nIncrease radii from ([0.33 0.89510399]) to ([0.363 0.98461438])\nIncrease radii from ([0.363 0.98461438]) to ([0.3993 1.08307582])\nIncrease radii from ([0.3993 1.08307582]) to ([0.43923 1.19138341])\nIncrease radii from ([0.43923 1.19138341]) to ([0.483153 1.31052175])\nIncrease radii from ([0.483153 1.31052175]) to ([0.5314683 1.44157392])\nIncrease radii from ([0.5314683 1.44157392]) to ([0.58461513 1.58573131])\nIncrease radii from ([0.58461513 1.58573131]) to ([0.64307664 1.74430444])\nIncrease radii from ([0.3 0.79033]) to ([0.33 0.869363])\nIncrease radii from ([0.33 0.869363]) to ([0.363 0.9562993])\nIncrease radii from ([0.363 0.9562993]) to ([0.3993 1.05192924])\nIncrease radii from ([0.3993 1.05192924]) to ([0.43923 1.15712216])\nIncrease radii from ([0.43923 1.15712216]) to ([0.483153 1.27283437])\nIncrease radii from ([0.483153 1.27283437]) to ([0.5314683 1.40011781])\nIncrease radii from ([0.5314683 1.40011781]) to ([0.58461513 1.54012959])\nIncrease radii from ([0.58461513 1.54012959]) to ([0.64307664 1.69414255])\nIncrease radii from ([0.25 1.33786564]) to ([0.275 1.4716522])\nIncrease radii from ([0.275 1.4716522]) to ([0.3025 1.61881742])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.94655727]) to ([0.275 1.041213])\nIncrease radii from ([0.275 1.041213]) to ([0.3025 1.1453343])\nIncrease radii from ([0.3025 1.1453343]) to ([0.33275 1.25986773])\nIncrease radii from ([0.33275 1.25986773]) to ([0.366025 1.3858545])\nIncrease radii from ([0.25 0.9311253]) to ([0.275 1.02423783])\nIncrease radii from ([0.275 1.02423783]) to ([0.3025 1.12666161])\nIncrease radii from ([0.25 0.67810908]) to ([0.275 0.74591999])\nIncrease radii from ([0.275 0.74591999]) to ([0.3025 0.82051199])\nIncrease radii from ([0.3025 0.82051199]) to ([0.33275 0.90256319])\nIncrease radii from ([0.33275 0.90256319]) to ([0.366025 0.9928195])\nIncrease radii from ([0.366025 0.9928195]) to ([0.4026275 1.09210145])\nIncrease radii from ([0.4026275 1.09210145]) to ([0.44289025 1.2013116 ])\nIncrease radii from ([0.44289025 1.2013116 ]) to ([0.48717928 1.32144276])\nIncrease radii from ([0.48717928 1.32144276]) to ([0.5358972 1.45358704])\nIncrease radii from ([0.5358972 1.45358704]) to ([0.58948692 1.59894574])\nIncrease radii from ([0.58948692 1.59894574]) to ([0.64843562 1.75884031])\nIncrease radii from ([0.4430987 0.76746946]) to ([0.48740857 0.8442164 ])\nIncrease radii from ([0.25 0.65860834]) to ([0.275 0.72446917])\nIncrease radii from ([0.275 0.72446917]) to ([0.3025 0.79691609])\nIncrease radii from ([0.3025 0.79691609]) to ([0.33275 0.8766077])\nIncrease radii from ([0.33275 0.8766077]) to ([0.366025 0.96426847])\nIncrease radii from ([0.366025 0.96426847]) to ([0.4026275 1.06069531])\nIncrease radii from ([0.4026275 1.06069531]) to ([0.44289025 1.16676484])\nIncrease radii from ([0.44289025 1.16676484]) to ([0.48717928 1.28344133])\nIncrease radii from ([0.48717928 1.28344133]) to ([0.5358972 1.41178546])\nIncrease radii from ([0.5358972 1.41178546]) to ([0.58948692 1.55296401])\nIncrease radii from ([0.58948692 1.55296401]) to ([0.64843562 1.70826041])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1728.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.49400731]) to ([0.605 1.64340804])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.5000638]) to ([0.605 1.65007018])\nIncrease radii from ([0.5 1.35818847]) to ([0.55 1.49400731])\nIncrease radii from ([0.55 1.49400731]) to ([0.605 1.64340804])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.36369437]) to ([0.55 1.5000638])\nIncrease radii from ([0.55 1.5000638]) to ([0.605 1.65007018])\nIncrease radii from ([0.45 1.22236962]) to ([0.495 1.34460658])\nIncrease radii from ([0.495 1.34460658]) to ([0.5445 1.47906724])\nIncrease radii from ([0.5445 1.47906724]) to ([0.59895 1.62697396])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.22732493]) to ([0.495 1.35005742])\nIncrease radii from ([0.495 1.35005742]) to ([0.5445 1.48506317])\nIncrease radii from ([0.5445 1.48506317]) to ([0.59895 1.63356948])\nIncrease radii from ([0.4 1.08655077]) to ([0.44 1.19520585])\nIncrease radii from ([0.44 1.19520585]) to ([0.484 1.31472644])\nIncrease radii from ([0.484 1.31472644]) to ([0.5324 1.44619908])\nIncrease radii from ([0.5324 1.44619908]) to ([0.58564 1.59081899])\nIncrease radii from ([0.58564 1.59081899]) to ([0.644204 1.74990089])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.09095549]) to ([0.44 1.20005104])\nIncrease radii from ([0.44 1.20005104]) to ([0.484 1.32005615])\nIncrease radii from ([0.484 1.32005615]) to ([0.5324 1.45206176])\nIncrease radii from ([0.5324 1.45206176]) to ([0.58564 1.59726794])\nIncrease radii from ([0.58564 1.59726794]) to ([0.644204 1.75699473])\nIncrease radii from ([0.35 0.95073193]) to ([0.385 1.04580512])\nIncrease radii from ([0.385 1.04580512]) to ([0.4235 1.15038563])\nIncrease radii from ([0.4235 1.15038563]) to ([0.46585 1.26542419])\nIncrease radii from ([0.46585 1.26542419]) to ([0.512435 1.39196661])\nIncrease radii from ([0.512435 1.39196661]) to ([0.5636785 1.53116328])\nIncrease radii from ([0.5636785 1.53116328]) to ([0.62004635 1.6842796 ])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.95458606]) to ([0.385 1.05004466])\nIncrease radii from ([0.385 1.05004466]) to ([0.4235 1.15504913])\nIncrease radii from ([0.4235 1.15504913]) to ([0.46585 1.27055404])\nIncrease radii from ([0.46585 1.27055404]) to ([0.512435 1.39760945])\nIncrease radii from ([0.512435 1.39760945]) to ([0.5636785 1.53737039])\nIncrease radii from ([0.5636785 1.53737039]) to ([0.62004635 1.69110743])\nIncrease radii from ([0.3 0.81491308]) to ([0.33 0.89640439])\nIncrease radii from ([0.33 0.89640439]) to ([0.363 0.98604483])\nIncrease radii from ([0.363 0.98604483]) to ([0.3993 1.08464931])\nIncrease radii from ([0.3993 1.08464931]) to ([0.43923 1.19311424])\nIncrease radii from ([0.43923 1.19311424]) to ([0.483153 1.31242566])\nIncrease radii from ([0.483153 1.31242566]) to ([0.5314683 1.44366823])\nIncrease radii from ([0.5314683 1.44366823]) to ([0.58461513 1.58803505])\nIncrease radii from ([0.58461513 1.58803505]) to ([0.64307664 1.74683856])\nIncrease radii from ([0.3 1.5994187]) to ([0.33 1.75936057])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 1.13586873]) to ([0.33 1.2494556])\nIncrease radii from ([0.33 1.2494556]) to ([0.363 1.37440116])\nIncrease radii from ([0.3 0.81821662]) to ([0.33 0.90003828])\nIncrease radii from ([0.33 0.90003828]) to ([0.363 0.99004211])\nIncrease radii from ([0.363 0.99004211]) to ([0.3993 1.08904632])\nIncrease radii from ([0.3993 1.08904632]) to ([0.43923 1.19795095])\nIncrease radii from ([0.43923 1.19795095]) to ([0.483153 1.31774605])\nIncrease radii from ([0.483153 1.31774605]) to ([0.5314683 1.44952065])\nIncrease radii from ([0.5314683 1.44952065]) to ([0.58461513 1.59447272])\nIncrease radii from ([0.58461513 1.59447272]) to ([0.64307664 1.75391999])\nIncrease radii from ([0.25 0.67909423]) to ([0.275 0.74700366])\nIncrease radii from ([0.275 0.74700366]) to ([0.3025 0.82170402])\nIncrease radii from ([0.3025 0.82170402]) to ([0.33275 0.90387442])\nIncrease radii from ([0.33275 0.90387442]) to ([0.366025 0.99426187])\nIncrease radii from ([0.366025 0.99426187]) to ([0.4026275 1.09368805])\nIncrease radii from ([0.4026275 1.09368805]) to ([0.44289025 1.20305686])\nIncrease radii from ([0.44289025 1.20305686]) to ([0.48717928 1.32336255])\nIncrease radii from ([0.48717928 1.32336255]) to ([0.5358972 1.4556988])\nIncrease radii from ([0.5358972 1.4556988]) to ([0.58948692 1.60126868])\nIncrease radii from ([0.58948692 1.60126868]) to ([0.64843562 1.76139555])\nIncrease radii from ([0.25 1.33284892]) to ([0.275 1.46613381])\nIncrease radii from ([0.275 1.46613381]) to ([0.3025 1.61274719])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.94655727]) to ([0.275 1.041213])\nIncrease radii from ([0.275 1.041213]) to ([0.3025 1.1453343])\nIncrease radii from ([0.3025 1.1453343]) to ([0.33275 1.25986773])\nIncrease radii from ([0.33275 1.25986773]) to ([0.366025 1.3858545])\nIncrease radii from ([0.25 0.9311253]) to ([0.275 1.02423783])\nIncrease radii from ([0.275 1.02423783]) to ([0.3025 1.12666161])\nIncrease radii from ([0.25 0.68184718]) to ([0.275 0.7500319])\nIncrease radii from ([0.275 0.7500319]) to ([0.3025 0.82503509])\nIncrease radii from ([0.3025 0.82503509]) to ([0.33275 0.9075386])\nIncrease radii from ([0.33275 0.9075386]) to ([0.366025 0.99829246])\nIncrease radii from ([0.366025 0.99829246]) to ([0.4026275 1.09812171])\nIncrease radii from ([0.4026275 1.09812171]) to ([0.44289025 1.20793388])\nIncrease radii from ([0.44289025 1.20793388]) to ([0.48717928 1.32872727])\nIncrease radii from ([0.48717928 1.32872727]) to ([0.5358972 1.46159999])\nIncrease radii from ([0.5358972 1.46159999]) to ([0.58948692 1.60775999])\nIncrease radii from ([0.58948692 1.60775999]) to ([0.64843562 1.76853599])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1729.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.4981409]) to ([0.605 1.64795499])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.5000638]) to ([0.605 1.65007018])\nIncrease radii from ([0.5 1.36194627]) to ([0.55 1.4981409])\nIncrease radii from ([0.55 1.4981409]) to ([0.605 1.64795499])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.36369437]) to ([0.55 1.5000638])\nIncrease radii from ([0.55 1.5000638]) to ([0.605 1.65007018])\nIncrease radii from ([0.45 1.22575165]) to ([0.495 1.34832681])\nIncrease radii from ([0.495 1.34832681]) to ([0.5445 1.48315949])\nIncrease radii from ([0.5445 1.48315949]) to ([0.59895 1.63147544])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.22732493]) to ([0.495 1.35005742])\nIncrease radii from ([0.495 1.35005742]) to ([0.5445 1.48506317])\nIncrease radii from ([0.5445 1.48506317]) to ([0.59895 1.63356948])\nIncrease radii from ([0.4 1.08955702]) to ([0.44 1.19851272])\nIncrease radii from ([0.44 1.19851272]) to ([0.484 1.31836399])\nIncrease radii from ([0.484 1.31836399]) to ([0.5324 1.45020039])\nIncrease radii from ([0.5324 1.45020039]) to ([0.58564 1.59522043])\nIncrease radii from ([0.58564 1.59522043]) to ([0.644204 1.75474247])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.09095549]) to ([0.44 1.20005104])\nIncrease radii from ([0.44 1.20005104]) to ([0.484 1.32005615])\nIncrease radii from ([0.484 1.32005615]) to ([0.5324 1.45206176])\nIncrease radii from ([0.5324 1.45206176]) to ([0.58564 1.59726794])\nIncrease radii from ([0.58564 1.59726794]) to ([0.644204 1.75699473])\nIncrease radii from ([0.35 0.95336239]) to ([0.385 1.04869863])\nIncrease radii from ([0.385 1.04869863]) to ([0.4235 1.15356849])\nIncrease radii from ([0.4235 1.15356849]) to ([0.46585 1.26892534])\nIncrease radii from ([0.46585 1.26892534]) to ([0.512435 1.39581788])\nIncrease radii from ([0.512435 1.39581788]) to ([0.5636785 1.53539966])\nIncrease radii from ([0.5636785 1.53539966]) to ([0.62004635 1.68893963])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.95458606]) to ([0.385 1.05004466])\nIncrease radii from ([0.385 1.05004466]) to ([0.4235 1.15504913])\nIncrease radii from ([0.4235 1.15504913]) to ([0.46585 1.27055404])\nIncrease radii from ([0.46585 1.27055404]) to ([0.512435 1.39760945])\nIncrease radii from ([0.512435 1.39760945]) to ([0.5636785 1.53737039])\nIncrease radii from ([0.5636785 1.53737039]) to ([0.62004635 1.69110743])\nIncrease radii from ([0.3 0.81716776]) to ([0.33 0.89888454])\nIncrease radii from ([0.33 0.89888454]) to ([0.363 0.98877299])\nIncrease radii from ([0.363 0.98877299]) to ([0.3993 1.08765029])\nIncrease radii from ([0.3993 1.08765029]) to ([0.43923 1.19641532])\nIncrease radii from ([0.43923 1.19641532]) to ([0.483153 1.31605686])\nIncrease radii from ([0.483153 1.31605686]) to ([0.5314683 1.44766254])\nIncrease radii from ([0.5314683 1.44766254]) to ([0.58461513 1.59242879])\nIncrease radii from ([0.58461513 1.59242879]) to ([0.64307664 1.75167167])\nIncrease radii from ([0.3 1.5994187]) to ([0.33 1.75936057])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 1.13586873]) to ([0.33 1.2494556])\nIncrease radii from ([0.33 1.2494556]) to ([0.363 1.37440116])\nIncrease radii from ([0.3 0.81821662]) to ([0.33 0.90003828])\nIncrease radii from ([0.33 0.90003828]) to ([0.363 0.99004211])\nIncrease radii from ([0.363 0.99004211]) to ([0.3993 1.08904632])\nIncrease radii from ([0.3993 1.08904632]) to ([0.43923 1.19795095])\nIncrease radii from ([0.43923 1.19795095]) to ([0.483153 1.31774605])\nIncrease radii from ([0.483153 1.31774605]) to ([0.5314683 1.44952065])\nIncrease radii from ([0.5314683 1.44952065]) to ([0.58461513 1.59447272])\nIncrease radii from ([0.58461513 1.59447272]) to ([0.64307664 1.75391999])\nIncrease radii from ([0.3 1.18299238]) to ([0.33 1.30129162])\nIncrease radii from ([0.25 0.68097314]) to ([0.275 0.74907045])\nIncrease radii from ([0.275 0.74907045]) to ([0.3025 0.8239775])\nIncrease radii from ([0.3025 0.8239775]) to ([0.33275 0.90637524])\nIncrease radii from ([0.33275 0.90637524]) to ([0.366025 0.99701277])\nIncrease radii from ([0.366025 0.99701277]) to ([0.4026275 1.09671405])\nIncrease radii from ([0.4026275 1.09671405]) to ([0.44289025 1.20638545])\nIncrease radii from ([0.44289025 1.20638545]) to ([0.48717928 1.327024 ])\nIncrease radii from ([0.48717928 1.327024 ]) to ([0.5358972 1.4597264])\nIncrease radii from ([0.5358972 1.4597264]) to ([0.58948692 1.60569903])\nIncrease radii from ([0.58948692 1.60569903]) to ([0.64843562 1.76626894])\nIncrease radii from ([0.44352313 0.7682046 ]) to ([0.48787544 0.84502505])\nIncrease radii from ([0.25 1.33284892]) to ([0.275 1.46613381])\nIncrease radii from ([0.275 1.46613381]) to ([0.3025 1.61274719])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.94655727]) to ([0.275 1.041213])\nIncrease radii from ([0.275 1.041213]) to ([0.3025 1.1453343])\nIncrease radii from ([0.3025 1.1453343]) to ([0.33275 1.25986773])\nIncrease radii from ([0.33275 1.25986773]) to ([0.366025 1.3858545])\nIncrease radii from ([0.25 0.9311253]) to ([0.275 1.02423783])\nIncrease radii from ([0.275 1.02423783]) to ([0.3025 1.12666161])\nIncrease radii from ([0.25 0.68184718]) to ([0.275 0.7500319])\nIncrease radii from ([0.275 0.7500319]) to ([0.3025 0.82503509])\nIncrease radii from ([0.3025 0.82503509]) to ([0.33275 0.9075386])\nIncrease radii from ([0.33275 0.9075386]) to ([0.366025 0.99829246])\nIncrease radii from ([0.366025 0.99829246]) to ([0.4026275 1.09812171])\nIncrease radii from ([0.4026275 1.09812171]) to ([0.44289025 1.20793388])\nIncrease radii from ([0.44289025 1.20793388]) to ([0.48717928 1.32872727])\nIncrease radii from ([0.48717928 1.32872727]) to ([0.5358972 1.46159999])\nIncrease radii from ([0.5358972 1.46159999]) to ([0.58948692 1.60775999])\nIncrease radii from ([0.58948692 1.60775999]) to ([0.64843562 1.76853599])\nIncrease radii from ([0.25 0.98582698]) to ([0.275 1.08440968])\nIncrease radii from ([0.275 1.08440968]) to ([0.3025 1.19285065])\nIncrease radii from ([0.3025 1.19285065]) to ([0.33275 1.31213571])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1730.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.56318422]) to ([0.66 1.71950264])\nIncrease radii from ([0.55 1.4831683]) to ([0.605 1.63148513])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.43291887]) to ([0.605 1.57621076])\nIncrease radii from ([0.605 1.57621076]) to ([0.6655 1.73383183])\nIncrease radii from ([0.5 1.34833482]) to ([0.55 1.4831683])\nIncrease radii from ([0.55 1.4831683]) to ([0.605 1.63148513])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.30265352]) to ([0.55 1.43291887])\nIncrease radii from ([0.55 1.43291887]) to ([0.605 1.57621076])\nIncrease radii from ([0.605 1.57621076]) to ([0.6655 1.73383183])\nIncrease radii from ([0.45 1.21350134]) to ([0.495 1.33485147])\nIncrease radii from ([0.495 1.33485147]) to ([0.5445 1.46833662])\nIncrease radii from ([0.5445 1.46833662]) to ([0.59895 1.61517028])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.17238817]) to ([0.495 1.28962698])\nIncrease radii from ([0.495 1.28962698]) to ([0.5445 1.41858968])\nIncrease radii from ([0.5445 1.41858968]) to ([0.59895 1.56044865])\nIncrease radii from ([0.59895 1.56044865]) to ([0.658845 1.71649351])\nIncrease radii from ([0.4 1.07866786]) to ([0.44 1.18653464])\nIncrease radii from ([0.44 1.18653464]) to ([0.484 1.30518811])\nIncrease radii from ([0.484 1.30518811]) to ([0.5324 1.43570692])\nIncrease radii from ([0.5324 1.43570692]) to ([0.58564 1.57927761])\nIncrease radii from ([0.58564 1.57927761]) to ([0.644204 1.73720537])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.04212281]) to ([0.44 1.14633509])\nIncrease radii from ([0.44 1.14633509]) to ([0.484 1.2609686])\nIncrease radii from ([0.484 1.2609686]) to ([0.5324 1.38706546])\nIncrease radii from ([0.5324 1.38706546]) to ([0.58564 1.52577201])\nIncrease radii from ([0.58564 1.52577201]) to ([0.644204 1.67834921])\nIncrease radii from ([0.35 0.94383438]) to ([0.385 1.03821781])\nIncrease radii from ([0.385 1.03821781]) to ([0.4235 1.14203959])\nIncrease radii from ([0.4235 1.14203959]) to ([0.46585 1.25624355])\nIncrease radii from ([0.46585 1.25624355]) to ([0.512435 1.38186791])\nIncrease radii from ([0.512435 1.38186791]) to ([0.5636785 1.5200547])\nIncrease radii from ([0.5636785 1.5200547]) to ([0.62004635 1.67206017])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.91185746]) to ([0.385 1.00304321])\nIncrease radii from ([0.385 1.00304321]) to ([0.4235 1.10334753])\nIncrease radii from ([0.4235 1.10334753]) to ([0.46585 1.21368228])\nIncrease radii from ([0.46585 1.21368228]) to ([0.512435 1.33505051])\nIncrease radii from ([0.512435 1.33505051]) to ([0.5636785 1.46855556])\nIncrease radii from ([0.5636785 1.46855556]) to ([0.62004635 1.61541112])\nIncrease radii from ([0.3 0.80900089]) to ([0.33 0.88990098])\nIncrease radii from ([0.33 0.88990098]) to ([0.363 0.97889108])\nIncrease radii from ([0.363 0.97889108]) to ([0.3993 1.07678019])\nIncrease radii from ([0.3993 1.07678019]) to ([0.43923 1.18445821])\nIncrease radii from ([0.43923 1.18445821]) to ([0.483153 1.30290403])\nIncrease radii from ([0.483153 1.30290403]) to ([0.5314683 1.43319443])\nIncrease radii from ([0.5314683 1.43319443]) to ([0.58461513 1.57651387])\nIncrease radii from ([0.58461513 1.57651387]) to ([0.64307664 1.73416526])\nIncrease radii from ([0.3 1.5994187]) to ([0.33 1.75936057])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.78159211]) to ([0.33 0.85975132])\nIncrease radii from ([0.33 0.85975132]) to ([0.363 0.94572645])\nIncrease radii from ([0.363 0.94572645]) to ([0.3993 1.0402991])\nIncrease radii from ([0.3993 1.0402991]) to ([0.43923 1.14432901])\nIncrease radii from ([0.43923 1.14432901]) to ([0.483153 1.25876191])\nIncrease radii from ([0.483153 1.25876191]) to ([0.5314683 1.3846381])\nIncrease radii from ([0.5314683 1.3846381]) to ([0.58461513 1.52310191])\nIncrease radii from ([0.58461513 1.52310191]) to ([0.64307664 1.6754121 ])\nIncrease radii from ([0.25 0.67416741]) to ([0.275 0.74158415])\nIncrease radii from ([0.275 0.74158415]) to ([0.3025 0.81574257])\nIncrease radii from ([0.3025 0.81574257]) to ([0.33275 0.89731682])\nIncrease radii from ([0.33275 0.89731682]) to ([0.366025 0.98704851])\nIncrease radii from ([0.366025 0.98704851]) to ([0.4026275 1.08575336])\nIncrease radii from ([0.4026275 1.08575336]) to ([0.44289025 1.19432869])\nIncrease radii from ([0.44289025 1.19432869]) to ([0.48717928 1.31376156])\nIncrease radii from ([0.48717928 1.31376156]) to ([0.5358972 1.44513772])\nIncrease radii from ([0.5358972 1.44513772]) to ([0.58948692 1.58965149])\nIncrease radii from ([0.58948692 1.58965149]) to ([0.64843562 1.74861664])\nIncrease radii from ([0.25 1.33284892]) to ([0.275 1.46613381])\nIncrease radii from ([0.275 1.46613381]) to ([0.3025 1.61274719])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.65132676]) to ([0.275 0.71645943])\nIncrease radii from ([0.275 0.71645943]) to ([0.3025 0.78810538])\nIncrease radii from ([0.3025 0.78810538]) to ([0.33275 0.86691592])\nIncrease radii from ([0.33275 0.86691592]) to ([0.366025 0.95360751])\nIncrease radii from ([0.366025 0.95360751]) to ([0.4026275 1.04896826])\nIncrease radii from ([0.4026275 1.04896826]) to ([0.44289025 1.15386508])\nIncrease radii from ([0.44289025 1.15386508]) to ([0.48717928 1.26925159])\nIncrease radii from ([0.48717928 1.26925159]) to ([0.5358972 1.39617675])\nIncrease radii from ([0.5358972 1.39617675]) to ([0.58948692 1.53579443])\nIncrease radii from ([0.58948692 1.53579443]) to ([0.64843562 1.68937387])\nIncrease radii from ([0.25 0.9311253]) to ([0.275 1.02423783])\nIncrease radii from ([0.275 1.02423783]) to ([0.3025 1.12666161])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1731.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.59913252]) to ([0.66 1.75904577])\nIncrease radii from ([0.6 1.56318422]) to ([0.66 1.71950264])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.46587148]) to ([0.605 1.61245863])\nIncrease radii from ([0.605 1.61245863]) to ([0.6655 1.77370449])\nIncrease radii from ([0.55 1.43291887]) to ([0.605 1.57621076])\nIncrease radii from ([0.605 1.57621076]) to ([0.6655 1.73383183])\nIncrease radii from ([0.55 1.49458076]) to ([0.605 1.64403883])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.33261043]) to ([0.55 1.46587148])\nIncrease radii from ([0.55 1.46587148]) to ([0.605 1.61245863])\nIncrease radii from ([0.605 1.61245863]) to ([0.6655 1.77370449])\nIncrease radii from ([0.5 1.30265352]) to ([0.55 1.43291887])\nIncrease radii from ([0.55 1.43291887]) to ([0.605 1.57621076])\nIncrease radii from ([0.605 1.57621076]) to ([0.6655 1.73383183])\nIncrease radii from ([0.5 1.35870978]) to ([0.55 1.49458076])\nIncrease radii from ([0.55 1.49458076]) to ([0.605 1.64403883])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.19934939]) to ([0.495 1.31928433])\nIncrease radii from ([0.495 1.31928433]) to ([0.5445 1.45121276])\nIncrease radii from ([0.5445 1.45121276]) to ([0.59895 1.59633404])\nIncrease radii from ([0.59895 1.59633404]) to ([0.658845 1.75596744])\nIncrease radii from ([0.45 1.17238817]) to ([0.495 1.28962698])\nIncrease radii from ([0.495 1.28962698]) to ([0.5445 1.41858968])\nIncrease radii from ([0.5445 1.41858968]) to ([0.59895 1.56044865])\nIncrease radii from ([0.59895 1.56044865]) to ([0.658845 1.71649351])\nIncrease radii from ([0.45 1.2228388]) to ([0.495 1.34512268])\nIncrease radii from ([0.495 1.34512268]) to ([0.5445 1.47963495])\nIncrease radii from ([0.5445 1.47963495]) to ([0.59895 1.62759845])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.06608835]) to ([0.44 1.17269718])\nIncrease radii from ([0.44 1.17269718]) to ([0.484 1.2899669])\nIncrease radii from ([0.484 1.2899669]) to ([0.5324 1.41896359])\nIncrease radii from ([0.5324 1.41896359]) to ([0.58564 1.56085995])\nIncrease radii from ([0.58564 1.56085995]) to ([0.644204 1.71694594])\nIncrease radii from ([0.4 1.04212281]) to ([0.44 1.14633509])\nIncrease radii from ([0.44 1.14633509]) to ([0.484 1.2609686])\nIncrease radii from ([0.484 1.2609686]) to ([0.5324 1.38706546])\nIncrease radii from ([0.5324 1.38706546]) to ([0.58564 1.52577201])\nIncrease radii from ([0.58564 1.52577201]) to ([0.644204 1.67834921])\nIncrease radii from ([0.4 1.08696782]) to ([0.44 1.19566461])\nIncrease radii from ([0.44 1.19566461]) to ([0.484 1.31523107])\nIncrease radii from ([0.484 1.31523107]) to ([0.5324 1.44675417])\nIncrease radii from ([0.5324 1.44675417]) to ([0.58564 1.59142959])\nIncrease radii from ([0.58564 1.59142959]) to ([0.644204 1.75057255])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.9328273]) to ([0.385 1.02611003])\nIncrease radii from ([0.385 1.02611003]) to ([0.4235 1.12872104])\nIncrease radii from ([0.4235 1.12872104]) to ([0.46585 1.24159314])\nIncrease radii from ([0.46585 1.24159314]) to ([0.512435 1.36575246])\nIncrease radii from ([0.512435 1.36575246]) to ([0.5636785 1.5023277])\nIncrease radii from ([0.5636785 1.5023277]) to ([0.62004635 1.65256047])\nIncrease radii from ([0.35 0.91185746]) to ([0.385 1.00304321])\nIncrease radii from ([0.385 1.00304321]) to ([0.4235 1.10334753])\nIncrease radii from ([0.4235 1.10334753]) to ([0.46585 1.21368228])\nIncrease radii from ([0.46585 1.21368228]) to ([0.512435 1.33505051])\nIncrease radii from ([0.512435 1.33505051]) to ([0.5636785 1.46855556])\nIncrease radii from ([0.5636785 1.46855556]) to ([0.62004635 1.61541112])\nIncrease radii from ([0.35 0.95109685]) to ([0.385 1.04620653])\nIncrease radii from ([0.385 1.04620653]) to ([0.4235 1.15082718])\nIncrease radii from ([0.4235 1.15082718]) to ([0.46585 1.2659099])\nIncrease radii from ([0.46585 1.2659099]) to ([0.512435 1.39250089])\nIncrease radii from ([0.512435 1.39250089]) to ([0.5636785 1.53175098])\nIncrease radii from ([0.5636785 1.53175098]) to ([0.62004635 1.68492608])\nIncrease radii from ([0.3 1.5994187]) to ([0.33 1.75936057])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.79956626]) to ([0.33 0.87952289])\nIncrease radii from ([0.33 0.87952289]) to ([0.363 0.96747518])\nIncrease radii from ([0.363 0.96747518]) to ([0.3993 1.06422269])\nIncrease radii from ([0.3993 1.06422269]) to ([0.43923 1.17064496])\nIncrease radii from ([0.43923 1.17064496]) to ([0.483153 1.28770946])\nIncrease radii from ([0.483153 1.28770946]) to ([0.5314683 1.4164804])\nIncrease radii from ([0.5314683 1.4164804]) to ([0.58461513 1.55812844])\nIncrease radii from ([0.58461513 1.55812844]) to ([0.64307664 1.71394129])\nIncrease radii from ([0.3 0.78159211]) to ([0.33 0.85975132])\nIncrease radii from ([0.33 0.85975132]) to ([0.363 0.94572645])\nIncrease radii from ([0.363 0.94572645]) to ([0.3993 1.0402991])\nIncrease radii from ([0.3993 1.0402991]) to ([0.43923 1.14432901])\nIncrease radii from ([0.43923 1.14432901]) to ([0.483153 1.25876191])\nIncrease radii from ([0.483153 1.25876191]) to ([0.5314683 1.3846381])\nIncrease radii from ([0.5314683 1.3846381]) to ([0.58461513 1.52310191])\nIncrease radii from ([0.58461513 1.52310191]) to ([0.64307664 1.6754121 ])\nIncrease radii from ([0.3 0.81522587]) to ([0.33 0.89674846])\nIncrease radii from ([0.33 0.89674846]) to ([0.363 0.9864233])\nIncrease radii from ([0.363 0.9864233]) to ([0.3993 1.08506563])\nIncrease radii from ([0.3993 1.08506563]) to ([0.43923 1.19357219])\nIncrease radii from ([0.43923 1.19357219]) to ([0.483153 1.31292941])\nIncrease radii from ([0.483153 1.31292941]) to ([0.5314683 1.44422235])\nIncrease radii from ([0.5314683 1.44422235]) to ([0.58461513 1.58864459])\nIncrease radii from ([0.58461513 1.58864459]) to ([0.64307664 1.74750905])\nIncrease radii from ([0.25 1.33284892]) to ([0.275 1.46613381])\nIncrease radii from ([0.275 1.46613381]) to ([0.3025 1.61274719])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.66630522]) to ([0.275 0.73293574])\nIncrease radii from ([0.275 0.73293574]) to ([0.3025 0.80622931])\nIncrease radii from ([0.3025 0.80622931]) to ([0.33275 0.88685224])\nIncrease radii from ([0.33275 0.88685224]) to ([0.366025 0.97553747])\nIncrease radii from ([0.366025 0.97553747]) to ([0.4026275 1.07309122])\nIncrease radii from ([0.4026275 1.07309122]) to ([0.44289025 1.18040034])\nIncrease radii from ([0.44289025 1.18040034]) to ([0.48717928 1.29844037])\nIncrease radii from ([0.48717928 1.29844037]) to ([0.5358972 1.42828441])\nIncrease radii from ([0.5358972 1.42828441]) to ([0.58948692 1.57111285])\nIncrease radii from ([0.58948692 1.57111285]) to ([0.64843562 1.72822413])\nIncrease radii from ([0.25 0.65132676]) to ([0.275 0.71645943])\nIncrease radii from ([0.275 0.71645943]) to ([0.3025 0.78810538])\nIncrease radii from ([0.3025 0.78810538]) to ([0.33275 0.86691592])\nIncrease radii from ([0.33275 0.86691592]) to ([0.366025 0.95360751])\nIncrease radii from ([0.366025 0.95360751]) to ([0.4026275 1.04896826])\nIncrease radii from ([0.4026275 1.04896826]) to ([0.44289025 1.15386508])\nIncrease radii from ([0.44289025 1.15386508]) to ([0.48717928 1.26925159])\nIncrease radii from ([0.48717928 1.26925159]) to ([0.5358972 1.39617675])\nIncrease radii from ([0.5358972 1.39617675]) to ([0.58948692 1.53579443])\nIncrease radii from ([0.58948692 1.53579443]) to ([0.64843562 1.68937387])\nIncrease radii from ([0.25 0.92328611]) to ([0.275 1.01561472])\nIncrease radii from ([0.275 1.01561472]) to ([0.3025 1.11717619])\nIncrease radii from ([0.25 0.67935489]) to ([0.275 0.74729038])\nIncrease radii from ([0.275 0.74729038]) to ([0.3025 0.82201942])\nIncrease radii from ([0.3025 0.82201942]) to ([0.33275 0.90422136])\nIncrease radii from ([0.33275 0.90422136]) to ([0.366025 0.9946435])\nIncrease radii from ([0.366025 0.9946435]) to ([0.4026275 1.09410784])\nIncrease radii from ([0.4026275 1.09410784]) to ([0.44289025 1.20351863])\nIncrease radii from ([0.44289025 1.20351863]) to ([0.48717928 1.32387049])\nIncrease radii from ([0.48717928 1.32387049]) to ([0.5358972 1.45625754])\nIncrease radii from ([0.5358972 1.45625754]) to ([0.58948692 1.6018833 ])\nIncrease radii from ([0.58948692 1.6018833 ]) to ([0.64843562 1.76207162])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1732.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.3 1.55793438]) to ([0.33 1.71372781])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.25 1.29827865]) to ([0.275 1.42810651])\nIncrease radii from ([0.275 1.42810651]) to ([0.3025 1.57091716])\nIncrease radii from ([0.3025 1.57091716]) to ([0.33275 1.72800888])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.92328611]) to ([0.275 1.01561472])\nIncrease radii from ([0.275 1.01561472]) to ([0.3025 1.11717619])\nIncrease radii from ([0.43821948 0.75901841]) to ([0.48204143 0.83492025])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1733.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.3 1.58067357]) to ([0.33 1.73874093])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.25 1.31722798]) to ([0.275 1.44895077])\nIncrease radii from ([0.275 1.44895077]) to ([0.3025 1.59384585])\nIncrease radii from ([0.3025 1.59384585]) to ([0.33275 1.75323044])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.93286532]) to ([0.275 1.02615185])\nIncrease radii from ([0.275 1.02615185]) to ([0.3025 1.12876704])\nIncrease radii from ([0.43821948 0.75901841]) to ([0.48204143 0.83492025])\nIncrease radii from ([0.25 0.96205219]) to ([0.275 1.05825741])\nIncrease radii from ([0.275 1.05825741]) to ([0.3025 1.16408315])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1734.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.57210273]) to ([0.66 1.729313])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.50783841]) to ([0.605 1.65862225])\nIncrease radii from ([0.55 1.44109417]) to ([0.605 1.58520358])\nIncrease radii from ([0.605 1.58520358]) to ([0.6655 1.74372394])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.37076219]) to ([0.55 1.50783841])\nIncrease radii from ([0.55 1.50783841]) to ([0.605 1.65862225])\nIncrease radii from ([0.5 1.31008561]) to ([0.55 1.44109417])\nIncrease radii from ([0.55 1.44109417]) to ([0.605 1.58520358])\nIncrease radii from ([0.605 1.58520358]) to ([0.6655 1.74372394])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.23368597]) to ([0.495 1.35705457])\nIncrease radii from ([0.495 1.35705457]) to ([0.5445 1.49276003])\nIncrease radii from ([0.5445 1.49276003]) to ([0.59895 1.64203603])\nIncrease radii from ([0.45 1.17907705]) to ([0.495 1.29698475])\nIncrease radii from ([0.495 1.29698475]) to ([0.5445 1.42668323])\nIncrease radii from ([0.5445 1.42668323]) to ([0.59895 1.56935155])\nIncrease radii from ([0.59895 1.56935155]) to ([0.658845 1.7262867])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.09660975]) to ([0.44 1.20627073])\nIncrease radii from ([0.44 1.20627073]) to ([0.484 1.3268978])\nIncrease radii from ([0.484 1.3268978]) to ([0.5324 1.45958758])\nIncrease radii from ([0.5324 1.45958758]) to ([0.58564 1.60554634])\nIncrease radii from ([0.58564 1.60554634]) to ([0.644204 1.76610097])\nIncrease radii from ([0.4 1.04806849]) to ([0.44 1.15287533])\nIncrease radii from ([0.44 1.15287533]) to ([0.484 1.26816287])\nIncrease radii from ([0.484 1.26816287]) to ([0.5324 1.39497915])\nIncrease radii from ([0.5324 1.39497915]) to ([0.58564 1.53447707])\nIncrease radii from ([0.58564 1.53447707]) to ([0.644204 1.68792478])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.95953353]) to ([0.385 1.05548689])\nIncrease radii from ([0.385 1.05548689]) to ([0.4235 1.16103558])\nIncrease radii from ([0.4235 1.16103558]) to ([0.46585 1.27713913])\nIncrease radii from ([0.46585 1.27713913]) to ([0.512435 1.40485305])\nIncrease radii from ([0.512435 1.40485305]) to ([0.5636785 1.54533835])\nIncrease radii from ([0.5636785 1.54533835]) to ([0.62004635 1.69987219])\nIncrease radii from ([0.35 0.91705993]) to ([0.385 1.00876592])\nIncrease radii from ([0.385 1.00876592]) to ([0.4235 1.10964251])\nIncrease radii from ([0.4235 1.10964251]) to ([0.46585 1.22060676])\nIncrease radii from ([0.46585 1.22060676]) to ([0.512435 1.34266744])\nIncrease radii from ([0.512435 1.34266744]) to ([0.5636785 1.47693418])\nIncrease radii from ([0.5636785 1.47693418]) to ([0.62004635 1.6246276 ])\nIncrease radii from ([0.3 1.58067357]) to ([0.33 1.73874093])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.82245731]) to ([0.33 0.90470305])\nIncrease radii from ([0.33 0.90470305]) to ([0.363 0.99517335])\nIncrease radii from ([0.363 0.99517335]) to ([0.3993 1.09469069])\nIncrease radii from ([0.3993 1.09469069]) to ([0.43923 1.20415975])\nIncrease radii from ([0.43923 1.20415975]) to ([0.483153 1.32457573])\nIncrease radii from ([0.483153 1.32457573]) to ([0.5314683 1.4570333])\nIncrease radii from ([0.5314683 1.4570333]) to ([0.58461513 1.60273663])\nIncrease radii from ([0.58461513 1.60273663]) to ([0.64307664 1.7630103 ])\nIncrease radii from ([0.3 0.78605136]) to ([0.33 0.8646565])\nIncrease radii from ([0.33 0.8646565]) to ([0.363 0.95112215])\nIncrease radii from ([0.363 0.95112215]) to ([0.3993 1.04623437])\nIncrease radii from ([0.3993 1.04623437]) to ([0.43923 1.1508578])\nIncrease radii from ([0.43923 1.1508578]) to ([0.483153 1.26594358])\nIncrease radii from ([0.483153 1.26594358]) to ([0.5314683 1.39253794])\nIncrease radii from ([0.5314683 1.39253794]) to ([0.58461513 1.53179174])\nIncrease radii from ([0.58461513 1.53179174]) to ([0.64307664 1.68497091])\nIncrease radii from ([0.25 1.31722798]) to ([0.275 1.44895077])\nIncrease radii from ([0.275 1.44895077]) to ([0.3025 1.59384585])\nIncrease radii from ([0.3025 1.59384585]) to ([0.33275 1.75323044])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.6853811]) to ([0.275 0.75391921])\nIncrease radii from ([0.275 0.75391921]) to ([0.3025 0.82931113])\nIncrease radii from ([0.3025 0.82931113]) to ([0.33275 0.91224224])\nIncrease radii from ([0.33275 0.91224224]) to ([0.366025 1.00346646])\nIncrease radii from ([0.366025 1.00346646]) to ([0.4026275 1.10381311])\nIncrease radii from ([0.4026275 1.10381311]) to ([0.44289025 1.21419442])\nIncrease radii from ([0.44289025 1.21419442]) to ([0.48717928 1.33561386])\nIncrease radii from ([0.48717928 1.33561386]) to ([0.5358972 1.46917525])\nIncrease radii from ([0.5358972 1.46917525]) to ([0.58948692 1.61609277])\nIncrease radii from ([0.25 0.93286532]) to ([0.275 1.02615185])\nIncrease radii from ([0.275 1.02615185]) to ([0.3025 1.12876704])\nIncrease radii from ([0.25 0.6550428]) to ([0.275 0.72054708])\nIncrease radii from ([0.275 0.72054708]) to ([0.3025 0.79260179])\nIncrease radii from ([0.3025 0.79260179]) to ([0.33275 0.87186197])\nIncrease radii from ([0.33275 0.87186197]) to ([0.366025 0.95904817])\nIncrease radii from ([0.366025 0.95904817]) to ([0.4026275 1.05495299])\nIncrease radii from ([0.4026275 1.05495299]) to ([0.44289025 1.16044828])\nIncrease radii from ([0.44289025 1.16044828]) to ([0.48717928 1.27649311])\nIncrease radii from ([0.48717928 1.27649311]) to ([0.5358972 1.40414242])\nIncrease radii from ([0.5358972 1.40414242]) to ([0.58948692 1.54455667])\nIncrease radii from ([0.58948692 1.54455667]) to ([0.64843562 1.69901233])\nIncrease radii from ([0.25 0.96205219]) to ([0.275 1.05825741])\nIncrease radii from ([0.275 1.05825741]) to ([0.3025 1.16408315])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1735.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.58010582]) to ([0.66 1.73811641])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.50783841]) to ([0.605 1.65862225])\nIncrease radii from ([0.55 1.44843034]) to ([0.605 1.59327337])\nIncrease radii from ([0.605 1.59327337]) to ([0.6655 1.75260071])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.37076219]) to ([0.55 1.50783841])\nIncrease radii from ([0.55 1.50783841]) to ([0.605 1.65862225])\nIncrease radii from ([0.5 1.31675485]) to ([0.55 1.44843034])\nIncrease radii from ([0.55 1.44843034]) to ([0.605 1.59327337])\nIncrease radii from ([0.605 1.59327337]) to ([0.6655 1.75260071])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.23368597]) to ([0.495 1.35705457])\nIncrease radii from ([0.495 1.35705457]) to ([0.5445 1.49276003])\nIncrease radii from ([0.5445 1.49276003]) to ([0.59895 1.64203603])\nIncrease radii from ([0.45 1.18507937]) to ([0.495 1.3035873])\nIncrease radii from ([0.495 1.3035873]) to ([0.5445 1.43394603])\nIncrease radii from ([0.5445 1.43394603]) to ([0.59895 1.57734064])\nIncrease radii from ([0.59895 1.57734064]) to ([0.658845 1.7350747])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.09660975]) to ([0.44 1.20627073])\nIncrease radii from ([0.44 1.20627073]) to ([0.484 1.3268978])\nIncrease radii from ([0.484 1.3268978]) to ([0.5324 1.45958758])\nIncrease radii from ([0.5324 1.45958758]) to ([0.58564 1.60554634])\nIncrease radii from ([0.58564 1.60554634]) to ([0.644204 1.76610097])\nIncrease radii from ([0.4 1.05340388]) to ([0.44 1.15874427])\nIncrease radii from ([0.44 1.15874427]) to ([0.484 1.2746187])\nIncrease radii from ([0.484 1.2746187]) to ([0.5324 1.40208057])\nIncrease radii from ([0.5324 1.40208057]) to ([0.58564 1.54228862])\nIncrease radii from ([0.58564 1.54228862]) to ([0.644204 1.69651749])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.95953353]) to ([0.385 1.05548689])\nIncrease radii from ([0.385 1.05548689]) to ([0.4235 1.16103558])\nIncrease radii from ([0.4235 1.16103558]) to ([0.46585 1.27713913])\nIncrease radii from ([0.46585 1.27713913]) to ([0.512435 1.40485305])\nIncrease radii from ([0.512435 1.40485305]) to ([0.5636785 1.54533835])\nIncrease radii from ([0.5636785 1.54533835]) to ([0.62004635 1.69987219])\nIncrease radii from ([0.93657338 0.93657338]) to ([1.03023072 1.03023072])\nIncrease radii from ([0.35 0.9217284]) to ([0.385 1.01390124])\nIncrease radii from ([0.385 1.01390124]) to ([0.4235 1.11529136])\nIncrease radii from ([0.4235 1.11529136]) to ([0.46585 1.2268205])\nIncrease radii from ([0.46585 1.2268205]) to ([0.512435 1.34950255])\nIncrease radii from ([0.512435 1.34950255]) to ([0.5636785 1.4844528])\nIncrease radii from ([0.5636785 1.4844528]) to ([0.62004635 1.63289808])\nIncrease radii from ([0.3 1.16991019]) to ([0.33 1.28690121])\nIncrease radii from ([0.33 1.28690121]) to ([0.363 1.41559134])\nIncrease radii from ([0.3 1.58067357]) to ([0.33 1.73874093])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.82245731]) to ([0.33 0.90470305])\nIncrease radii from ([0.33 0.90470305]) to ([0.363 0.99517335])\nIncrease radii from ([0.363 0.99517335]) to ([0.3993 1.09469069])\nIncrease radii from ([0.3993 1.09469069]) to ([0.43923 1.20415975])\nIncrease radii from ([0.43923 1.20415975]) to ([0.483153 1.32457573])\nIncrease radii from ([0.483153 1.32457573]) to ([0.5314683 1.4570333])\nIncrease radii from ([0.5314683 1.4570333]) to ([0.58461513 1.60273663])\nIncrease radii from ([0.58461513 1.60273663]) to ([0.64307664 1.7630103 ])\nIncrease radii from ([0.80277719 0.80277719]) to ([0.88305491 0.88305491])\nIncrease radii from ([0.88305491 0.88305491]) to ([0.9713604 0.9713604])\nIncrease radii from ([0.3 0.79005291]) to ([0.33 0.8690582])\nIncrease radii from ([0.33 0.8690582]) to ([0.363 0.95596402])\nIncrease radii from ([0.363 0.95596402]) to ([0.3993 1.05156043])\nIncrease radii from ([0.3993 1.05156043]) to ([0.43923 1.15671647])\nIncrease radii from ([0.43923 1.15671647]) to ([0.483153 1.27238811])\nIncrease radii from ([0.483153 1.27238811]) to ([0.5314683 1.39962693])\nIncrease radii from ([0.5314683 1.39962693]) to ([0.58461513 1.53958962])\nIncrease radii from ([0.58461513 1.53958962]) to ([0.64307664 1.69354858])\nIncrease radii from ([0.25 0.97492516]) to ([0.275 1.07241768])\nIncrease radii from ([0.275 1.07241768]) to ([0.3025 1.17965945])\nIncrease radii from ([0.3025 1.17965945]) to ([0.33275 1.29762539])\nIncrease radii from ([0.25 1.31722798]) to ([0.275 1.44895077])\nIncrease radii from ([0.275 1.44895077]) to ([0.3025 1.59384585])\nIncrease radii from ([0.3025 1.59384585]) to ([0.33275 1.75323044])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.6853811]) to ([0.275 0.75391921])\nIncrease radii from ([0.275 0.75391921]) to ([0.3025 0.82931113])\nIncrease radii from ([0.3025 0.82931113]) to ([0.33275 0.91224224])\nIncrease radii from ([0.33275 0.91224224]) to ([0.366025 1.00346646])\nIncrease radii from ([0.366025 1.00346646]) to ([0.4026275 1.10381311])\nIncrease radii from ([0.4026275 1.10381311]) to ([0.44289025 1.21419442])\nIncrease radii from ([0.44289025 1.21419442]) to ([0.48717928 1.33561386])\nIncrease radii from ([0.48717928 1.33561386]) to ([0.5358972 1.46917525])\nIncrease radii from ([0.5358972 1.46917525]) to ([0.58948692 1.61609277])\nIncrease radii from ([0.25 0.96205219]) to ([0.275 1.05825741])\nIncrease radii from ([0.275 1.05825741]) to ([0.3025 1.16408315])\nIncrease radii from ([0.66898099 0.66898099]) to ([0.73587909 0.73587909])\nIncrease radii from ([0.73587909 0.73587909]) to ([0.809467 0.809467])\nIncrease radii from ([0.809467 0.809467]) to ([0.8904137 0.8904137])\nIncrease radii from ([0.8904137 0.8904137]) to ([0.97945507 0.97945507])\nIncrease radii from ([0.25 0.65837743]) to ([0.275 0.72421517])\nIncrease radii from ([0.275 0.72421517]) to ([0.3025 0.79663669])\nIncrease radii from ([0.3025 0.79663669]) to ([0.33275 0.87630035])\nIncrease radii from ([0.33275 0.87630035]) to ([0.366025 0.96393039])\nIncrease radii from ([0.366025 0.96393039]) to ([0.4026275 1.06032343])\nIncrease radii from ([0.4026275 1.06032343]) to ([0.44289025 1.16635577])\nIncrease radii from ([0.44289025 1.16635577]) to ([0.48717928 1.28299135])\nIncrease radii from ([0.48717928 1.28299135]) to ([0.5358972 1.41129048])\nIncrease radii from ([0.5358972 1.41129048]) to ([0.58948692 1.55241953])\nIncrease radii from ([0.58948692 1.55241953]) to ([0.64843562 1.70766149])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1736.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.55 1.50783841]) to ([0.605 1.65862225])\nIncrease radii from ([0.5 1.44605104]) to ([0.55 1.59065615])\nIncrease radii from ([0.55 1.59065615]) to ([0.605 1.74972176])\nIncrease radii from ([0.5 1.37076219]) to ([0.55 1.50783841])\nIncrease radii from ([0.55 1.50783841]) to ([0.605 1.65862225])\nIncrease radii from ([0.45 1.30144594]) to ([0.495 1.43159053])\nIncrease radii from ([0.495 1.43159053]) to ([0.5445 1.57474959])\nIncrease radii from ([0.5445 1.57474959]) to ([0.59895 1.73222455])\nIncrease radii from ([0.45 1.23368597]) to ([0.495 1.35705457])\nIncrease radii from ([0.495 1.35705457]) to ([0.5445 1.49276003])\nIncrease radii from ([0.5445 1.49276003]) to ([0.59895 1.64203603])\nIncrease radii from ([0.4 1.15684084]) to ([0.44 1.27252492])\nIncrease radii from ([0.44 1.27252492]) to ([0.484 1.39977741])\nIncrease radii from ([0.484 1.39977741]) to ([0.5324 1.53975515])\nIncrease radii from ([0.5324 1.53975515]) to ([0.58564 1.69373067])\nIncrease radii from ([0.4 1.09660975]) to ([0.44 1.20627073])\nIncrease radii from ([0.44 1.20627073]) to ([0.484 1.3268978])\nIncrease radii from ([0.484 1.3268978]) to ([0.5324 1.45958758])\nIncrease radii from ([0.5324 1.45958758]) to ([0.58564 1.60554634])\nIncrease radii from ([0.58564 1.60554634]) to ([0.644204 1.76610097])\nIncrease radii from ([0.35 1.01223573]) to ([0.385 1.1134593])\nIncrease radii from ([0.385 1.1134593]) to ([0.4235 1.22480523])\nIncrease radii from ([0.4235 1.22480523]) to ([0.46585 1.34728576])\nIncrease radii from ([0.46585 1.34728576]) to ([0.512435 1.48201433])\nIncrease radii from ([0.512435 1.48201433]) to ([0.5636785 1.63021577])\nIncrease radii from ([0.35 0.95953353]) to ([0.385 1.05548689])\nIncrease radii from ([0.385 1.05548689]) to ([0.4235 1.16103558])\nIncrease radii from ([0.4235 1.16103558]) to ([0.46585 1.27713913])\nIncrease radii from ([0.46585 1.27713913]) to ([0.512435 1.40485305])\nIncrease radii from ([0.512435 1.40485305]) to ([0.5636785 1.54533835])\nIncrease radii from ([0.5636785 1.54533835]) to ([0.62004635 1.69987219])\nIncrease radii from ([0.922183 0.922183]) to ([1.0144013 1.0144013])\nIncrease radii from ([0.3 1.58067357]) to ([0.33 1.73874093])\nIncrease radii from ([0.3 0.86763063]) to ([0.33 0.95439369])\nIncrease radii from ([0.33 0.95439369]) to ([0.363 1.04983306])\nIncrease radii from ([0.363 1.04983306]) to ([0.3993 1.15481636])\nIncrease radii from ([0.3993 1.15481636]) to ([0.43923 1.270298])\nIncrease radii from ([0.43923 1.270298]) to ([0.483153 1.3973278])\nIncrease radii from ([0.483153 1.3973278]) to ([0.5314683 1.53706058])\nIncrease radii from ([0.5314683 1.53706058]) to ([0.58461513 1.69076664])\nIncrease radii from ([0.3 0.82245731]) to ([0.33 0.90470305])\nIncrease radii from ([0.33 0.90470305]) to ([0.363 0.99517335])\nIncrease radii from ([0.363 0.99517335]) to ([0.3993 1.09469069])\nIncrease radii from ([0.3993 1.09469069]) to ([0.43923 1.20415975])\nIncrease radii from ([0.43923 1.20415975]) to ([0.483153 1.32457573])\nIncrease radii from ([0.483153 1.32457573]) to ([0.5314683 1.4570333])\nIncrease radii from ([0.5314683 1.4570333]) to ([0.58461513 1.60273663])\nIncrease radii from ([0.58461513 1.60273663]) to ([0.64307664 1.7630103 ])\nIncrease radii from ([0.79044257 0.79044257]) to ([0.86948683 0.86948683])\nIncrease radii from ([0.86948683 0.86948683]) to ([0.95643551 0.95643551])\nIncrease radii from ([0.95643551 0.95643551]) to ([1.05207906 1.05207906])\nIncrease radii from ([0.25 1.31722798]) to ([0.275 1.44895077])\nIncrease radii from ([0.275 1.44895077]) to ([0.3025 1.59384585])\nIncrease radii from ([0.3025 1.59384585]) to ([0.33275 1.75323044])\nIncrease radii from ([0.25 0.72302552]) to ([0.275 0.79532807])\nIncrease radii from ([0.275 0.79532807]) to ([0.3025 0.87486088])\nIncrease radii from ([0.3025 0.87486088]) to ([0.33275 0.96234697])\nIncrease radii from ([0.33275 0.96234697]) to ([0.366025 1.05858167])\nIncrease radii from ([0.366025 1.05858167]) to ([0.4026275 1.16443983])\nIncrease radii from ([0.4026275 1.16443983]) to ([0.44289025 1.28088382])\nIncrease radii from ([0.44289025 1.28088382]) to ([0.48717928 1.4089722 ])\nIncrease radii from ([0.48717928 1.4089722 ]) to ([0.5358972 1.54986942])\nIncrease radii from ([0.5358972 1.54986942]) to ([0.58948692 1.70485636])\nIncrease radii from ([0.25 0.6853811]) to ([0.275 0.75391921])\nIncrease radii from ([0.275 0.75391921]) to ([0.3025 0.82931113])\nIncrease radii from ([0.3025 0.82931113]) to ([0.33275 0.91224224])\nIncrease radii from ([0.33275 0.91224224]) to ([0.366025 1.00346646])\nIncrease radii from ([0.366025 1.00346646]) to ([0.4026275 1.10381311])\nIncrease radii from ([0.4026275 1.10381311]) to ([0.44289025 1.21419442])\nIncrease radii from ([0.44289025 1.21419442]) to ([0.48717928 1.33561386])\nIncrease radii from ([0.48717928 1.33561386]) to ([0.5358972 1.46917525])\nIncrease radii from ([0.5358972 1.46917525]) to ([0.58948692 1.61609277])\nIncrease radii from ([0.65870214 0.65870214]) to ([0.72457236 0.72457236])\nIncrease radii from ([0.72457236 0.72457236]) to ([0.79702959 0.79702959])\nIncrease radii from ([0.79702959 0.79702959]) to ([0.87673255 0.87673255])\nIncrease radii from ([0.87673255 0.87673255]) to ([0.96440581 0.96440581])\nIncrease radii from ([0.96440581 0.96440581]) to ([1.06084639 1.06084639])\nIncrease radii from ([0.25 0.93266647]) to ([0.275 1.02593311])\nIncrease radii from ([0.275 1.02593311]) to ([0.3025 1.12852642])\nIncrease radii from ([0.25 0.92467885]) to ([0.275 1.01714673])\nIncrease radii from ([0.275 1.01714673]) to ([0.3025 1.11886141])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1737.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.922183 0.922183]) to ([1.0144013 1.0144013])\nIncrease radii from ([0.79044257 0.79044257]) to ([0.86948683 0.86948683])\nIncrease radii from ([0.86948683 0.86948683]) to ([0.95643551 0.95643551])\nIncrease radii from ([0.95643551 0.95643551]) to ([1.05207906 1.05207906])\nIncrease radii from ([0.65870214 0.65870214]) to ([0.72457236 0.72457236])\nIncrease radii from ([0.72457236 0.72457236]) to ([0.79702959 0.79702959])\nIncrease radii from ([0.79702959 0.79702959]) to ([0.87673255 0.87673255])\nIncrease radii from ([0.87673255 0.87673255]) to ([0.96440581 0.96440581])\nIncrease radii from ([0.96440581 0.96440581]) to ([1.06084639 1.06084639])\nIncrease radii from ([0.25 0.93266647]) to ([0.275 1.02593311])\nIncrease radii from ([0.275 1.02593311]) to ([0.3025 1.12852642])\nIncrease radii from ([0.25 0.92467885]) to ([0.275 1.01714673])\nIncrease radii from ([0.275 1.01714673]) to ([0.3025 1.11886141])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1738.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.922183 0.922183]) to ([1.0144013 1.0144013])\nIncrease radii from ([0.3 1.17975659]) to ([0.33 1.29773225])\nIncrease radii from ([0.79044257 0.79044257]) to ([0.86948683 0.86948683])\nIncrease radii from ([0.86948683 0.86948683]) to ([0.95643551 0.95643551])\nIncrease radii from ([0.95643551 0.95643551]) to ([1.05207906 1.05207906])\nIncrease radii from ([0.25 0.98313049]) to ([0.275 1.08144354])\nIncrease radii from ([0.275 1.08144354]) to ([0.3025 1.1895879])\nIncrease radii from ([0.3025 1.1895879]) to ([0.33275 1.30854669])\nIncrease radii from ([0.65870214 0.65870214]) to ([0.72457236 0.72457236])\nIncrease radii from ([0.72457236 0.72457236]) to ([0.79702959 0.79702959])\nIncrease radii from ([0.79702959 0.79702959]) to ([0.87673255 0.87673255])\nIncrease radii from ([0.87673255 0.87673255]) to ([0.96440581 0.96440581])\nIncrease radii from ([0.96440581 0.96440581]) to ([1.06084639 1.06084639])\nIncrease radii from ([0.25 0.93332909]) to ([0.275 1.026662])\nIncrease radii from ([0.275 1.026662]) to ([0.3025 1.1293282])\nIncrease radii from ([0.25 0.92467885]) to ([0.275 1.01714673])\nIncrease radii from ([0.275 1.01714673]) to ([0.3025 1.11886141])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1739.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.922183 0.922183]) to ([1.0144013 1.0144013])\nIncrease radii from ([0.79044257 0.79044257]) to ([0.86948683 0.86948683])\nIncrease radii from ([0.86948683 0.86948683]) to ([0.95643551 0.95643551])\nIncrease radii from ([0.95643551 0.95643551]) to ([1.05207906 1.05207906])\nIncrease radii from ([0.65870214 0.65870214]) to ([0.72457236 0.72457236])\nIncrease radii from ([0.72457236 0.72457236]) to ([0.79702959 0.79702959])\nIncrease radii from ([0.79702959 0.79702959]) to ([0.87673255 0.87673255])\nIncrease radii from ([0.87673255 0.87673255]) to ([0.96440581 0.96440581])\nIncrease radii from ([0.96440581 0.96440581]) to ([1.06084639 1.06084639])\nIncrease radii from ([0.43638123 0.75583447]) to ([0.48001936 0.83141792])\nIncrease radii from ([0.25 0.92467885]) to ([0.275 1.01714673])\nIncrease radii from ([0.275 1.01714673]) to ([0.3025 1.11886141])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1740.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.5836225]) to ([0.66 1.74198475])\nIncrease radii from ([0.6 1.57356783]) to ([0.66 1.73092461])\nIncrease radii from ([0.55 1.45165396]) to ([0.605 1.59681936])\nIncrease radii from ([0.605 1.59681936]) to ([0.6655 1.75650129])\nIncrease radii from ([0.55 1.44243718]) to ([0.605 1.58668089])\nIncrease radii from ([0.605 1.58668089]) to ([0.6655 1.74534898])\nIncrease radii from ([0.5 1.31968542]) to ([0.55 1.45165396])\nIncrease radii from ([0.55 1.45165396]) to ([0.605 1.59681936])\nIncrease radii from ([0.605 1.59681936]) to ([0.6655 1.75650129])\nIncrease radii from ([0.5 1.31130652]) to ([0.55 1.44243718])\nIncrease radii from ([0.55 1.44243718]) to ([0.605 1.58668089])\nIncrease radii from ([0.605 1.58668089]) to ([0.6655 1.74534898])\nIncrease radii from ([0.45 1.18771688]) to ([0.495 1.30648856])\nIncrease radii from ([0.495 1.30648856]) to ([0.5445 1.43713742])\nIncrease radii from ([0.5445 1.43713742]) to ([0.59895 1.58085116])\nIncrease radii from ([0.59895 1.58085116]) to ([0.658845 1.73893628])\nIncrease radii from ([0.45 1.18017587]) to ([0.495 1.29819346])\nIncrease radii from ([0.495 1.29819346]) to ([0.5445 1.4280128])\nIncrease radii from ([0.5445 1.4280128]) to ([0.59895 1.57081409])\nIncrease radii from ([0.59895 1.57081409]) to ([0.658845 1.72789549])\nIncrease radii from ([0.4 1.05574834]) to ([0.44 1.16132317])\nIncrease radii from ([0.44 1.16132317]) to ([0.484 1.27745549])\nIncrease radii from ([0.484 1.27745549]) to ([0.5324 1.40520103])\nIncrease radii from ([0.5324 1.40520103]) to ([0.58564 1.54572114])\nIncrease radii from ([0.58564 1.54572114]) to ([0.644204 1.70029325])\nIncrease radii from ([0.4 1.04904522]) to ([0.44 1.15394974])\nIncrease radii from ([0.44 1.15394974]) to ([0.484 1.26934472])\nIncrease radii from ([0.484 1.26934472]) to ([0.5324 1.39627919])\nIncrease radii from ([0.5324 1.39627919]) to ([0.58564 1.53590711])\nIncrease radii from ([0.58564 1.53590711]) to ([0.644204 1.68949782])\nIncrease radii from ([0.35 0.92377979]) to ([0.385 1.01615777])\nIncrease radii from ([0.385 1.01615777]) to ([0.4235 1.11777355])\nIncrease radii from ([0.4235 1.11777355]) to ([0.46585 1.2295509])\nIncrease radii from ([0.46585 1.2295509]) to ([0.512435 1.352506])\nIncrease radii from ([0.512435 1.352506]) to ([0.5636785 1.48775659])\nIncrease radii from ([0.5636785 1.48775659]) to ([0.62004635 1.63653225])\nIncrease radii from ([0.35 0.91791457]) to ([0.385 1.00970602])\nIncrease radii from ([0.385 1.00970602]) to ([0.4235 1.11067663])\nIncrease radii from ([0.4235 1.11067663]) to ([0.46585 1.22174429])\nIncrease radii from ([0.46585 1.22174429]) to ([0.512435 1.34391872])\nIncrease radii from ([0.512435 1.34391872]) to ([0.5636785 1.47831059])\nIncrease radii from ([0.5636785 1.47831059]) to ([0.62004635 1.62614165])\nIncrease radii from ([0.3 0.79181125]) to ([0.33 0.87099238])\nIncrease radii from ([0.33 0.87099238]) to ([0.363 0.95809161])\nIncrease radii from ([0.363 0.95809161]) to ([0.3993 1.05390078])\nIncrease radii from ([0.3993 1.05390078]) to ([0.43923 1.15929085])\nIncrease radii from ([0.43923 1.15929085]) to ([0.483153 1.27521994])\nIncrease radii from ([0.483153 1.27521994]) to ([0.5314683 1.40274193])\nIncrease radii from ([0.5314683 1.40274193]) to ([0.58461513 1.54301613])\nIncrease radii from ([0.58461513 1.54301613]) to ([0.64307664 1.69731774])\nIncrease radii from ([0.3 0.78678391]) to ([0.33 0.86546231])\nIncrease radii from ([0.33 0.86546231]) to ([0.363 0.95200854])\nIncrease radii from ([0.363 0.95200854]) to ([0.3993 1.04720939])\nIncrease radii from ([0.3993 1.04720939]) to ([0.43923 1.15193033])\nIncrease radii from ([0.43923 1.15193033]) to ([0.483153 1.26712336])\nIncrease radii from ([0.483153 1.26712336]) to ([0.5314683 1.3938357])\nIncrease radii from ([0.5314683 1.3938357]) to ([0.58461513 1.53321927])\nIncrease radii from ([0.58461513 1.53321927]) to ([0.64307664 1.6865412 ])\nIncrease radii from ([0.25 0.65984271]) to ([0.275 0.72582698])\nIncrease radii from ([0.275 0.72582698]) to ([0.3025 0.79840968])\nIncrease radii from ([0.3025 0.79840968]) to ([0.33275 0.87825065])\nIncrease radii from ([0.33275 0.87825065]) to ([0.366025 0.96607571])\nIncrease radii from ([0.366025 0.96607571]) to ([0.4026275 1.06268328])\nIncrease radii from ([0.4026275 1.06268328]) to ([0.44289025 1.16895161])\nIncrease radii from ([0.44289025 1.16895161]) to ([0.48717928 1.28584677])\nIncrease radii from ([0.48717928 1.28584677]) to ([0.5358972 1.41443145])\nIncrease radii from ([0.5358972 1.41443145]) to ([0.58948692 1.55587459])\nIncrease radii from ([0.58948692 1.55587459]) to ([0.64843562 1.71146205])\nIncrease radii from ([0.25 0.65565326]) to ([0.275 0.72121859])\nIncrease radii from ([0.275 0.72121859]) to ([0.3025 0.79334045])\nIncrease radii from ([0.3025 0.79334045]) to ([0.33275 0.87267449])\nIncrease radii from ([0.33275 0.87267449]) to ([0.366025 0.95994194])\nIncrease radii from ([0.366025 0.95994194]) to ([0.4026275 1.05593614])\nIncrease radii from ([0.4026275 1.05593614]) to ([0.44289025 1.16152975])\nIncrease radii from ([0.44289025 1.16152975]) to ([0.48717928 1.27768272])\nIncrease radii from ([0.48717928 1.27768272]) to ([0.5358972 1.405451 ])\nIncrease radii from ([0.5358972 1.405451 ]) to ([0.58948692 1.5459961 ])\nIncrease radii from ([0.58948692 1.5459961 ]) to ([0.64843562 1.70059571])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1741.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.59895825]) to ([0.66 1.75885407])\nIncrease radii from ([0.55 1.46571173]) to ([0.605 1.6122829])\nIncrease radii from ([0.605 1.6122829]) to ([0.6655 1.77351119])\nIncrease radii from ([0.5 1.33246521]) to ([0.55 1.46571173])\nIncrease radii from ([0.55 1.46571173]) to ([0.605 1.6122829])\nIncrease radii from ([0.605 1.6122829]) to ([0.6655 1.77351119])\nIncrease radii from ([0.45 1.19921869]) to ([0.495 1.31914056])\nIncrease radii from ([0.495 1.31914056]) to ([0.5445 1.45105461])\nIncrease radii from ([0.5445 1.45105461]) to ([0.59895 1.59616007])\nIncrease radii from ([0.59895 1.59616007]) to ([0.658845 1.75577608])\nIncrease radii from ([0.4 1.06597217]) to ([0.44 1.17256938])\nIncrease radii from ([0.44 1.17256938]) to ([0.484 1.28982632])\nIncrease radii from ([0.484 1.28982632]) to ([0.5324 1.41880895])\nIncrease radii from ([0.5324 1.41880895]) to ([0.58564 1.56068985])\nIncrease radii from ([0.58564 1.56068985]) to ([0.644204 1.71675883])\nIncrease radii from ([0.35 0.93272565]) to ([0.385 1.02599821])\nIncrease radii from ([0.385 1.02599821]) to ([0.4235 1.12859803])\nIncrease radii from ([0.4235 1.12859803]) to ([0.46585 1.24145783])\nIncrease radii from ([0.46585 1.24145783]) to ([0.512435 1.36560362])\nIncrease radii from ([0.512435 1.36560362]) to ([0.5636785 1.50216398])\nIncrease radii from ([0.5636785 1.50216398]) to ([0.62004635 1.65238038])\nIncrease radii from ([0.3 0.79947912]) to ([0.33 0.87942704])\nIncrease radii from ([0.33 0.87942704]) to ([0.363 0.96736974])\nIncrease radii from ([0.363 0.96736974]) to ([0.3993 1.06410672])\nIncrease radii from ([0.3993 1.06410672]) to ([0.43923 1.17051739])\nIncrease radii from ([0.43923 1.17051739]) to ([0.483153 1.28756913])\nIncrease radii from ([0.483153 1.28756913]) to ([0.5314683 1.41632604])\nIncrease radii from ([0.5314683 1.41632604]) to ([0.58461513 1.55795864])\nIncrease radii from ([0.58461513 1.55795864]) to ([0.64307664 1.71375451])\nIncrease radii from ([0.25 0.6662326]) to ([0.275 0.73285586])\nIncrease radii from ([0.275 0.73285586]) to ([0.3025 0.80614145])\nIncrease radii from ([0.3025 0.80614145]) to ([0.33275 0.8867556])\nIncrease radii from ([0.33275 0.8867556]) to ([0.366025 0.97543116])\nIncrease radii from ([0.366025 0.97543116]) to ([0.4026275 1.07297427])\nIncrease radii from ([0.4026275 1.07297427]) to ([0.44289025 1.1802717 ])\nIncrease radii from ([0.44289025 1.1802717 ]) to ([0.48717928 1.29829887])\nIncrease radii from ([0.48717928 1.29829887]) to ([0.5358972 1.42812876])\nIncrease radii from ([0.5358972 1.42812876]) to ([0.58948692 1.57094163])\nIncrease radii from ([0.58948692 1.57094163]) to ([0.64843562 1.72803579])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1742.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.57266314]) to ([0.66 1.72992946])\nIncrease radii from ([0.55 1.44160788]) to ([0.605 1.58576867])\nIncrease radii from ([0.605 1.58576867]) to ([0.6655 1.74434553])\nIncrease radii from ([0.5 1.31055262]) to ([0.55 1.44160788])\nIncrease radii from ([0.55 1.44160788]) to ([0.605 1.58576867])\nIncrease radii from ([0.605 1.58576867]) to ([0.6655 1.74434553])\nIncrease radii from ([0.45 1.17949736]) to ([0.495 1.29744709])\nIncrease radii from ([0.495 1.29744709]) to ([0.5445 1.4271918])\nIncrease radii from ([0.5445 1.4271918]) to ([0.59895 1.56991098])\nIncrease radii from ([0.59895 1.56991098]) to ([0.658845 1.72690208])\nIncrease radii from ([0.4 1.04844209]) to ([0.44 1.1532863])\nIncrease radii from ([0.44 1.1532863]) to ([0.484 1.26861493])\nIncrease radii from ([0.484 1.26861493]) to ([0.5324 1.39547643])\nIncrease radii from ([0.5324 1.39547643]) to ([0.58564 1.53502407])\nIncrease radii from ([0.58564 1.53502407]) to ([0.644204 1.68852648])\nIncrease radii from ([0.35 0.91738683]) to ([0.385 1.00912552])\nIncrease radii from ([0.385 1.00912552]) to ([0.4235 1.11003807])\nIncrease radii from ([0.4235 1.11003807]) to ([0.46585 1.22104187])\nIncrease radii from ([0.46585 1.22104187]) to ([0.512435 1.34314606])\nIncrease radii from ([0.512435 1.34314606]) to ([0.5636785 1.47746067])\nIncrease radii from ([0.5636785 1.47746067]) to ([0.62004635 1.62520673])\nIncrease radii from ([0.3 0.78633157]) to ([0.33 0.86496473])\nIncrease radii from ([0.33 0.86496473]) to ([0.363 0.9514612])\nIncrease radii from ([0.363 0.9514612]) to ([0.3993 1.04660732])\nIncrease radii from ([0.3993 1.04660732]) to ([0.43923 1.15126805])\nIncrease radii from ([0.43923 1.15126805]) to ([0.483153 1.26639486])\nIncrease radii from ([0.483153 1.26639486]) to ([0.5314683 1.39303434])\nIncrease radii from ([0.5314683 1.39303434]) to ([0.58461513 1.53233778])\nIncrease radii from ([0.58461513 1.53233778]) to ([0.64307664 1.68557156])\nIncrease radii from ([0.25 0.65527631]) to ([0.275 0.72080394])\nIncrease radii from ([0.275 0.72080394]) to ([0.3025 0.79288433])\nIncrease radii from ([0.3025 0.79288433]) to ([0.33275 0.87217277])\nIncrease radii from ([0.33275 0.87217277]) to ([0.366025 0.95939004])\nIncrease radii from ([0.366025 0.95939004]) to ([0.4026275 1.05532905])\nIncrease radii from ([0.4026275 1.05532905]) to ([0.44289025 1.16086195])\nIncrease radii from ([0.44289025 1.16086195]) to ([0.48717928 1.27694815])\nIncrease radii from ([0.48717928 1.27694815]) to ([0.5358972 1.40464296])\nIncrease radii from ([0.5358972 1.40464296]) to ([0.58948692 1.54510726])\nIncrease radii from ([0.58948692 1.54510726]) to ([0.64843562 1.69961799])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1743.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.94127509 0.94127509]) to ([1.0354026 1.0354026])\nIncrease radii from ([0.80680722 0.80680722]) to ([0.88748794 0.88748794])\nIncrease radii from ([0.88748794 0.88748794]) to ([0.97623674 0.97623674])\nIncrease radii from ([0.67233935 0.67233935]) to ([0.73957329 0.73957329])\nIncrease radii from ([0.73957329 0.73957329]) to ([0.81353062 0.81353062])\nIncrease radii from ([0.81353062 0.81353062]) to ([0.89488368 0.89488368])\nIncrease radii from ([0.89488368 0.89488368]) to ([0.98437204 0.98437204])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1744.tif is a low contrast image\n compress=6,\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1745.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.3 1.14207922]) to ([0.33 1.25628714])\nIncrease radii from ([0.33 1.25628714]) to ([0.363 1.38191585])\nIncrease radii from ([0.25 0.95173268]) to ([0.275 1.04690595])\nIncrease radii from ([0.275 1.04690595]) to ([0.3025 1.15159654])\nIncrease radii from ([0.3025 1.15159654]) to ([0.33275 1.2667562])\nIncrease radii from ([0.33275 1.2667562]) to ([0.366025 1.39343182])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1746.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.569237]) to ([0.66 1.7261607])\nIncrease radii from ([0.6 1.5805268]) to ([0.66 1.73857948])\nIncrease radii from ([0.55 1.43846725]) to ([0.605 1.58231398])\nIncrease radii from ([0.605 1.58231398]) to ([0.6655 1.74054537])\nIncrease radii from ([0.55 1.44881623]) to ([0.605 1.59369785])\nIncrease radii from ([0.605 1.59369785]) to ([0.6655 1.75306764])\nIncrease radii from ([0.5 1.3076975]) to ([0.55 1.43846725])\nIncrease radii from ([0.55 1.43846725]) to ([0.605 1.58231398])\nIncrease radii from ([0.605 1.58231398]) to ([0.6655 1.74054537])\nIncrease radii from ([0.5 1.31710566]) to ([0.55 1.44881623])\nIncrease radii from ([0.55 1.44881623]) to ([0.605 1.59369785])\nIncrease radii from ([0.605 1.59369785]) to ([0.6655 1.75306764])\nIncrease radii from ([0.45 1.17692775]) to ([0.495 1.29462053])\nIncrease radii from ([0.495 1.29462053]) to ([0.5445 1.42408258])\nIncrease radii from ([0.5445 1.42408258]) to ([0.59895 1.56649084])\nIncrease radii from ([0.59895 1.56649084]) to ([0.658845 1.72313992])\nIncrease radii from ([0.45 1.1853951]) to ([0.495 1.30393461])\nIncrease radii from ([0.495 1.30393461]) to ([0.5445 1.43432807])\nIncrease radii from ([0.5445 1.43432807]) to ([0.59895 1.57776087])\nIncrease radii from ([0.59895 1.57776087]) to ([0.658845 1.73553696])\nIncrease radii from ([0.4 1.046158]) to ([0.44 1.1507738])\nIncrease radii from ([0.44 1.1507738]) to ([0.484 1.26585118])\nIncrease radii from ([0.484 1.26585118]) to ([0.5324 1.3924363])\nIncrease radii from ([0.5324 1.3924363]) to ([0.58564 1.53167993])\nIncrease radii from ([0.58564 1.53167993]) to ([0.644204 1.68484792])\nIncrease radii from ([0.4 1.05368453]) to ([0.44 1.15905298])\nIncrease radii from ([0.44 1.15905298]) to ([0.484 1.27495828])\nIncrease radii from ([0.484 1.27495828]) to ([0.5324 1.40245411])\nIncrease radii from ([0.5324 1.40245411]) to ([0.58564 1.54269952])\nIncrease radii from ([0.58564 1.54269952]) to ([0.644204 1.69696947])\nIncrease radii from ([0.35 0.91538825]) to ([0.385 1.00692708])\nIncrease radii from ([0.385 1.00692708]) to ([0.4235 1.10761978])\nIncrease radii from ([0.4235 1.10761978]) to ([0.46585 1.21838176])\nIncrease radii from ([0.46585 1.21838176]) to ([0.512435 1.34021994])\nIncrease radii from ([0.512435 1.34021994]) to ([0.5636785 1.47424193])\nIncrease radii from ([0.5636785 1.47424193]) to ([0.62004635 1.62166612])\nIncrease radii from ([0.35 0.92197396]) to ([0.385 1.01417136])\nIncrease radii from ([0.385 1.01417136]) to ([0.4235 1.1155885])\nIncrease radii from ([0.4235 1.1155885]) to ([0.46585 1.22714735])\nIncrease radii from ([0.46585 1.22714735]) to ([0.512435 1.34986208])\nIncrease radii from ([0.512435 1.34986208]) to ([0.5636785 1.48484829])\nIncrease radii from ([0.5636785 1.48484829]) to ([0.62004635 1.63333312])\nIncrease radii from ([0.3 0.7846185]) to ([0.33 0.86308035])\nIncrease radii from ([0.33 0.86308035]) to ([0.363 0.94938839])\nIncrease radii from ([0.363 0.94938839]) to ([0.3993 1.04432722])\nIncrease radii from ([0.3993 1.04432722]) to ([0.43923 1.14875995])\nIncrease radii from ([0.43923 1.14875995]) to ([0.483153 1.26363594])\nIncrease radii from ([0.483153 1.26363594]) to ([0.5314683 1.38999953])\nIncrease radii from ([0.5314683 1.38999953]) to ([0.58461513 1.52899949])\nIncrease radii from ([0.58461513 1.52899949]) to ([0.64307664 1.68189944])\nIncrease radii from ([0.3 0.7902634]) to ([0.33 0.86928974])\nIncrease radii from ([0.33 0.86928974]) to ([0.363 0.95621871])\nIncrease radii from ([0.363 0.95621871]) to ([0.3993 1.05184058])\nIncrease radii from ([0.3993 1.05184058]) to ([0.43923 1.15702464])\nIncrease radii from ([0.43923 1.15702464]) to ([0.483153 1.27272711])\nIncrease radii from ([0.483153 1.27272711]) to ([0.5314683 1.39999982])\nIncrease radii from ([0.5314683 1.39999982]) to ([0.58461513 1.5399998 ])\nIncrease radii from ([0.58461513 1.5399998 ]) to ([0.64307664 1.69399978])\nIncrease radii from ([0.3 1.14207922]) to ([0.33 1.25628714])\nIncrease radii from ([0.33 1.25628714]) to ([0.363 1.38191585])\nIncrease radii from ([0.25 0.93712122]) to ([0.275 1.03083335])\nIncrease radii from ([0.275 1.03083335]) to ([0.3025 1.13391668])\nIncrease radii from ([0.25 0.65384875]) to ([0.275 0.71923363])\nIncrease radii from ([0.275 0.71923363]) to ([0.3025 0.79115699])\nIncrease radii from ([0.3025 0.79115699]) to ([0.33275 0.87027269])\nIncrease radii from ([0.33275 0.87027269]) to ([0.366025 0.95729996])\nIncrease radii from ([0.366025 0.95729996]) to ([0.4026275 1.05302995])\nIncrease radii from ([0.4026275 1.05302995]) to ([0.44289025 1.15833295])\nIncrease radii from ([0.44289025 1.15833295]) to ([0.48717928 1.27416624])\nIncrease radii from ([0.48717928 1.27416624]) to ([0.5358972 1.40158286])\nIncrease radii from ([0.5358972 1.40158286]) to ([0.58948692 1.54174115])\nIncrease radii from ([0.58948692 1.54174115]) to ([0.64843562 1.69591527])\nIncrease radii from ([0.25 0.65855283]) to ([0.275 0.72440811])\nIncrease radii from ([0.275 0.72440811]) to ([0.3025 0.79684893])\nIncrease radii from ([0.3025 0.79684893]) to ([0.33275 0.87653382])\nIncrease radii from ([0.33275 0.87653382]) to ([0.366025 0.9641872])\nIncrease radii from ([0.366025 0.9641872]) to ([0.4026275 1.06060592])\nIncrease radii from ([0.4026275 1.06060592]) to ([0.44289025 1.16666651])\nIncrease radii from ([0.44289025 1.16666651]) to ([0.48717928 1.28333316])\nIncrease radii from ([0.48717928 1.28333316]) to ([0.5358972 1.41166648])\nIncrease radii from ([0.5358972 1.41166648]) to ([0.58948692 1.55283313])\nIncrease radii from ([0.58948692 1.55283313]) to ([0.64843562 1.70811644])\nIncrease radii from ([0.25 0.95173268]) to ([0.275 1.04690595])\nIncrease radii from ([0.275 1.04690595]) to ([0.3025 1.15159654])\nIncrease radii from ([0.3025 1.15159654]) to ([0.33275 1.2667562])\nIncrease radii from ([0.33275 1.2667562]) to ([0.366025 1.39343182])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1747.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.6104011]) to ([0.66 1.7714412])\nIncrease radii from ([0.6 1.5805268]) to ([0.66 1.73857948])\nIncrease radii from ([0.55 1.476201]) to ([0.605 1.6238211])\nIncrease radii from ([0.55 1.44881623]) to ([0.605 1.59369785])\nIncrease radii from ([0.605 1.59369785]) to ([0.6655 1.75306764])\nIncrease radii from ([0.5 1.34200091]) to ([0.55 1.476201])\nIncrease radii from ([0.55 1.476201]) to ([0.605 1.6238211])\nIncrease radii from ([0.5 1.31710566]) to ([0.55 1.44881623])\nIncrease radii from ([0.55 1.44881623]) to ([0.605 1.59369785])\nIncrease radii from ([0.605 1.59369785]) to ([0.6655 1.75306764])\nIncrease radii from ([0.45 1.20780082]) to ([0.495 1.3285809])\nIncrease radii from ([0.495 1.3285809]) to ([0.5445 1.46143899])\nIncrease radii from ([0.5445 1.46143899]) to ([0.59895 1.60758289])\nIncrease radii from ([0.59895 1.60758289]) to ([0.658845 1.76834118])\nIncrease radii from ([0.45 1.1853951]) to ([0.495 1.30393461])\nIncrease radii from ([0.495 1.30393461]) to ([0.5445 1.43432807])\nIncrease radii from ([0.5445 1.43432807]) to ([0.59895 1.57776087])\nIncrease radii from ([0.59895 1.57776087]) to ([0.658845 1.73553696])\nIncrease radii from ([0.4 1.07360073]) to ([0.44 1.1809608])\nIncrease radii from ([0.44 1.1809608]) to ([0.484 1.29905688])\nIncrease radii from ([0.484 1.29905688]) to ([0.5324 1.42896257])\nIncrease radii from ([0.5324 1.42896257]) to ([0.58564 1.57185883])\nIncrease radii from ([0.58564 1.57185883]) to ([0.644204 1.72904471])\nIncrease radii from ([0.4 1.05368453]) to ([0.44 1.15905298])\nIncrease radii from ([0.44 1.15905298]) to ([0.484 1.27495828])\nIncrease radii from ([0.484 1.27495828]) to ([0.5324 1.40245411])\nIncrease radii from ([0.5324 1.40245411]) to ([0.58564 1.54269952])\nIncrease radii from ([0.58564 1.54269952]) to ([0.644204 1.69696947])\nIncrease radii from ([0.35 0.93940064]) to ([0.385 1.0333407])\nIncrease radii from ([0.385 1.0333407]) to ([0.4235 1.13667477])\nIncrease radii from ([0.4235 1.13667477]) to ([0.46585 1.25034225])\nIncrease radii from ([0.46585 1.25034225]) to ([0.512435 1.37537648])\nIncrease radii from ([0.512435 1.37537648]) to ([0.5636785 1.51291412])\nIncrease radii from ([0.5636785 1.51291412]) to ([0.62004635 1.66420554])\nIncrease radii from ([0.35 0.92197396]) to ([0.385 1.01417136])\nIncrease radii from ([0.385 1.01417136]) to ([0.4235 1.1155885])\nIncrease radii from ([0.4235 1.1155885]) to ([0.46585 1.22714735])\nIncrease radii from ([0.46585 1.22714735]) to ([0.512435 1.34986208])\nIncrease radii from ([0.512435 1.34986208]) to ([0.5636785 1.48484829])\nIncrease radii from ([0.5636785 1.48484829]) to ([0.62004635 1.63333312])\nIncrease radii from ([0.3 0.80520055]) to ([0.33 0.8857206])\nIncrease radii from ([0.33 0.8857206]) to ([0.363 0.97429266])\nIncrease radii from ([0.363 0.97429266]) to ([0.3993 1.07172193])\nIncrease radii from ([0.3993 1.07172193]) to ([0.43923 1.17889412])\nIncrease radii from ([0.43923 1.17889412]) to ([0.483153 1.29678353])\nIncrease radii from ([0.483153 1.29678353]) to ([0.5314683 1.42646189])\nIncrease radii from ([0.5314683 1.42646189]) to ([0.58461513 1.56910808])\nIncrease radii from ([0.58461513 1.56910808]) to ([0.64307664 1.72601888])\nIncrease radii from ([0.3 0.7902634]) to ([0.33 0.86928974])\nIncrease radii from ([0.33 0.86928974]) to ([0.363 0.95621871])\nIncrease radii from ([0.363 0.95621871]) to ([0.3993 1.05184058])\nIncrease radii from ([0.3993 1.05184058]) to ([0.43923 1.15702464])\nIncrease radii from ([0.43923 1.15702464]) to ([0.483153 1.27272711])\nIncrease radii from ([0.483153 1.27272711]) to ([0.5314683 1.39999982])\nIncrease radii from ([0.5314683 1.39999982]) to ([0.58461513 1.5399998 ])\nIncrease radii from ([0.58461513 1.5399998 ]) to ([0.64307664 1.69399978])\nIncrease radii from ([0.3 1.14207922]) to ([0.33 1.25628714])\nIncrease radii from ([0.33 1.25628714]) to ([0.363 1.38191585])\nIncrease radii from ([0.25 0.93712122]) to ([0.275 1.03083335])\nIncrease radii from ([0.275 1.03083335]) to ([0.3025 1.13391668])\nIncrease radii from ([0.25 0.67100046]) to ([0.275 0.7381005])\nIncrease radii from ([0.275 0.7381005]) to ([0.3025 0.81191055])\nIncrease radii from ([0.3025 0.81191055]) to ([0.33275 0.89310161])\nIncrease radii from ([0.33275 0.89310161]) to ([0.366025 0.98241177])\nIncrease radii from ([0.366025 0.98241177]) to ([0.4026275 1.08065295])\nIncrease radii from ([0.4026275 1.08065295]) to ([0.44289025 1.18871824])\nIncrease radii from ([0.44289025 1.18871824]) to ([0.48717928 1.30759006])\nIncrease radii from ([0.48717928 1.30759006]) to ([0.5358972 1.43834907])\nIncrease radii from ([0.5358972 1.43834907]) to ([0.58948692 1.58218398])\nIncrease radii from ([0.58948692 1.58218398]) to ([0.64843562 1.74040237])\nIncrease radii from ([0.25 0.65855283]) to ([0.275 0.72440811])\nIncrease radii from ([0.275 0.72440811]) to ([0.3025 0.79684893])\nIncrease radii from ([0.3025 0.79684893]) to ([0.33275 0.87653382])\nIncrease radii from ([0.33275 0.87653382]) to ([0.366025 0.9641872])\nIncrease radii from ([0.366025 0.9641872]) to ([0.4026275 1.06060592])\nIncrease radii from ([0.4026275 1.06060592]) to ([0.44289025 1.16666651])\nIncrease radii from ([0.44289025 1.16666651]) to ([0.48717928 1.28333316])\nIncrease radii from ([0.48717928 1.28333316]) to ([0.5358972 1.41166648])\nIncrease radii from ([0.5358972 1.41166648]) to ([0.58948692 1.55283313])\nIncrease radii from ([0.58948692 1.55283313]) to ([0.64843562 1.70811644])\nIncrease radii from ([0.25 0.95173268]) to ([0.275 1.04690595])\nIncrease radii from ([0.275 1.04690595]) to ([0.3025 1.15159654])\nIncrease radii from ([0.3025 1.15159654]) to ([0.33275 1.2667562])\nIncrease radii from ([0.33275 1.2667562]) to ([0.366025 1.39343182])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1748.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.6104011]) to ([0.66 1.7714412])\nIncrease radii from ([0.6 1.5805268]) to ([0.66 1.73857948])\nIncrease radii from ([0.55 1.476201]) to ([0.605 1.6238211])\nIncrease radii from ([0.55 1.44881623]) to ([0.605 1.59369785])\nIncrease radii from ([0.605 1.59369785]) to ([0.6655 1.75306764])\nIncrease radii from ([0.5 1.34200091]) to ([0.55 1.476201])\nIncrease radii from ([0.55 1.476201]) to ([0.605 1.6238211])\nIncrease radii from ([0.5 1.31710566]) to ([0.55 1.44881623])\nIncrease radii from ([0.55 1.44881623]) to ([0.605 1.59369785])\nIncrease radii from ([0.605 1.59369785]) to ([0.6655 1.75306764])\nIncrease radii from ([0.45 1.20780082]) to ([0.495 1.3285809])\nIncrease radii from ([0.495 1.3285809]) to ([0.5445 1.46143899])\nIncrease radii from ([0.5445 1.46143899]) to ([0.59895 1.60758289])\nIncrease radii from ([0.59895 1.60758289]) to ([0.658845 1.76834118])\nIncrease radii from ([0.45 1.1853951]) to ([0.495 1.30393461])\nIncrease radii from ([0.495 1.30393461]) to ([0.5445 1.43432807])\nIncrease radii from ([0.5445 1.43432807]) to ([0.59895 1.57776087])\nIncrease radii from ([0.59895 1.57776087]) to ([0.658845 1.73553696])\nIncrease radii from ([0.4 1.07360073]) to ([0.44 1.1809608])\nIncrease radii from ([0.44 1.1809608]) to ([0.484 1.29905688])\nIncrease radii from ([0.484 1.29905688]) to ([0.5324 1.42896257])\nIncrease radii from ([0.5324 1.42896257]) to ([0.58564 1.57185883])\nIncrease radii from ([0.58564 1.57185883]) to ([0.644204 1.72904471])\nIncrease radii from ([0.4 1.05368453]) to ([0.44 1.15905298])\nIncrease radii from ([0.44 1.15905298]) to ([0.484 1.27495828])\nIncrease radii from ([0.484 1.27495828]) to ([0.5324 1.40245411])\nIncrease radii from ([0.5324 1.40245411]) to ([0.58564 1.54269952])\nIncrease radii from ([0.58564 1.54269952]) to ([0.644204 1.69696947])\nIncrease radii from ([0.35 0.93940064]) to ([0.385 1.0333407])\nIncrease radii from ([0.385 1.0333407]) to ([0.4235 1.13667477])\nIncrease radii from ([0.4235 1.13667477]) to ([0.46585 1.25034225])\nIncrease radii from ([0.46585 1.25034225]) to ([0.512435 1.37537648])\nIncrease radii from ([0.512435 1.37537648]) to ([0.5636785 1.51291412])\nIncrease radii from ([0.5636785 1.51291412]) to ([0.62004635 1.66420554])\nIncrease radii from ([0.35 0.92197396]) to ([0.385 1.01417136])\nIncrease radii from ([0.385 1.01417136]) to ([0.4235 1.1155885])\nIncrease radii from ([0.4235 1.1155885]) to ([0.46585 1.22714735])\nIncrease radii from ([0.46585 1.22714735]) to ([0.512435 1.34986208])\nIncrease radii from ([0.512435 1.34986208]) to ([0.5636785 1.48484829])\nIncrease radii from ([0.5636785 1.48484829]) to ([0.62004635 1.63333312])\nIncrease radii from ([0.3 0.80520055]) to ([0.33 0.8857206])\nIncrease radii from ([0.33 0.8857206]) to ([0.363 0.97429266])\nIncrease radii from ([0.363 0.97429266]) to ([0.3993 1.07172193])\nIncrease radii from ([0.3993 1.07172193]) to ([0.43923 1.17889412])\nIncrease radii from ([0.43923 1.17889412]) to ([0.483153 1.29678353])\nIncrease radii from ([0.483153 1.29678353]) to ([0.5314683 1.42646189])\nIncrease radii from ([0.5314683 1.42646189]) to ([0.58461513 1.56910808])\nIncrease radii from ([0.58461513 1.56910808]) to ([0.64307664 1.72601888])\nIncrease radii from ([0.3 0.7902634]) to ([0.33 0.86928974])\nIncrease radii from ([0.33 0.86928974]) to ([0.363 0.95621871])\nIncrease radii from ([0.363 0.95621871]) to ([0.3993 1.05184058])\nIncrease radii from ([0.3993 1.05184058]) to ([0.43923 1.15702464])\nIncrease radii from ([0.43923 1.15702464]) to ([0.483153 1.27272711])\nIncrease radii from ([0.483153 1.27272711]) to ([0.5314683 1.39999982])\nIncrease radii from ([0.5314683 1.39999982]) to ([0.58461513 1.5399998 ])\nIncrease radii from ([0.58461513 1.5399998 ]) to ([0.64307664 1.69399978])\nIncrease radii from ([0.3 1.14207922]) to ([0.33 1.25628714])\nIncrease radii from ([0.33 1.25628714]) to ([0.363 1.38191585])\nIncrease radii from ([0.25 0.93712122]) to ([0.275 1.03083335])\nIncrease radii from ([0.275 1.03083335]) to ([0.3025 1.13391668])\nIncrease radii from ([0.25 0.67100046]) to ([0.275 0.7381005])\nIncrease radii from ([0.275 0.7381005]) to ([0.3025 0.81191055])\nIncrease radii from ([0.3025 0.81191055]) to ([0.33275 0.89310161])\nIncrease radii from ([0.33275 0.89310161]) to ([0.366025 0.98241177])\nIncrease radii from ([0.366025 0.98241177]) to ([0.4026275 1.08065295])\nIncrease radii from ([0.4026275 1.08065295]) to ([0.44289025 1.18871824])\nIncrease radii from ([0.44289025 1.18871824]) to ([0.48717928 1.30759006])\nIncrease radii from ([0.48717928 1.30759006]) to ([0.5358972 1.43834907])\nIncrease radii from ([0.5358972 1.43834907]) to ([0.58948692 1.58218398])\nIncrease radii from ([0.58948692 1.58218398]) to ([0.64843562 1.74040237])\nIncrease radii from ([0.25 0.65855283]) to ([0.275 0.72440811])\nIncrease radii from ([0.275 0.72440811]) to ([0.3025 0.79684893])\nIncrease radii from ([0.3025 0.79684893]) to ([0.33275 0.87653382])\nIncrease radii from ([0.33275 0.87653382]) to ([0.366025 0.9641872])\nIncrease radii from ([0.366025 0.9641872]) to ([0.4026275 1.06060592])\nIncrease radii from ([0.4026275 1.06060592]) to ([0.44289025 1.16666651])\nIncrease radii from ([0.44289025 1.16666651]) to ([0.48717928 1.28333316])\nIncrease radii from ([0.48717928 1.28333316]) to ([0.5358972 1.41166648])\nIncrease radii from ([0.5358972 1.41166648]) to ([0.58948692 1.55283313])\nIncrease radii from ([0.58948692 1.55283313]) to ([0.64843562 1.70811644])\nIncrease radii from ([0.25 0.95173268]) to ([0.275 1.04690595])\nIncrease radii from ([0.275 1.04690595]) to ([0.3025 1.15159654])\nIncrease radii from ([0.3025 1.15159654]) to ([0.33275 1.2667562])\nIncrease radii from ([0.33275 1.2667562]) to ([0.366025 1.39343182])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1749.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.5805268]) to ([0.66 1.73857948])\nIncrease radii from ([0.55 1.44881623]) to ([0.605 1.59369785])\nIncrease radii from ([0.605 1.59369785]) to ([0.6655 1.75306764])\nIncrease radii from ([0.5 1.31710566]) to ([0.55 1.44881623])\nIncrease radii from ([0.55 1.44881623]) to ([0.605 1.59369785])\nIncrease radii from ([0.605 1.59369785]) to ([0.6655 1.75306764])\nIncrease radii from ([0.45 1.1853951]) to ([0.495 1.30393461])\nIncrease radii from ([0.495 1.30393461]) to ([0.5445 1.43432807])\nIncrease radii from ([0.5445 1.43432807]) to ([0.59895 1.57776087])\nIncrease radii from ([0.59895 1.57776087]) to ([0.658845 1.73553696])\nIncrease radii from ([0.4 1.05368453]) to ([0.44 1.15905298])\nIncrease radii from ([0.44 1.15905298]) to ([0.484 1.27495828])\nIncrease radii from ([0.484 1.27495828]) to ([0.5324 1.40245411])\nIncrease radii from ([0.5324 1.40245411]) to ([0.58564 1.54269952])\nIncrease radii from ([0.58564 1.54269952]) to ([0.644204 1.69696947])\nIncrease radii from ([0.35 0.92197396]) to ([0.385 1.01417136])\nIncrease radii from ([0.385 1.01417136]) to ([0.4235 1.1155885])\nIncrease radii from ([0.4235 1.1155885]) to ([0.46585 1.22714735])\nIncrease radii from ([0.46585 1.22714735]) to ([0.512435 1.34986208])\nIncrease radii from ([0.512435 1.34986208]) to ([0.5636785 1.48484829])\nIncrease radii from ([0.5636785 1.48484829]) to ([0.62004635 1.63333312])\nIncrease radii from ([0.3 1.15612797]) to ([0.33 1.27174076])\nIncrease radii from ([0.33 1.27174076]) to ([0.363 1.39891484])\nIncrease radii from ([0.3 0.7902634]) to ([0.33 0.86928974])\nIncrease radii from ([0.33 0.86928974]) to ([0.363 0.95621871])\nIncrease radii from ([0.363 0.95621871]) to ([0.3993 1.05184058])\nIncrease radii from ([0.3993 1.05184058]) to ([0.43923 1.15702464])\nIncrease radii from ([0.43923 1.15702464]) to ([0.483153 1.27272711])\nIncrease radii from ([0.483153 1.27272711]) to ([0.5314683 1.39999982])\nIncrease radii from ([0.5314683 1.39999982]) to ([0.58461513 1.5399998 ])\nIncrease radii from ([0.58461513 1.5399998 ]) to ([0.64307664 1.69399978])\nIncrease radii from ([0.3 1.14207922]) to ([0.33 1.25628714])\nIncrease radii from ([0.33 1.25628714]) to ([0.363 1.38191585])\nIncrease radii from ([0.25 0.93712122]) to ([0.275 1.03083335])\nIncrease radii from ([0.275 1.03083335]) to ([0.3025 1.13391668])\nIncrease radii from ([0.25 0.96343997]) to ([0.275 1.05978397])\nIncrease radii from ([0.275 1.05978397]) to ([0.3025 1.16576237])\nIncrease radii from ([0.3025 1.16576237]) to ([0.33275 1.2823386])\nIncrease radii from ([0.33275 1.2823386]) to ([0.366025 1.41057247])\nIncrease radii from ([0.25 0.9402826]) to ([0.275 1.03431086])\nIncrease radii from ([0.275 1.03431086]) to ([0.3025 1.13774195])\nIncrease radii from ([0.25 0.65855283]) to ([0.275 0.72440811])\nIncrease radii from ([0.275 0.72440811]) to ([0.3025 0.79684893])\nIncrease radii from ([0.3025 0.79684893]) to ([0.33275 0.87653382])\nIncrease radii from ([0.33275 0.87653382]) to ([0.366025 0.9641872])\nIncrease radii from ([0.366025 0.9641872]) to ([0.4026275 1.06060592])\nIncrease radii from ([0.4026275 1.06060592]) to ([0.44289025 1.16666651])\nIncrease radii from ([0.44289025 1.16666651]) to ([0.48717928 1.28333316])\nIncrease radii from ([0.48717928 1.28333316]) to ([0.5358972 1.41166648])\nIncrease radii from ([0.5358972 1.41166648]) to ([0.58948692 1.55283313])\nIncrease radii from ([0.58948692 1.55283313]) to ([0.64843562 1.70811644])\nIncrease radii from ([0.25 0.95173268]) to ([0.275 1.04690595])\nIncrease radii from ([0.275 1.04690595]) to ([0.3025 1.15159654])\nIncrease radii from ([0.3025 1.15159654]) to ([0.33275 1.2667562])\nIncrease radii from ([0.33275 1.2667562]) to ([0.366025 1.39343182])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1750.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.5805268]) to ([0.66 1.73857948])\nIncrease radii from ([0.55 1.44881623]) to ([0.605 1.59369785])\nIncrease radii from ([0.605 1.59369785]) to ([0.6655 1.75306764])\nIncrease radii from ([0.5 1.31710566]) to ([0.55 1.44881623])\nIncrease radii from ([0.55 1.44881623]) to ([0.605 1.59369785])\nIncrease radii from ([0.605 1.59369785]) to ([0.6655 1.75306764])\nIncrease radii from ([0.45 1.1853951]) to ([0.495 1.30393461])\nIncrease radii from ([0.495 1.30393461]) to ([0.5445 1.43432807])\nIncrease radii from ([0.5445 1.43432807]) to ([0.59895 1.57776087])\nIncrease radii from ([0.59895 1.57776087]) to ([0.658845 1.73553696])\nIncrease radii from ([0.4 1.05368453]) to ([0.44 1.15905298])\nIncrease radii from ([0.44 1.15905298]) to ([0.484 1.27495828])\nIncrease radii from ([0.484 1.27495828]) to ([0.5324 1.40245411])\nIncrease radii from ([0.5324 1.40245411]) to ([0.58564 1.54269952])\nIncrease radii from ([0.58564 1.54269952]) to ([0.644204 1.69696947])\nIncrease radii from ([0.35 0.92197396]) to ([0.385 1.01417136])\nIncrease radii from ([0.385 1.01417136]) to ([0.4235 1.1155885])\nIncrease radii from ([0.4235 1.1155885]) to ([0.46585 1.22714735])\nIncrease radii from ([0.46585 1.22714735]) to ([0.512435 1.34986208])\nIncrease radii from ([0.512435 1.34986208]) to ([0.5636785 1.48484829])\nIncrease radii from ([0.5636785 1.48484829]) to ([0.62004635 1.63333312])\nIncrease radii from ([0.3 1.15612797]) to ([0.33 1.27174076])\nIncrease radii from ([0.33 1.27174076]) to ([0.363 1.39891484])\nIncrease radii from ([0.3 0.7902634]) to ([0.33 0.86928974])\nIncrease radii from ([0.33 0.86928974]) to ([0.363 0.95621871])\nIncrease radii from ([0.363 0.95621871]) to ([0.3993 1.05184058])\nIncrease radii from ([0.3993 1.05184058]) to ([0.43923 1.15702464])\nIncrease radii from ([0.43923 1.15702464]) to ([0.483153 1.27272711])\nIncrease radii from ([0.483153 1.27272711]) to ([0.5314683 1.39999982])\nIncrease radii from ([0.5314683 1.39999982]) to ([0.58461513 1.5399998 ])\nIncrease radii from ([0.58461513 1.5399998 ]) to ([0.64307664 1.69399978])\nIncrease radii from ([0.3 1.14207922]) to ([0.33 1.25628714])\nIncrease radii from ([0.33 1.25628714]) to ([0.363 1.38191585])\nIncrease radii from ([0.25 0.93712122]) to ([0.275 1.03083335])\nIncrease radii from ([0.275 1.03083335]) to ([0.3025 1.13391668])\nIncrease radii from ([0.25 0.96343997]) to ([0.275 1.05978397])\nIncrease radii from ([0.275 1.05978397]) to ([0.3025 1.16576237])\nIncrease radii from ([0.3025 1.16576237]) to ([0.33275 1.2823386])\nIncrease radii from ([0.33275 1.2823386]) to ([0.366025 1.41057247])\nIncrease radii from ([0.25 0.93012415]) to ([0.275 1.02313657])\nIncrease radii from ([0.275 1.02313657]) to ([0.3025 1.12545022])\nIncrease radii from ([0.25 0.65855283]) to ([0.275 0.72440811])\nIncrease radii from ([0.275 0.72440811]) to ([0.3025 0.79684893])\nIncrease radii from ([0.3025 0.79684893]) to ([0.33275 0.87653382])\nIncrease radii from ([0.33275 0.87653382]) to ([0.366025 0.9641872])\nIncrease radii from ([0.366025 0.9641872]) to ([0.4026275 1.06060592])\nIncrease radii from ([0.4026275 1.06060592]) to ([0.44289025 1.16666651])\nIncrease radii from ([0.44289025 1.16666651]) to ([0.48717928 1.28333316])\nIncrease radii from ([0.48717928 1.28333316]) to ([0.5358972 1.41166648])\nIncrease radii from ([0.5358972 1.41166648]) to ([0.58948692 1.55283313])\nIncrease radii from ([0.58948692 1.55283313]) to ([0.64843562 1.70811644])\nIncrease radii from ([0.25 0.95173268]) to ([0.275 1.04690595])\nIncrease radii from ([0.275 1.04690595]) to ([0.3025 1.15159654])\nIncrease radii from ([0.3025 1.15159654]) to ([0.33275 1.2667562])\nIncrease radii from ([0.33275 1.2667562]) to ([0.366025 1.39343182])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1751.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.5805268]) to ([0.66 1.73857948])\nIncrease radii from ([0.55 1.5171163]) to ([0.605 1.66882793])\nIncrease radii from ([0.55 1.44881623]) to ([0.605 1.59369785])\nIncrease radii from ([0.605 1.59369785]) to ([0.6655 1.75306764])\nIncrease radii from ([0.5 1.37919664]) to ([0.55 1.5171163])\nIncrease radii from ([0.55 1.5171163]) to ([0.605 1.66882793])\nIncrease radii from ([0.5 1.31710566]) to ([0.55 1.44881623])\nIncrease radii from ([0.55 1.44881623]) to ([0.605 1.59369785])\nIncrease radii from ([0.605 1.59369785]) to ([0.6655 1.75306764])\nIncrease radii from ([0.45 1.24127697]) to ([0.495 1.36540467])\nIncrease radii from ([0.495 1.36540467]) to ([0.5445 1.50194514])\nIncrease radii from ([0.5445 1.50194514]) to ([0.59895 1.65213965])\nIncrease radii from ([0.45 1.1853951]) to ([0.495 1.30393461])\nIncrease radii from ([0.495 1.30393461]) to ([0.5445 1.43432807])\nIncrease radii from ([0.5445 1.43432807]) to ([0.59895 1.57776087])\nIncrease radii from ([0.59895 1.57776087]) to ([0.658845 1.73553696])\nIncrease radii from ([0.4 1.10335731]) to ([0.44 1.21369304])\nIncrease radii from ([0.44 1.21369304]) to ([0.484 1.33506235])\nIncrease radii from ([0.484 1.33506235]) to ([0.5324 1.46856858])\nIncrease radii from ([0.5324 1.46856858]) to ([0.58564 1.61542544])\nIncrease radii from ([0.4 1.05368453]) to ([0.44 1.15905298])\nIncrease radii from ([0.44 1.15905298]) to ([0.484 1.27495828])\nIncrease radii from ([0.484 1.27495828]) to ([0.5324 1.40245411])\nIncrease radii from ([0.5324 1.40245411]) to ([0.58564 1.54269952])\nIncrease radii from ([0.58564 1.54269952]) to ([0.644204 1.69696947])\nIncrease radii from ([0.35 0.96543765]) to ([0.385 1.06198141])\nIncrease radii from ([0.385 1.06198141]) to ([0.4235 1.16817955])\nIncrease radii from ([0.4235 1.16817955]) to ([0.46585 1.28499751])\nIncrease radii from ([0.46585 1.28499751]) to ([0.512435 1.41349726])\nIncrease radii from ([0.512435 1.41349726]) to ([0.5636785 1.55484698])\nIncrease radii from ([0.5636785 1.55484698]) to ([0.62004635 1.71033168])\nIncrease radii from ([0.35 0.92197396]) to ([0.385 1.01417136])\nIncrease radii from ([0.385 1.01417136]) to ([0.4235 1.1155885])\nIncrease radii from ([0.4235 1.1155885]) to ([0.46585 1.22714735])\nIncrease radii from ([0.46585 1.22714735]) to ([0.512435 1.34986208])\nIncrease radii from ([0.512435 1.34986208]) to ([0.5636785 1.48484829])\nIncrease radii from ([0.5636785 1.48484829]) to ([0.62004635 1.63333312])\nIncrease radii from ([0.3 0.82751798]) to ([0.33 0.91026978])\nIncrease radii from ([0.33 0.91026978]) to ([0.363 1.00129676])\nIncrease radii from ([0.363 1.00129676]) to ([0.3993 1.10142644])\nIncrease radii from ([0.3993 1.10142644]) to ([0.43923 1.21156908])\nIncrease radii from ([0.43923 1.21156908]) to ([0.483153 1.33272599])\nIncrease radii from ([0.483153 1.33272599]) to ([0.5314683 1.46599859])\nIncrease radii from ([0.5314683 1.46599859]) to ([0.58461513 1.61259844])\nIncrease radii from ([0.3 1.15612797]) to ([0.33 1.27174076])\nIncrease radii from ([0.33 1.27174076]) to ([0.363 1.39891484])\nIncrease radii from ([0.3 0.7902634]) to ([0.33 0.86928974])\nIncrease radii from ([0.33 0.86928974]) to ([0.363 0.95621871])\nIncrease radii from ([0.363 0.95621871]) to ([0.3993 1.05184058])\nIncrease radii from ([0.3993 1.05184058]) to ([0.43923 1.15702464])\nIncrease radii from ([0.43923 1.15702464]) to ([0.483153 1.27272711])\nIncrease radii from ([0.483153 1.27272711]) to ([0.5314683 1.39999982])\nIncrease radii from ([0.5314683 1.39999982]) to ([0.58461513 1.5399998 ])\nIncrease radii from ([0.58461513 1.5399998 ]) to ([0.64307664 1.69399978])\nIncrease radii from ([0.3 1.14207922]) to ([0.33 1.25628714])\nIncrease radii from ([0.33 1.25628714]) to ([0.363 1.38191585])\nIncrease radii from ([0.25 0.68959832]) to ([0.275 0.75855815])\nIncrease radii from ([0.275 0.75855815]) to ([0.3025 0.83441397])\nIncrease radii from ([0.3025 0.83441397]) to ([0.33275 0.91785536])\nIncrease radii from ([0.33275 0.91785536]) to ([0.366025 1.0096409])\nIncrease radii from ([0.366025 1.0096409]) to ([0.4026275 1.11060499])\nIncrease radii from ([0.4026275 1.11060499]) to ([0.44289025 1.22166549])\nIncrease radii from ([0.44289025 1.22166549]) to ([0.48717928 1.34383204])\nIncrease radii from ([0.48717928 1.34383204]) to ([0.5358972 1.47821524])\nIncrease radii from ([0.5358972 1.47821524]) to ([0.58948692 1.62603676])\nIncrease radii from ([0.25 0.93712122]) to ([0.275 1.03083335])\nIncrease radii from ([0.275 1.03083335]) to ([0.3025 1.13391668])\nIncrease radii from ([0.25 0.96343997]) to ([0.275 1.05978397])\nIncrease radii from ([0.275 1.05978397]) to ([0.3025 1.16576237])\nIncrease radii from ([0.3025 1.16576237]) to ([0.33275 1.2823386])\nIncrease radii from ([0.33275 1.2823386]) to ([0.366025 1.41057247])\nIncrease radii from ([0.25 0.93012415]) to ([0.275 1.02313657])\nIncrease radii from ([0.275 1.02313657]) to ([0.3025 1.12545022])\nIncrease radii from ([0.25 0.65855283]) to ([0.275 0.72440811])\nIncrease radii from ([0.275 0.72440811]) to ([0.3025 0.79684893])\nIncrease radii from ([0.3025 0.79684893]) to ([0.33275 0.87653382])\nIncrease radii from ([0.33275 0.87653382]) to ([0.366025 0.9641872])\nIncrease radii from ([0.366025 0.9641872]) to ([0.4026275 1.06060592])\nIncrease radii from ([0.4026275 1.06060592]) to ([0.44289025 1.16666651])\nIncrease radii from ([0.44289025 1.16666651]) to ([0.48717928 1.28333316])\nIncrease radii from ([0.48717928 1.28333316]) to ([0.5358972 1.41166648])\nIncrease radii from ([0.5358972 1.41166648]) to ([0.58948692 1.55283313])\nIncrease radii from ([0.58948692 1.55283313]) to ([0.64843562 1.70811644])\nIncrease radii from ([0.25 0.95173268]) to ([0.275 1.04690595])\nIncrease radii from ([0.275 1.04690595]) to ([0.3025 1.15159654])\nIncrease radii from ([0.3025 1.15159654]) to ([0.33275 1.2667562])\nIncrease radii from ([0.33275 1.2667562]) to ([0.366025 1.39343182])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1752.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.5171163]) to ([0.605 1.66882793])\nIncrease radii from ([0.5 1.37919664]) to ([0.55 1.5171163])\nIncrease radii from ([0.55 1.5171163]) to ([0.605 1.66882793])\nIncrease radii from ([0.45 1.24127697]) to ([0.495 1.36540467])\nIncrease radii from ([0.495 1.36540467]) to ([0.5445 1.50194514])\nIncrease radii from ([0.5445 1.50194514]) to ([0.59895 1.65213965])\nIncrease radii from ([0.4 1.10335731]) to ([0.44 1.21369304])\nIncrease radii from ([0.44 1.21369304]) to ([0.484 1.33506235])\nIncrease radii from ([0.484 1.33506235]) to ([0.5324 1.46856858])\nIncrease radii from ([0.5324 1.46856858]) to ([0.58564 1.61542544])\nIncrease radii from ([0.35 0.96543765]) to ([0.385 1.06198141])\nIncrease radii from ([0.385 1.06198141]) to ([0.4235 1.16817955])\nIncrease radii from ([0.4235 1.16817955]) to ([0.46585 1.28499751])\nIncrease radii from ([0.46585 1.28499751]) to ([0.512435 1.41349726])\nIncrease radii from ([0.512435 1.41349726]) to ([0.5636785 1.55484698])\nIncrease radii from ([0.5636785 1.55484698]) to ([0.62004635 1.71033168])\nIncrease radii from ([0.3 1.56774359]) to ([0.33 1.72451795])\nIncrease radii from ([0.3 0.82751798]) to ([0.33 0.91026978])\nIncrease radii from ([0.33 0.91026978]) to ([0.363 1.00129676])\nIncrease radii from ([0.363 1.00129676]) to ([0.3993 1.10142644])\nIncrease radii from ([0.3993 1.10142644]) to ([0.43923 1.21156908])\nIncrease radii from ([0.43923 1.21156908]) to ([0.483153 1.33272599])\nIncrease radii from ([0.483153 1.33272599]) to ([0.5314683 1.46599859])\nIncrease radii from ([0.5314683 1.46599859]) to ([0.58461513 1.61259844])\nIncrease radii from ([0.3 1.14207922]) to ([0.33 1.25628714])\nIncrease radii from ([0.33 1.25628714]) to ([0.363 1.38191585])\nIncrease radii from ([0.25 1.30645299]) to ([0.275 1.43709829])\nIncrease radii from ([0.275 1.43709829]) to ([0.3025 1.58080812])\nIncrease radii from ([0.3025 1.58080812]) to ([0.33275 1.73888893])\nIncrease radii from ([0.25 0.68959832]) to ([0.275 0.75855815])\nIncrease radii from ([0.275 0.75855815]) to ([0.3025 0.83441397])\nIncrease radii from ([0.3025 0.83441397]) to ([0.33275 0.91785536])\nIncrease radii from ([0.33275 0.91785536]) to ([0.366025 1.0096409])\nIncrease radii from ([0.366025 1.0096409]) to ([0.4026275 1.11060499])\nIncrease radii from ([0.4026275 1.11060499]) to ([0.44289025 1.22166549])\nIncrease radii from ([0.44289025 1.22166549]) to ([0.48717928 1.34383204])\nIncrease radii from ([0.48717928 1.34383204]) to ([0.5358972 1.47821524])\nIncrease radii from ([0.5358972 1.47821524]) to ([0.58948692 1.62603676])\nIncrease radii from ([0.44449528 0.7698884 ]) to ([0.4889448 0.84687724])\nIncrease radii from ([0.25 0.95173268]) to ([0.275 1.04690595])\nIncrease radii from ([0.275 1.04690595]) to ([0.3025 1.15159654])\nIncrease radii from ([0.3025 1.15159654]) to ([0.33275 1.2667562])\nIncrease radii from ([0.33275 1.2667562]) to ([0.366025 1.39343182])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1753.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.55 1.50044885]) to ([0.605 1.65049373])\nIncrease radii from ([0.5 1.36404441]) to ([0.55 1.50044885])\nIncrease radii from ([0.55 1.50044885]) to ([0.605 1.65049373])\nIncrease radii from ([0.45 1.22763997]) to ([0.495 1.35040396])\nIncrease radii from ([0.495 1.35040396]) to ([0.5445 1.48544436])\nIncrease radii from ([0.5445 1.48544436]) to ([0.59895 1.6339888])\nIncrease radii from ([0.4 1.09123553]) to ([0.44 1.20035908])\nIncrease radii from ([0.44 1.20035908]) to ([0.484 1.32039499])\nIncrease radii from ([0.484 1.32039499]) to ([0.5324 1.45243449])\nIncrease radii from ([0.5324 1.45243449]) to ([0.58564 1.59767793])\nIncrease radii from ([0.58564 1.59767793]) to ([0.644204 1.75744573])\nIncrease radii from ([0.35 0.95483109]) to ([0.385 1.05031419])\nIncrease radii from ([0.385 1.05031419]) to ([0.4235 1.15534561])\nIncrease radii from ([0.4235 1.15534561]) to ([0.46585 1.27088018])\nIncrease radii from ([0.46585 1.27088018]) to ([0.512435 1.39796819])\nIncrease radii from ([0.512435 1.39796819]) to ([0.5636785 1.53776501])\nIncrease radii from ([0.5636785 1.53776501]) to ([0.62004635 1.69154151])\nIncrease radii from ([0.3 1.56774359]) to ([0.33 1.72451795])\nIncrease radii from ([0.3 0.81842665]) to ([0.33 0.90026931])\nIncrease radii from ([0.33 0.90026931]) to ([0.363 0.99029624])\nIncrease radii from ([0.363 0.99029624]) to ([0.3993 1.08932586])\nIncrease radii from ([0.3993 1.08932586]) to ([0.43923 1.19825845])\nIncrease radii from ([0.43923 1.19825845]) to ([0.483153 1.3180843])\nIncrease radii from ([0.483153 1.3180843]) to ([0.5314683 1.44989273])\nIncrease radii from ([0.5314683 1.44989273]) to ([0.58461513 1.594882 ])\nIncrease radii from ([0.58461513 1.594882 ]) to ([0.64307664 1.7543702 ])\nIncrease radii from ([0.3 1.14570717]) to ([0.33 1.26027789])\nIncrease radii from ([0.33 1.26027789]) to ([0.363 1.38630568])\nIncrease radii from ([0.25 1.30645299]) to ([0.275 1.43709829])\nIncrease radii from ([0.275 1.43709829]) to ([0.3025 1.58080812])\nIncrease radii from ([0.3025 1.58080812]) to ([0.33275 1.73888893])\nIncrease radii from ([0.25 0.6820222]) to ([0.275 0.75022442])\nIncrease radii from ([0.275 0.75022442]) to ([0.3025 0.82524687])\nIncrease radii from ([0.3025 0.82524687]) to ([0.33275 0.90777155])\nIncrease radii from ([0.33275 0.90777155]) to ([0.366025 0.99854871])\nIncrease radii from ([0.366025 0.99854871]) to ([0.4026275 1.09840358])\nIncrease radii from ([0.4026275 1.09840358]) to ([0.44289025 1.20824394])\nIncrease radii from ([0.44289025 1.20824394]) to ([0.48717928 1.32906833])\nIncrease radii from ([0.48717928 1.32906833]) to ([0.5358972 1.46197517])\nIncrease radii from ([0.5358972 1.46197517]) to ([0.58948692 1.60817268])\nIncrease radii from ([0.58948692 1.60817268]) to ([0.64843562 1.76898995])\nIncrease radii from ([0.44449528 0.7698884 ]) to ([0.4889448 0.84687724])\nIncrease radii from ([0.25 0.95475598]) to ([0.275 1.05023157])\nIncrease radii from ([0.275 1.05023157]) to ([0.3025 1.15525473])\nIncrease radii from ([0.3025 1.15525473]) to ([0.33275 1.27078021])\nIncrease radii from ([0.33275 1.27078021]) to ([0.366025 1.39785823])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1754.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.3 1.17529851]) to ([0.33 1.29282836])\nIncrease radii from ([0.3 1.56774359]) to ([0.33 1.72451795])\nIncrease radii from ([0.3 1.14570717]) to ([0.33 1.26027789])\nIncrease radii from ([0.33 1.26027789]) to ([0.363 1.38630568])\nIncrease radii from ([0.25 0.97941543]) to ([0.275 1.07735697])\nIncrease radii from ([0.275 1.07735697]) to ([0.3025 1.18509267])\nIncrease radii from ([0.3025 1.18509267]) to ([0.33275 1.30360193])\nIncrease radii from ([0.25 1.30645299]) to ([0.275 1.43709829])\nIncrease radii from ([0.275 1.43709829]) to ([0.3025 1.58080812])\nIncrease radii from ([0.3025 1.58080812]) to ([0.33275 1.73888893])\nIncrease radii from ([0.25 0.94715103]) to ([0.275 1.04186613])\nIncrease radii from ([0.275 1.04186613]) to ([0.3025 1.14605274])\nIncrease radii from ([0.25 0.95475598]) to ([0.275 1.05023157])\nIncrease radii from ([0.275 1.05023157]) to ([0.3025 1.15525473])\nIncrease radii from ([0.3025 1.15525473]) to ([0.33275 1.27078021])\nIncrease radii from ([0.33275 1.27078021]) to ([0.366025 1.39785823])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1755.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.55303682]) to ([0.66 1.7083405])\nIncrease radii from ([0.55 1.49841346]) to ([0.605 1.6482548])\nIncrease radii from ([0.55 1.47952908]) to ([0.605 1.62748199])\nIncrease radii from ([0.55 1.42361708]) to ([0.605 1.56597879])\nIncrease radii from ([0.605 1.56597879]) to ([0.6655 1.72257667])\nIncrease radii from ([0.5 1.36219405]) to ([0.55 1.49841346])\nIncrease radii from ([0.55 1.49841346]) to ([0.605 1.6482548])\nIncrease radii from ([0.5 1.34502644]) to ([0.55 1.47952908])\nIncrease radii from ([0.55 1.47952908]) to ([0.605 1.62748199])\nIncrease radii from ([0.5 1.29419735]) to ([0.55 1.42361708])\nIncrease radii from ([0.55 1.42361708]) to ([0.605 1.56597879])\nIncrease radii from ([0.605 1.56597879]) to ([0.6655 1.72257667])\nIncrease radii from ([0.45 1.22597465]) to ([0.495 1.34857211])\nIncrease radii from ([0.495 1.34857211]) to ([0.5445 1.48342932])\nIncrease radii from ([0.5445 1.48342932]) to ([0.59895 1.63177225])\nIncrease radii from ([0.45 1.21052379]) to ([0.495 1.33157617])\nIncrease radii from ([0.495 1.33157617]) to ([0.5445 1.46473379])\nIncrease radii from ([0.5445 1.46473379]) to ([0.59895 1.61120717])\nIncrease radii from ([0.59895 1.61120717]) to ([0.658845 1.77232789])\nIncrease radii from ([0.45 1.16477761]) to ([0.495 1.28125537])\nIncrease radii from ([0.495 1.28125537]) to ([0.5445 1.40938091])\nIncrease radii from ([0.5445 1.40938091]) to ([0.59895 1.550319])\nIncrease radii from ([0.59895 1.550319]) to ([0.658845 1.7053509])\nIncrease radii from ([0.4 1.08975524]) to ([0.44 1.19873076])\nIncrease radii from ([0.44 1.19873076]) to ([0.484 1.31860384])\nIncrease radii from ([0.484 1.31860384]) to ([0.5324 1.45046423])\nIncrease radii from ([0.5324 1.45046423]) to ([0.58564 1.59551065])\nIncrease radii from ([0.58564 1.59551065]) to ([0.644204 1.75506171])\nIncrease radii from ([0.4 1.07602115]) to ([0.44 1.18362327])\nIncrease radii from ([0.44 1.18362327]) to ([0.484 1.30198559])\nIncrease radii from ([0.484 1.30198559]) to ([0.5324 1.43218415])\nIncrease radii from ([0.5324 1.43218415]) to ([0.58564 1.57540257])\nIncrease radii from ([0.58564 1.57540257]) to ([0.644204 1.73294282])\nIncrease radii from ([0.4 1.03535788]) to ([0.44 1.13889367])\nIncrease radii from ([0.44 1.13889367]) to ([0.484 1.25278303])\nIncrease radii from ([0.484 1.25278303]) to ([0.5324 1.37806134])\nIncrease radii from ([0.5324 1.37806134]) to ([0.58564 1.51586747])\nIncrease radii from ([0.58564 1.51586747]) to ([0.644204 1.66745422])\nIncrease radii from ([0.35 0.95353584]) to ([0.385 1.04888942])\nIncrease radii from ([0.385 1.04888942]) to ([0.4235 1.15377836])\nIncrease radii from ([0.4235 1.15377836]) to ([0.46585 1.2691562])\nIncrease radii from ([0.46585 1.2691562]) to ([0.512435 1.39607182])\nIncrease radii from ([0.512435 1.39607182]) to ([0.5636785 1.535679 ])\nIncrease radii from ([0.5636785 1.535679 ]) to ([0.62004635 1.6892469 ])\nIncrease radii from ([0.35 0.94151851]) to ([0.385 1.03567036])\nIncrease radii from ([0.385 1.03567036]) to ([0.4235 1.13923739])\nIncrease radii from ([0.4235 1.13923739]) to ([0.46585 1.25316113])\nIncrease radii from ([0.46585 1.25316113]) to ([0.512435 1.37847725])\nIncrease radii from ([0.512435 1.37847725]) to ([0.5636785 1.51632497])\nIncrease radii from ([0.5636785 1.51632497]) to ([0.62004635 1.66795747])\nIncrease radii from ([0.35 0.90593814]) to ([0.385 0.99653196])\nIncrease radii from ([0.385 0.99653196]) to ([0.4235 1.09618515])\nIncrease radii from ([0.4235 1.09618515]) to ([0.46585 1.20580367])\nIncrease radii from ([0.46585 1.20580367]) to ([0.512435 1.32638404])\nIncrease radii from ([0.512435 1.32638404]) to ([0.5636785 1.45902244])\nIncrease radii from ([0.5636785 1.45902244]) to ([0.62004635 1.60492468])\nIncrease radii from ([0.62004635 1.60492468]) to ([0.68205099 1.76541715])\nIncrease radii from ([0.3 0.81731643]) to ([0.33 0.89904807])\nIncrease radii from ([0.33 0.89904807]) to ([0.363 0.98895288])\nIncrease radii from ([0.363 0.98895288]) to ([0.3993 1.08784817])\nIncrease radii from ([0.3993 1.08784817]) to ([0.43923 1.19663299])\nIncrease radii from ([0.43923 1.19663299]) to ([0.483153 1.31629628])\nIncrease radii from ([0.483153 1.31629628]) to ([0.5314683 1.44792591])\nIncrease radii from ([0.5314683 1.44792591]) to ([0.58461513 1.5927185 ])\nIncrease radii from ([0.58461513 1.5927185 ]) to ([0.64307664 1.75199035])\nIncrease radii from ([0.3 1.56774359]) to ([0.33 1.72451795])\nIncrease radii from ([0.3 0.80701586]) to ([0.33 0.88771745])\nIncrease radii from ([0.33 0.88771745]) to ([0.363 0.97648919])\nIncrease radii from ([0.363 0.97648919]) to ([0.3993 1.07413811])\nIncrease radii from ([0.3993 1.07413811]) to ([0.43923 1.18155192])\nIncrease radii from ([0.43923 1.18155192]) to ([0.483153 1.29970712])\nIncrease radii from ([0.483153 1.29970712]) to ([0.5314683 1.42967783])\nIncrease radii from ([0.5314683 1.42967783]) to ([0.58461513 1.57264561])\nIncrease radii from ([0.58461513 1.57264561]) to ([0.64307664 1.72991017])\nIncrease radii from ([0.3 0.77651841]) to ([0.33 0.85417025])\nIncrease radii from ([0.33 0.85417025]) to ([0.363 0.93958727])\nIncrease radii from ([0.363 0.93958727]) to ([0.3993 1.033546])\nIncrease radii from ([0.3993 1.033546]) to ([0.43923 1.1369006])\nIncrease radii from ([0.43923 1.1369006]) to ([0.483153 1.25059066])\nIncrease radii from ([0.483153 1.25059066]) to ([0.5314683 1.37564973])\nIncrease radii from ([0.5314683 1.37564973]) to ([0.58461513 1.5132147 ])\nIncrease radii from ([0.58461513 1.5132147 ]) to ([0.64307664 1.66453617])\nIncrease radii from ([0.3 1.15290214]) to ([0.33 1.26819236])\nIncrease radii from ([0.33 1.26819236]) to ([0.363 1.39501159])\nIncrease radii from ([0.3 1.14570717]) to ([0.33 1.26027789])\nIncrease radii from ([0.33 1.26027789]) to ([0.363 1.38630568])\nIncrease radii from ([0.25 0.68109703]) to ([0.275 0.74920673])\nIncrease radii from ([0.275 0.74920673]) to ([0.3025 0.8241274])\nIncrease radii from ([0.3025 0.8241274]) to ([0.33275 0.90654014])\nIncrease radii from ([0.33275 0.90654014]) to ([0.366025 0.99719415])\nIncrease radii from ([0.366025 0.99719415]) to ([0.4026275 1.09691357])\nIncrease radii from ([0.4026275 1.09691357]) to ([0.44289025 1.20660493])\nIncrease radii from ([0.44289025 1.20660493]) to ([0.48717928 1.32726542])\nIncrease radii from ([0.48717928 1.32726542]) to ([0.5358972 1.45999196])\nIncrease radii from ([0.5358972 1.45999196]) to ([0.58948692 1.60599116])\nIncrease radii from ([0.58948692 1.60599116]) to ([0.64843562 1.76659027])\nIncrease radii from ([0.25 1.30645299]) to ([0.275 1.43709829])\nIncrease radii from ([0.275 1.43709829]) to ([0.3025 1.58080812])\nIncrease radii from ([0.3025 1.58080812]) to ([0.33275 1.73888893])\nIncrease radii from ([0.25 0.67251322]) to ([0.275 0.73976454])\nIncrease radii from ([0.275 0.73976454]) to ([0.3025 0.81374099])\nIncrease radii from ([0.3025 0.81374099]) to ([0.33275 0.89511509])\nIncrease radii from ([0.33275 0.89511509]) to ([0.366025 0.9846266])\nIncrease radii from ([0.366025 0.9846266]) to ([0.4026275 1.08308926])\nIncrease radii from ([0.4026275 1.08308926]) to ([0.44289025 1.19139819])\nIncrease radii from ([0.44289025 1.19139819]) to ([0.48717928 1.31053801])\nIncrease radii from ([0.48717928 1.31053801]) to ([0.5358972 1.44159181])\nIncrease radii from ([0.5358972 1.44159181]) to ([0.58948692 1.58575099])\nIncrease radii from ([0.58948692 1.58575099]) to ([0.64843562 1.74432609])\nIncrease radii from ([0.25 0.64709867]) to ([0.275 0.71180854])\nIncrease radii from ([0.275 0.71180854]) to ([0.3025 0.7829894])\nIncrease radii from ([0.3025 0.7829894]) to ([0.33275 0.86128833])\nIncrease radii from ([0.33275 0.86128833]) to ([0.366025 0.94741717])\nIncrease radii from ([0.366025 0.94741717]) to ([0.4026275 1.04215889])\nIncrease radii from ([0.4026275 1.04215889]) to ([0.44289025 1.14637477])\nIncrease radii from ([0.44289025 1.14637477]) to ([0.48717928 1.26101225])\nIncrease radii from ([0.48717928 1.26101225]) to ([0.5358972 1.38711348])\nIncrease radii from ([0.5358972 1.38711348]) to ([0.58948692 1.52582482])\nIncrease radii from ([0.58948692 1.52582482]) to ([0.64843562 1.67840731])\nIncrease radii from ([0.25 0.96075179]) to ([0.275 1.05682696])\nIncrease radii from ([0.275 1.05682696]) to ([0.3025 1.16250966])\nIncrease radii from ([0.3025 1.16250966]) to ([0.33275 1.27876063])\nIncrease radii from ([0.33275 1.27876063]) to ([0.366025 1.40663669])\nIncrease radii from ([0.25 0.95475598]) to ([0.275 1.05023157])\nIncrease radii from ([0.275 1.05023157]) to ([0.3025 1.15525473])\nIncrease radii from ([0.3025 1.15525473]) to ([0.33275 1.27078021])\nIncrease radii from ([0.33275 1.27078021]) to ([0.366025 1.39785823])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1756.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.56895651]) to ([0.66 1.72585216])\nIncrease radii from ([0.55 1.43821013]) to ([0.605 1.58203114])\nIncrease radii from ([0.605 1.58203114]) to ([0.6655 1.74023426])\nIncrease radii from ([0.5 1.30746375]) to ([0.55 1.43821013])\nIncrease radii from ([0.55 1.43821013]) to ([0.605 1.58203114])\nIncrease radii from ([0.605 1.58203114]) to ([0.6655 1.74023426])\nIncrease radii from ([0.45 1.17671738]) to ([0.495 1.29438912])\nIncrease radii from ([0.495 1.29438912]) to ([0.5445 1.42382803])\nIncrease radii from ([0.5445 1.42382803]) to ([0.59895 1.56621083])\nIncrease radii from ([0.59895 1.56621083]) to ([0.658845 1.72283191])\nIncrease radii from ([0.4 1.045971]) to ([0.44 1.1505681])\nIncrease radii from ([0.44 1.1505681]) to ([0.484 1.26562491])\nIncrease radii from ([0.484 1.26562491]) to ([0.5324 1.39218741])\nIncrease radii from ([0.5324 1.39218741]) to ([0.58564 1.53140615])\nIncrease radii from ([0.58564 1.53140615]) to ([0.644204 1.68454676])\nIncrease radii from ([0.35 0.91522463]) to ([0.385 1.00674709])\nIncrease radii from ([0.385 1.00674709]) to ([0.4235 1.1074218])\nIncrease radii from ([0.4235 1.1074218]) to ([0.46585 1.21816398])\nIncrease radii from ([0.46585 1.21816398]) to ([0.512435 1.33998038])\nIncrease radii from ([0.512435 1.33998038]) to ([0.5636785 1.47397842])\nIncrease radii from ([0.5636785 1.47397842]) to ([0.62004635 1.62137626])\nIncrease radii from ([0.3 1.5598772]) to ([0.33 1.71586492])\nIncrease radii from ([0.3 0.78447825]) to ([0.33 0.86292608])\nIncrease radii from ([0.33 0.86292608]) to ([0.363 0.94921869])\nIncrease radii from ([0.363 0.94921869]) to ([0.3993 1.04414055])\nIncrease radii from ([0.3993 1.04414055]) to ([0.43923 1.14855461])\nIncrease radii from ([0.43923 1.14855461]) to ([0.483153 1.26341007])\nIncrease radii from ([0.483153 1.26341007]) to ([0.5314683 1.38975108])\nIncrease radii from ([0.5314683 1.38975108]) to ([0.58461513 1.52872619])\nIncrease radii from ([0.58461513 1.52872619]) to ([0.64307664 1.6815988 ])\nIncrease radii from ([0.3 1.14570717]) to ([0.33 1.26027789])\nIncrease radii from ([0.33 1.26027789]) to ([0.363 1.38630568])\nIncrease radii from ([0.25 1.29989767]) to ([0.275 1.42988743])\nIncrease radii from ([0.275 1.42988743]) to ([0.3025 1.57287618])\nIncrease radii from ([0.3025 1.57287618]) to ([0.33275 1.73016379])\nIncrease radii from ([0.25 0.65373188]) to ([0.275 0.71910507])\nIncrease radii from ([0.275 0.71910507]) to ([0.3025 0.79101557])\nIncrease radii from ([0.3025 0.79101557]) to ([0.33275 0.87011713])\nIncrease radii from ([0.33275 0.87011713]) to ([0.366025 0.95712884])\nIncrease radii from ([0.366025 0.95712884]) to ([0.4026275 1.05284173])\nIncrease radii from ([0.4026275 1.05284173]) to ([0.44289025 1.1581259 ])\nIncrease radii from ([0.44289025 1.1581259 ]) to ([0.48717928 1.27393849])\nIncrease radii from ([0.48717928 1.27393849]) to ([0.5358972 1.40133234])\nIncrease radii from ([0.5358972 1.40133234]) to ([0.58948692 1.54146557])\nIncrease radii from ([0.58948692 1.54146557]) to ([0.64843562 1.69561213])\nIncrease radii from ([0.25 0.95475598]) to ([0.275 1.05023157])\nIncrease radii from ([0.275 1.05023157]) to ([0.3025 1.15525473])\nIncrease radii from ([0.3025 1.15525473]) to ([0.33275 1.27078021])\nIncrease radii from ([0.33275 1.27078021]) to ([0.366025 1.39785823])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1757.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.58722049]) to ([0.66 1.74594254])\nIncrease radii from ([0.55 1.49576825]) to ([0.605 1.64534507])\nIncrease radii from ([0.55 1.45495212]) to ([0.605 1.60044733])\nIncrease radii from ([0.605 1.60044733]) to ([0.6655 1.76049206])\nIncrease radii from ([0.5 1.35978932]) to ([0.55 1.49576825])\nIncrease radii from ([0.55 1.49576825]) to ([0.605 1.64534507])\nIncrease radii from ([0.5 1.32268374]) to ([0.55 1.45495212])\nIncrease radii from ([0.55 1.45495212]) to ([0.605 1.60044733])\nIncrease radii from ([0.605 1.60044733]) to ([0.6655 1.76049206])\nIncrease radii from ([0.45 1.22381038]) to ([0.495 1.34619142])\nIncrease radii from ([0.495 1.34619142]) to ([0.5445 1.48081057])\nIncrease radii from ([0.5445 1.48081057]) to ([0.59895 1.62889162])\nIncrease radii from ([0.45 1.19041537]) to ([0.495 1.30945691])\nIncrease radii from ([0.495 1.30945691]) to ([0.5445 1.4404026])\nIncrease radii from ([0.5445 1.4404026]) to ([0.59895 1.58444286])\nIncrease radii from ([0.59895 1.58444286]) to ([0.658845 1.74288714])\nIncrease radii from ([0.4 1.08783145]) to ([0.44 1.1966146])\nIncrease radii from ([0.44 1.1966146]) to ([0.484 1.31627606])\nIncrease radii from ([0.484 1.31627606]) to ([0.5324 1.44790366])\nIncrease radii from ([0.5324 1.44790366]) to ([0.58564 1.59269403])\nIncrease radii from ([0.58564 1.59269403]) to ([0.644204 1.75196343])\nIncrease radii from ([0.4 1.058147]) to ([0.44 1.16396169])\nIncrease radii from ([0.44 1.16396169]) to ([0.484 1.28035786])\nIncrease radii from ([0.484 1.28035786]) to ([0.5324 1.40839365])\nIncrease radii from ([0.5324 1.40839365]) to ([0.58564 1.54923302])\nIncrease radii from ([0.58564 1.54923302]) to ([0.644204 1.70415632])\nIncrease radii from ([0.35 0.95185252]) to ([0.385 1.04703777])\nIncrease radii from ([0.385 1.04703777]) to ([0.4235 1.15174155])\nIncrease radii from ([0.4235 1.15174155]) to ([0.46585 1.26691571])\nIncrease radii from ([0.46585 1.26691571]) to ([0.512435 1.39360728])\nIncrease radii from ([0.512435 1.39360728]) to ([0.5636785 1.532968 ])\nIncrease radii from ([0.5636785 1.532968 ]) to ([0.62004635 1.6862648 ])\nIncrease radii from ([0.35 0.92587862]) to ([0.385 1.01846648])\nIncrease radii from ([0.385 1.01846648]) to ([0.4235 1.12031313])\nIncrease radii from ([0.4235 1.12031313]) to ([0.46585 1.23234444])\nIncrease radii from ([0.46585 1.23234444]) to ([0.512435 1.35557889])\nIncrease radii from ([0.512435 1.35557889]) to ([0.5636785 1.49113678])\nIncrease radii from ([0.5636785 1.49113678]) to ([0.62004635 1.64025046])\nIncrease radii from ([0.3 0.81587359]) to ([0.33 0.89746095])\nIncrease radii from ([0.33 0.89746095]) to ([0.363 0.98720704])\nIncrease radii from ([0.363 0.98720704]) to ([0.3993 1.08592775])\nIncrease radii from ([0.3993 1.08592775]) to ([0.43923 1.19452052])\nIncrease radii from ([0.43923 1.19452052]) to ([0.483153 1.31397258])\nIncrease radii from ([0.483153 1.31397258]) to ([0.5314683 1.44536983])\nIncrease radii from ([0.5314683 1.44536983]) to ([0.58461513 1.58990682])\nIncrease radii from ([0.58461513 1.58990682]) to ([0.64307664 1.7488975 ])\nIncrease radii from ([0.3 1.58690791]) to ([0.33 1.7455987])\nIncrease radii from ([0.3 0.79361025]) to ([0.33 0.87297127])\nIncrease radii from ([0.33 0.87297127]) to ([0.363 0.9602684])\nIncrease radii from ([0.363 0.9602684]) to ([0.3993 1.05629524])\nIncrease radii from ([0.3993 1.05629524]) to ([0.43923 1.16192476])\nIncrease radii from ([0.43923 1.16192476]) to ([0.483153 1.27811724])\nIncrease radii from ([0.483153 1.27811724]) to ([0.5314683 1.40592896])\nIncrease radii from ([0.5314683 1.40592896]) to ([0.58461513 1.54652186])\nIncrease radii from ([0.58461513 1.54652186]) to ([0.64307664 1.70117404])\nIncrease radii from ([0.3 1.14570717]) to ([0.33 1.26027789])\nIncrease radii from ([0.33 1.26027789]) to ([0.363 1.38630568])\nIncrease radii from ([0.25 0.67989466]) to ([0.275 0.74788412])\nIncrease radii from ([0.275 0.74788412]) to ([0.3025 0.82267254])\nIncrease radii from ([0.3025 0.82267254]) to ([0.33275 0.90493979])\nIncrease radii from ([0.33275 0.90493979]) to ([0.366025 0.99543377])\nIncrease radii from ([0.366025 0.99543377]) to ([0.4026275 1.09497715])\nIncrease radii from ([0.4026275 1.09497715]) to ([0.44289025 1.20447486])\nIncrease radii from ([0.44289025 1.20447486]) to ([0.48717928 1.32492235])\nIncrease radii from ([0.48717928 1.32492235]) to ([0.5358972 1.45741458])\nIncrease radii from ([0.5358972 1.45741458]) to ([0.58948692 1.60315604])\nIncrease radii from ([0.58948692 1.60315604]) to ([0.64843562 1.76347164])\nIncrease radii from ([0.25 1.32242326]) to ([0.275 1.45466558])\nIncrease radii from ([0.275 1.45466558]) to ([0.3025 1.60013214])\nIncrease radii from ([0.3025 1.60013214]) to ([0.33275 1.76014535])\nIncrease radii from ([0.25 0.66134187]) to ([0.275 0.72747606])\nIncrease radii from ([0.275 0.72747606]) to ([0.3025 0.80022367])\nIncrease radii from ([0.3025 0.80022367]) to ([0.33275 0.88024603])\nIncrease radii from ([0.33275 0.88024603]) to ([0.366025 0.96827063])\nIncrease radii from ([0.366025 0.96827063]) to ([0.4026275 1.0650977])\nIncrease radii from ([0.4026275 1.0650977]) to ([0.44289025 1.17160747])\nIncrease radii from ([0.44289025 1.17160747]) to ([0.48717928 1.28876822])\nIncrease radii from ([0.48717928 1.28876822]) to ([0.5358972 1.41764504])\nIncrease radii from ([0.5358972 1.41764504]) to ([0.58948692 1.55940954])\nIncrease radii from ([0.58948692 1.55940954]) to ([0.64843562 1.71535049])\nIncrease radii from ([0.25 0.95475598]) to ([0.275 1.05023157])\nIncrease radii from ([0.275 1.05023157]) to ([0.3025 1.15525473])\nIncrease radii from ([0.3025 1.15525473]) to ([0.33275 1.27078021])\nIncrease radii from ([0.33275 1.27078021]) to ([0.366025 1.39785823])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1758.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.59407624]) to ([0.66 1.75348386])\nIncrease radii from ([0.6 1.57354845]) to ([0.66 1.7309033])\nIncrease radii from ([0.55 1.46123655]) to ([0.605 1.6073602])\nIncrease radii from ([0.605 1.6073602]) to ([0.6655 1.76809622])\nIncrease radii from ([0.55 1.44241942]) to ([0.605 1.58666136])\nIncrease radii from ([0.605 1.58666136]) to ([0.6655 1.74532749])\nIncrease radii from ([0.5 1.32839686]) to ([0.55 1.46123655])\nIncrease radii from ([0.55 1.46123655]) to ([0.605 1.6073602])\nIncrease radii from ([0.605 1.6073602]) to ([0.6655 1.76809622])\nIncrease radii from ([0.5 1.31129038]) to ([0.55 1.44241942])\nIncrease radii from ([0.55 1.44241942]) to ([0.605 1.58666136])\nIncrease radii from ([0.605 1.58666136]) to ([0.6655 1.74532749])\nIncrease radii from ([0.45 1.19555718]) to ([0.495 1.31511289])\nIncrease radii from ([0.495 1.31511289]) to ([0.5445 1.44662418])\nIncrease radii from ([0.5445 1.44662418]) to ([0.59895 1.5912866])\nIncrease radii from ([0.59895 1.5912866]) to ([0.658845 1.75041526])\nIncrease radii from ([0.45 1.18016134]) to ([0.495 1.29817747])\nIncrease radii from ([0.495 1.29817747]) to ([0.5445 1.42799522])\nIncrease radii from ([0.5445 1.42799522]) to ([0.59895 1.57079474])\nIncrease radii from ([0.59895 1.57079474]) to ([0.658845 1.72787422])\nIncrease radii from ([0.4 1.06271749]) to ([0.44 1.16898924])\nIncrease radii from ([0.44 1.16898924]) to ([0.484 1.28588816])\nIncrease radii from ([0.484 1.28588816]) to ([0.5324 1.41447698])\nIncrease radii from ([0.5324 1.41447698]) to ([0.58564 1.55592468])\nIncrease radii from ([0.58564 1.55592468]) to ([0.644204 1.71151715])\nIncrease radii from ([0.4 1.0490323]) to ([0.44 1.15393553])\nIncrease radii from ([0.44 1.15393553]) to ([0.484 1.26932909])\nIncrease radii from ([0.484 1.26932909]) to ([0.5324 1.39626199])\nIncrease radii from ([0.5324 1.39626199]) to ([0.58564 1.53588819])\nIncrease radii from ([0.58564 1.53588819]) to ([0.644204 1.68947701])\nIncrease radii from ([0.35 0.9298778]) to ([0.385 1.02286558])\nIncrease radii from ([0.385 1.02286558]) to ([0.4235 1.12515214])\nIncrease radii from ([0.4235 1.12515214]) to ([0.46585 1.23766736])\nIncrease radii from ([0.46585 1.23766736]) to ([0.512435 1.36143409])\nIncrease radii from ([0.512435 1.36143409]) to ([0.5636785 1.4975775])\nIncrease radii from ([0.5636785 1.4975775]) to ([0.62004635 1.64733525])\nIncrease radii from ([0.35 0.91790326]) to ([0.385 1.00969359])\nIncrease radii from ([0.385 1.00969359]) to ([0.4235 1.11066295])\nIncrease radii from ([0.4235 1.11066295]) to ([0.46585 1.22172924])\nIncrease radii from ([0.46585 1.22172924]) to ([0.512435 1.34390217])\nIncrease radii from ([0.512435 1.34390217]) to ([0.5636785 1.47829239])\nIncrease radii from ([0.5636785 1.47829239]) to ([0.62004635 1.62612162])\nIncrease radii from ([0.3 0.79703812]) to ([0.33 0.87674193])\nIncrease radii from ([0.33 0.87674193]) to ([0.363 0.96441612])\nIncrease radii from ([0.363 0.96441612]) to ([0.3993 1.06085773])\nIncrease radii from ([0.3993 1.06085773]) to ([0.43923 1.16694351])\nIncrease radii from ([0.43923 1.16694351]) to ([0.483153 1.28363786])\nIncrease radii from ([0.483153 1.28363786]) to ([0.5314683 1.41200164])\nIncrease radii from ([0.5314683 1.41200164]) to ([0.58461513 1.55320181])\nIncrease radii from ([0.58461513 1.55320181]) to ([0.64307664 1.70852199])\nIncrease radii from ([0.3 0.78677423]) to ([0.33 0.86545165])\nIncrease radii from ([0.33 0.86545165]) to ([0.363 0.95199681])\nIncrease radii from ([0.363 0.95199681]) to ([0.3993 1.0471965])\nIncrease radii from ([0.3993 1.0471965]) to ([0.43923 1.15191614])\nIncrease radii from ([0.43923 1.15191614]) to ([0.483153 1.26710776])\nIncrease radii from ([0.483153 1.26710776]) to ([0.5314683 1.39381854])\nIncrease radii from ([0.5314683 1.39381854]) to ([0.58461513 1.53320039])\nIncrease radii from ([0.58461513 1.53320039]) to ([0.64307664 1.68652043])\nIncrease radii from ([0.25 0.66419843]) to ([0.275 0.73061827])\nIncrease radii from ([0.275 0.73061827]) to ([0.3025 0.8036801])\nIncrease radii from ([0.3025 0.8036801]) to ([0.33275 0.88404811])\nIncrease radii from ([0.33275 0.88404811]) to ([0.366025 0.97245292])\nIncrease radii from ([0.366025 0.97245292]) to ([0.4026275 1.06969822])\nIncrease radii from ([0.4026275 1.06969822]) to ([0.44289025 1.17666804])\nIncrease radii from ([0.44289025 1.17666804]) to ([0.48717928 1.29433484])\nIncrease radii from ([0.48717928 1.29433484]) to ([0.5358972 1.42376833])\nIncrease radii from ([0.5358972 1.42376833]) to ([0.58948692 1.56614516])\nIncrease radii from ([0.58948692 1.56614516]) to ([0.64843562 1.72275967])\nIncrease radii from ([0.25 0.65564519]) to ([0.275 0.72120971])\nIncrease radii from ([0.275 0.72120971]) to ([0.3025 0.79333068])\nIncrease radii from ([0.3025 0.79333068]) to ([0.33275 0.87266375])\nIncrease radii from ([0.33275 0.87266375]) to ([0.366025 0.95993012])\nIncrease radii from ([0.366025 0.95993012]) to ([0.4026275 1.05592313])\nIncrease radii from ([0.4026275 1.05592313]) to ([0.44289025 1.16151545])\nIncrease radii from ([0.44289025 1.16151545]) to ([0.48717928 1.27766699])\nIncrease radii from ([0.48717928 1.27766699]) to ([0.5358972 1.40543369])\nIncrease radii from ([0.5358972 1.40543369]) to ([0.58948692 1.54597706])\nIncrease radii from ([0.58948692 1.54597706]) to ([0.64843562 1.70057476])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1759.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.59407624]) to ([0.66 1.75348386])\nIncrease radii from ([0.6 1.5676508]) to ([0.66 1.72441588])\nIncrease radii from ([0.55 1.46123655]) to ([0.605 1.6073602])\nIncrease radii from ([0.605 1.6073602]) to ([0.6655 1.76809622])\nIncrease radii from ([0.55 1.43701323]) to ([0.605 1.58071455])\nIncrease radii from ([0.605 1.58071455]) to ([0.6655 1.73878601])\nIncrease radii from ([0.5 1.32839686]) to ([0.55 1.46123655])\nIncrease radii from ([0.55 1.46123655]) to ([0.605 1.6073602])\nIncrease radii from ([0.605 1.6073602]) to ([0.6655 1.76809622])\nIncrease radii from ([0.5 1.30637567]) to ([0.55 1.43701323])\nIncrease radii from ([0.55 1.43701323]) to ([0.605 1.58071455])\nIncrease radii from ([0.605 1.58071455]) to ([0.6655 1.73878601])\nIncrease radii from ([0.45 1.19555718]) to ([0.495 1.31511289])\nIncrease radii from ([0.495 1.31511289]) to ([0.5445 1.44662418])\nIncrease radii from ([0.5445 1.44662418]) to ([0.59895 1.5912866])\nIncrease radii from ([0.59895 1.5912866]) to ([0.658845 1.75041526])\nIncrease radii from ([0.45 1.1757381]) to ([0.495 1.29331191])\nIncrease radii from ([0.495 1.29331191]) to ([0.5445 1.4226431])\nIncrease radii from ([0.5445 1.4226431]) to ([0.59895 1.56490741])\nIncrease radii from ([0.59895 1.56490741]) to ([0.658845 1.72139815])\nIncrease radii from ([0.4 1.06271749]) to ([0.44 1.16898924])\nIncrease radii from ([0.44 1.16898924]) to ([0.484 1.28588816])\nIncrease radii from ([0.484 1.28588816]) to ([0.5324 1.41447698])\nIncrease radii from ([0.5324 1.41447698]) to ([0.58564 1.55592468])\nIncrease radii from ([0.58564 1.55592468]) to ([0.644204 1.71151715])\nIncrease radii from ([0.4 1.04510053]) to ([0.44 1.14961059])\nIncrease radii from ([0.44 1.14961059]) to ([0.484 1.26457164])\nIncrease radii from ([0.484 1.26457164]) to ([0.5324 1.39102881])\nIncrease radii from ([0.5324 1.39102881]) to ([0.58564 1.53013169])\nIncrease radii from ([0.58564 1.53013169]) to ([0.644204 1.68314486])\nIncrease radii from ([0.35 0.9298778]) to ([0.385 1.02286558])\nIncrease radii from ([0.385 1.02286558]) to ([0.4235 1.12515214])\nIncrease radii from ([0.4235 1.12515214]) to ([0.46585 1.23766736])\nIncrease radii from ([0.46585 1.23766736]) to ([0.512435 1.36143409])\nIncrease radii from ([0.512435 1.36143409]) to ([0.5636785 1.4975775])\nIncrease radii from ([0.5636785 1.4975775]) to ([0.62004635 1.64733525])\nIncrease radii from ([0.35 0.91446297]) to ([0.385 1.00590926])\nIncrease radii from ([0.385 1.00590926]) to ([0.4235 1.10650019])\nIncrease radii from ([0.4235 1.10650019]) to ([0.46585 1.21715021])\nIncrease radii from ([0.46585 1.21715021]) to ([0.512435 1.33886523])\nIncrease radii from ([0.512435 1.33886523]) to ([0.5636785 1.47275175])\nIncrease radii from ([0.5636785 1.47275175]) to ([0.62004635 1.62002693])\nIncrease radii from ([0.3 0.79703812]) to ([0.33 0.87674193])\nIncrease radii from ([0.33 0.87674193]) to ([0.363 0.96441612])\nIncrease radii from ([0.363 0.96441612]) to ([0.3993 1.06085773])\nIncrease radii from ([0.3993 1.06085773]) to ([0.43923 1.16694351])\nIncrease radii from ([0.43923 1.16694351]) to ([0.483153 1.28363786])\nIncrease radii from ([0.483153 1.28363786]) to ([0.5314683 1.41200164])\nIncrease radii from ([0.5314683 1.41200164]) to ([0.58461513 1.55320181])\nIncrease radii from ([0.58461513 1.55320181]) to ([0.64307664 1.70852199])\nIncrease radii from ([0.3 0.7838254]) to ([0.33 0.86220794])\nIncrease radii from ([0.33 0.86220794]) to ([0.363 0.94842873])\nIncrease radii from ([0.363 0.94842873]) to ([0.3993 1.04327161])\nIncrease radii from ([0.3993 1.04327161]) to ([0.43923 1.14759877])\nIncrease radii from ([0.43923 1.14759877]) to ([0.483153 1.26235864])\nIncrease radii from ([0.483153 1.26235864]) to ([0.5314683 1.38859451])\nIncrease radii from ([0.5314683 1.38859451]) to ([0.58461513 1.52745396])\nIncrease radii from ([0.58461513 1.52745396]) to ([0.64307664 1.68019935])\nIncrease radii from ([0.25 0.66419843]) to ([0.275 0.73061827])\nIncrease radii from ([0.275 0.73061827]) to ([0.3025 0.8036801])\nIncrease radii from ([0.3025 0.8036801]) to ([0.33275 0.88404811])\nIncrease radii from ([0.33275 0.88404811]) to ([0.366025 0.97245292])\nIncrease radii from ([0.366025 0.97245292]) to ([0.4026275 1.06969822])\nIncrease radii from ([0.4026275 1.06969822]) to ([0.44289025 1.17666804])\nIncrease radii from ([0.44289025 1.17666804]) to ([0.48717928 1.29433484])\nIncrease radii from ([0.48717928 1.29433484]) to ([0.5358972 1.42376833])\nIncrease radii from ([0.5358972 1.42376833]) to ([0.58948692 1.56614516])\nIncrease radii from ([0.58948692 1.56614516]) to ([0.64843562 1.72275967])\nIncrease radii from ([0.25 0.65318783]) to ([0.275 0.71850662])\nIncrease radii from ([0.275 0.71850662]) to ([0.3025 0.79035728])\nIncrease radii from ([0.3025 0.79035728]) to ([0.33275 0.86939301])\nIncrease radii from ([0.33275 0.86939301]) to ([0.366025 0.95633231])\nIncrease radii from ([0.366025 0.95633231]) to ([0.4026275 1.05196554])\nIncrease radii from ([0.4026275 1.05196554]) to ([0.44289025 1.15716209])\nIncrease radii from ([0.44289025 1.15716209]) to ([0.48717928 1.2728783 ])\nIncrease radii from ([0.48717928 1.2728783 ]) to ([0.5358972 1.40016613])\nIncrease radii from ([0.5358972 1.40016613]) to ([0.58948692 1.54018274])\nIncrease radii from ([0.58948692 1.54018274]) to ([0.64843562 1.69420102])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1760.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.5891151]) to ([0.66 1.74802661])\nIncrease radii from ([0.55 1.45668884]) to ([0.605 1.60235773])\nIncrease radii from ([0.605 1.60235773]) to ([0.6655 1.7625935])\nIncrease radii from ([0.5 1.32426258]) to ([0.55 1.45668884])\nIncrease radii from ([0.55 1.45668884]) to ([0.605 1.60235773])\nIncrease radii from ([0.605 1.60235773]) to ([0.6655 1.7625935])\nIncrease radii from ([0.45 1.19183633]) to ([0.495 1.31101996])\nIncrease radii from ([0.495 1.31101996]) to ([0.5445 1.44212195])\nIncrease radii from ([0.5445 1.44212195]) to ([0.59895 1.58633415])\nIncrease radii from ([0.59895 1.58633415]) to ([0.658845 1.74496756])\nIncrease radii from ([0.4 1.05941007]) to ([0.44 1.16535107])\nIncrease radii from ([0.44 1.16535107]) to ([0.484 1.28188618])\nIncrease radii from ([0.484 1.28188618]) to ([0.5324 1.4100748])\nIncrease radii from ([0.5324 1.4100748]) to ([0.58564 1.55108228])\nIncrease radii from ([0.58564 1.55108228]) to ([0.644204 1.70619051])\nIncrease radii from ([0.35 0.92698381]) to ([0.385 1.01968219])\nIncrease radii from ([0.385 1.01968219]) to ([0.4235 1.12165041])\nIncrease radii from ([0.4235 1.12165041]) to ([0.46585 1.23381545])\nIncrease radii from ([0.46585 1.23381545]) to ([0.512435 1.35719699])\nIncrease radii from ([0.512435 1.35719699]) to ([0.5636785 1.49291669])\nIncrease radii from ([0.5636785 1.49291669]) to ([0.62004635 1.64220836])\nIncrease radii from ([0.3 0.79455755]) to ([0.33 0.87401331])\nIncrease radii from ([0.33 0.87401331]) to ([0.363 0.96141464])\nIncrease radii from ([0.363 0.96141464]) to ([0.3993 1.0575561])\nIncrease radii from ([0.3993 1.0575561]) to ([0.43923 1.16331171])\nIncrease radii from ([0.43923 1.16331171]) to ([0.483153 1.27964288])\nIncrease radii from ([0.483153 1.27964288]) to ([0.5314683 1.40760717])\nIncrease radii from ([0.5314683 1.40760717]) to ([0.58461513 1.54836789])\nIncrease radii from ([0.58461513 1.54836789]) to ([0.64307664 1.70320467])\nIncrease radii from ([0.25 0.66213129]) to ([0.275 0.72834442])\nIncrease radii from ([0.275 0.72834442]) to ([0.3025 0.80117886])\nIncrease radii from ([0.3025 0.80117886]) to ([0.33275 0.88129675])\nIncrease radii from ([0.33275 0.88129675]) to ([0.366025 0.96942642])\nIncrease radii from ([0.366025 0.96942642]) to ([0.4026275 1.06636907])\nIncrease radii from ([0.4026275 1.06636907]) to ([0.44289025 1.17300597])\nIncrease radii from ([0.44289025 1.17300597]) to ([0.48717928 1.29030657])\nIncrease radii from ([0.48717928 1.29030657]) to ([0.5358972 1.41933723])\nIncrease radii from ([0.5358972 1.41933723]) to ([0.58948692 1.56127095])\nIncrease radii from ([0.58948692 1.56127095]) to ([0.64843562 1.71739805])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1761.tif is a low contrast image\n compress=6,\nIncrease radii from ([0.6 1.58802253]) to ([0.66 1.74682479])\nIncrease radii from ([0.55 1.45568732]) to ([0.605 1.60125605])\nIncrease radii from ([0.605 1.60125605]) to ([0.6655 1.76138166])\nIncrease radii from ([0.5 1.32335211]) to ([0.55 1.45568732])\nIncrease radii from ([0.55 1.45568732]) to ([0.605 1.60125605])\nIncrease radii from ([0.605 1.60125605]) to ([0.6655 1.76138166])\nIncrease radii from ([0.45 1.1910169]) to ([0.495 1.31011859])\nIncrease radii from ([0.495 1.31011859]) to ([0.5445 1.44113045])\nIncrease radii from ([0.5445 1.44113045]) to ([0.59895 1.58524349])\nIncrease radii from ([0.59895 1.58524349]) to ([0.658845 1.74376784])\nIncrease radii from ([0.4 1.05868169]) to ([0.44 1.16454986])\nIncrease radii from ([0.44 1.16454986]) to ([0.484 1.28100484])\nIncrease radii from ([0.484 1.28100484]) to ([0.5324 1.40910533])\nIncrease radii from ([0.5324 1.40910533]) to ([0.58564 1.55001586])\nIncrease radii from ([0.58564 1.55001586]) to ([0.644204 1.70501745])\nIncrease radii from ([0.35 0.92634648]) to ([0.385 1.01898112])\nIncrease radii from ([0.385 1.01898112]) to ([0.4235 1.12087924])\nIncrease radii from ([0.4235 1.12087924]) to ([0.46585 1.23296716])\nIncrease radii from ([0.46585 1.23296716]) to ([0.512435 1.35626388])\nIncrease radii from ([0.512435 1.35626388]) to ([0.5636785 1.49189027])\nIncrease radii from ([0.5636785 1.49189027]) to ([0.62004635 1.64107929])\nIncrease radii from ([0.3 0.79401127]) to ([0.33 0.87341239])\nIncrease radii from ([0.33 0.87341239]) to ([0.363 0.96075363])\nIncrease radii from ([0.363 0.96075363]) to ([0.3993 1.056829])\nIncrease radii from ([0.3993 1.056829]) to ([0.43923 1.16251189])\nIncrease radii from ([0.43923 1.16251189]) to ([0.483153 1.27876308])\nIncrease radii from ([0.483153 1.27876308]) to ([0.5314683 1.40663939])\nIncrease radii from ([0.5314683 1.40663939]) to ([0.58461513 1.54730333])\nIncrease radii from ([0.58461513 1.54730333]) to ([0.64307664 1.70203367])\nIncrease radii from ([0.25 0.66167606]) to ([0.275 0.72784366])\nIncrease radii from ([0.275 0.72784366]) to ([0.3025 0.80062803])\nIncrease radii from ([0.3025 0.80062803]) to ([0.33275 0.88069083])\nIncrease radii from ([0.33275 0.88069083]) to ([0.366025 0.96875991])\nIncrease radii from ([0.366025 0.96875991]) to ([0.4026275 1.0656359])\nIncrease radii from ([0.4026275 1.0656359]) to ([0.44289025 1.17219949])\nIncrease radii from ([0.44289025 1.17219949]) to ([0.48717928 1.28941944])\nIncrease radii from ([0.48717928 1.28941944]) to ([0.5358972 1.41836139])\nIncrease radii from ([0.5358972 1.41836139]) to ([0.58948692 1.56019753])\nIncrease radii from ([0.58948692 1.56019753]) to ([0.64843562 1.71621728])\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1762.tif is a low contrast image\n compress=6,\n/content/CTC-IGFL-FR-main/src/SW/miniconda/lib/python3.7/site-packages/elephant/common.py:693: UserWarning: ../../Data/BF-C2DL-HSC/01_RES-allGT/mask1763.tif is a low contrast image\n compress=6,\nrun_export completed\n" ], [ "", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ] ]

ec7b024999582f3f3a7afc01fe6d51513f8bb292

61,701

ipynb

Jupyter Notebook

youtube_api.ipynb

wo1ever/youtube-api-web

0754aee7dd9ced0d8404aac6a860de3948ce753b

[ "MIT" ]

null

youtube_api.ipynb

wo1ever/youtube-api-web

0754aee7dd9ced0d8404aac6a860de3948ce753b

[ "MIT" ]

null

youtube_api.ipynb

wo1ever/youtube-api-web

0754aee7dd9ced0d8404aac6a860de3948ce753b

[ "MIT" ]

null

86.658708

1,079

0.606149

[ [ [ "import requests", "_____no_output_____" ], [ "url = 'https://www.googleapis.com/youtube/v3/search'\n\nparams = dict(\n part='id,snippet',\n q='휘인',\n order='viewCount', #relevance\n type=\"video\",\n maxResults=50,\n key='AIzaSyBK7aaQgSA9mYbdQiJ5Eg6uu4oC_VdmD_s'\n)\n\nresponse = requests.get(url=url, params=params)\ndata = response.json()", "_____no_output_____" ], [ "data\n# items에 영상이 들어있음을 확인.\n# 찍어본 후 items만 가져오도록 할 것\n\nitems = data['items']\nfor item in items:\n print(item)", "{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/P1EH66-5qN8MRx9fzc0vfQO51uo\"', 'id': {'kind': 'youtube#video', 'videoId': 'wxQZ48TgqZk'}, 'snippet': {'publishedAt': '2017-03-08T03:00:02.000Z', 'channelId': 'UCweOkPb1wVVH0Q0Tlj4a5Pw', 'title': '[MV] Jung Key(정키) _ Anymore(부담이 돼) (feat. Whee In of MAMAMOO(휘인 of 마마무))', 'description': \"[MV] Jung Key(정키) _ Anymore(부담이 돼) (feat. Whee In of MAMAMOO(휘인 of 마마무)) *English subtitles are now available. :D (Please click on 'CC' button or ...\", 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/wxQZ48TgqZk/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/wxQZ48TgqZk/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/wxQZ48TgqZk/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '1theK (원더케이)', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/YPrhlY14sf0yU5QcQhUAmvI4GRI\"', 'id': {'kind': 'youtube#video', 'videoId': 'KNWLQI5_JLc'}, 'snippet': {'publishedAt': '2018-04-17T09:00:01.000Z', 'channelId': 'UCweOkPb1wVVH0Q0Tlj4a5Pw', 'title': '[MV] Whee In(휘인) _ EASY (Feat. Sik-K)', 'description': \"[MV] Whee In(휘인) _ EASY (Feat. Sik-K) *English subtitles are now available. (Please click on 'CC' button or activate 'Interactive Transcript' function) [Notice] ...\", 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/KNWLQI5_JLc/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/KNWLQI5_JLc/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/KNWLQI5_JLc/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '1theK (원더케이)', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/am9Ry7LuPFCgXMf_nnOcHXKD5-Y\"', 'id': {'kind': 'youtube#video', 'videoId': 'ZpK9KkOk8fU'}, 'snippet': {'publishedAt': '2016-04-14T05:20:37.000Z', 'channelId': 'UCMgBzk25SliK5UpVvjOmnIQ', 'title': '마마무 휘인 ((센스쩔었던)) 정형돈 삼행시 인정인정!!!ㅋㅋ', 'description': '마마무 휘인 ((센스쩔었던)) 정형돈 삼행시 인정인정!!!ㅋㅋ.', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/ZpK9KkOk8fU/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/ZpK9KkOk8fU/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/ZpK9KkOk8fU/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'atom드라곤', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/3_uQ6jFwSlQsR0WnU5-R0s070TI\"', 'id': {'kind': 'youtube#video', 'videoId': 'W16RU4e_Kf0'}, 'snippet': {'publishedAt': '2016-04-14T15:00:03.000Z', 'channelId': 'UCEf_Bc-KVd7onSeifS3py9g', 'title': \"[STATION] 김희철&김정모 X 휘인 (of 마마무) '나르시스 (Narcissus)' MV\", 'description': \"SM STATION's new track “나르시스 (Narcissus)”, sung by Kim Hee Chul & and WheeIn (of MAMAMOO),and produced by Kim Jung Mo, has been released.\", 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/W16RU4e_Kf0/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/W16RU4e_Kf0/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/W16RU4e_Kf0/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'SMTOWN', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/mMDEoDSbjMquoci8lr5P4qDxD7M\"', 'id': {'kind': 'youtube#video', 'videoId': 'TZlewwpQ0Uw'}, 'snippet': {'publishedAt': '2017-02-18T01:09:34.000Z', 'channelId': 'UCSebqQZeRFFnqfi12pP_4GA', 'title': '마마무 나쁜말 모음 (휘인 응! 시파로마 추가)', 'description': '설명 마마무 나쁜말 모음 (휘인 응! 시파로마 추가)', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/TZlewwpQ0Uw/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/TZlewwpQ0Uw/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/TZlewwpQ0Uw/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '영상편집자', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/xURE8lwjvQuEV8AE6Y0zhDjGLpc\"', 'id': {'kind': 'youtube#video', 'videoId': 'KYrj7cs8oyE'}, 'snippet': {'publishedAt': '2015-07-28T04:27:50.000Z', 'channelId': 'UCgGl-tG-O6wD2Feob8Qkn1g', 'title': '[마마무] 중학생때부터친구인 휘인 화사 (수정)', 'description': '행쇼.....', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/KYrj7cs8oyE/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/KYrj7cs8oyE/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/KYrj7cs8oyE/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '쿠쿠하세요', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/YgvGqQt86iNc61sCaL04luj4J1E\"', 'id': {'kind': 'youtube#video', 'videoId': '9kL8yGPCQKY'}, 'snippet': {'publishedAt': '2018-04-08T02:25:18.000Z', 'channelId': 'UCrHEYpoaF1xKaPfThZVDQUA', 'title': '마마무 휘인이 부르는 멜로망스의 선물', 'description': '옐로우 다이어리에서...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/9kL8yGPCQKY/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/9kL8yGPCQKY/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/9kL8yGPCQKY/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '마로말', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/ekrKG2RmHmoOCARQgor7uVKCPUE\"', 'id': {'kind': 'youtube#video', 'videoId': 'NNNckYgmcA4'}, 'snippet': {'publishedAt': '2017-03-09T08:00:02.000Z', 'channelId': 'UCweOkPb1wVVH0Q0Tlj4a5Pw', 'title': 'Color LIVE(컬러라이브): Jung Key(정키) _ Anymore(부담이 돼)(feat. Whee In of MAMAMOO(휘인 of 마마무))', 'description': 'Color LIVE(컬러라이브): Jung Key(정키) _ Anymore(부담이 돼)(feat. Whee In of MAMAMOO(휘인 of 마마무)) ***** Hello, this is 1theK. We are working on subtitles ...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/NNNckYgmcA4/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/NNNckYgmcA4/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/NNNckYgmcA4/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '1theK (원더케이)', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/dAG2fioIaixPXjzpZUCZ-vVwZnw\"', 'id': {'kind': 'youtube#video', 'videoId': 'pj5zLSX8OII'}, 'snippet': {'publishedAt': '2017-10-06T09:00:02.000Z', 'channelId': 'UCweOkPb1wVVH0Q0Tlj4a5Pw', 'title': '[MV] Whee In(휘인)(MAMAMOO(마마무)) _ Shadow(그림자) (Yellow OST part.1)', 'description': \"[MV] Whee In(휘인)(MAMAMOO(마마무)) _ Shadow(그림자) (Yellow OST part.1) *English subtitles are now available. (Please click on 'CC' button or activate ...\", 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/pj5zLSX8OII/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/pj5zLSX8OII/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/pj5zLSX8OII/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '1theK (원더케이)', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/TLXounVsE5CFVm7m9VFLb5Lc9VE\"', 'id': {'kind': 'youtube#video', 'videoId': '6pl3BZ-3WBs'}, 'snippet': {'publishedAt': '2016-08-31T03:00:02.000Z', 'channelId': 'UC_pwIXKXNm5KGhdEVzmY60A', 'title': '마마무 (MAMAMOO) - Angel (솔라 Solar & 휘인 Whee In) MV', 'description': \"걸그룹 마마무가 데뷔 후 첫 유닛 앨범을 발매 한다. 이번 유닛 곡은 지난 8월 13일~14일 성공적으로 개최 되었던 마마무 첫 단독 콘서트 'MOOSICAL'에...\", 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/6pl3BZ-3WBs/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/6pl3BZ-3WBs/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/6pl3BZ-3WBs/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'Stone Music Entertainment', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/Fuum81WX1sgPGqZx1BLa8gehPX8\"', 'id': {'kind': 'youtube#video', 'videoId': 'rnxgV6dxbag'}, 'snippet': {'publishedAt': '2016-02-06T11:17:07.000Z', 'channelId': 'UCnpn9pPIxINzxJCVfWS6lhw', 'title': '[마마무] 만화더빙해도 될거같은 휘인목소리 모음영상', 'description': '', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/rnxgV6dxbag/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/rnxgV6dxbag/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/rnxgV6dxbag/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '꾸무무', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/IudWJCfWYiSsVS2tuXHQkdFlAv4\"', 'id': {'kind': 'youtube#video', 'videoId': 'U_tQ-GIC0O8'}, 'snippet': {'publishedAt': '2018-04-17T09:00:01.000Z', 'channelId': 'UCuhAUMLzJxlP1W7mEk0_6lA', 'title': '[MV] 휘인(Whee In) - EASY (Feat. 식 케이)', 'description': '[MV] 휘인(Whee In) - EASY (Feat. 식 케이)', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/U_tQ-GIC0O8/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/U_tQ-GIC0O8/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/U_tQ-GIC0O8/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'MAMAMOO', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/OBWmsLC1W8R6_5kwtsK2Tqk5h4w\"', 'id': {'kind': 'youtube#video', 'videoId': 'JmJ8LUjWjt8'}, 'snippet': {'publishedAt': '2018-05-04T10:30:01.000Z', 'channelId': 'UCq0pZFhnpKAEISB6mH03YeA', 'title': '[마마무 휘인,화사] 투영걸만의 분휘기를 느껴보는 방송', 'description': '늦었지만 다시한번더 생일축하해 휘이나ㅠㅠㅠ DO NOT CROP the LOGO & EDIT & REUPLOAD. 불펌 금지, 2차가공 금지, 재업로드 금지입니다. 휘인 - easy ...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/JmJ8LUjWjt8/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/JmJ8LUjWjt8/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/JmJ8LUjWjt8/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '무무옥탑방', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/DoL0tT_0woqq4EHKRAV5PtW2sps\"', 'id': {'kind': 'youtube#video', 'videoId': '9to5Fts8oKU'}, 'snippet': {'publishedAt': '2017-06-15T15:00:02.000Z', 'channelId': 'UCuhAUMLzJxlP1W7mEk0_6lA', 'title': \"[Teaser] 마마무(MAMAMOO) - '나로 말할 것 같으면' 2차 티저(휘인/화사)\", 'description': \"[Teaser] 마마무(MAMAMOO) - '나로 말할 것 같으면' 2차 티저 (휘인/화사)\", 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/9to5Fts8oKU/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/9to5Fts8oKU/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/9to5Fts8oKU/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'MAMAMOO', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/Uc7rsb5c0D8JcqQBiEaBwM22XFM\"', 'id': {'kind': 'youtube#video', 'videoId': 'W7SBbGafiT4'}, 'snippet': {'publishedAt': '2017-09-22T08:00:00.000Z', 'channelId': 'UCq0pZFhnpKAEISB6mH03YeA', 'title': '[마마무 휘인] 실제 인성 모음 (中字)', 'description': '휘인아 놓지않을게!!!!!!!!!!!!!!!! 정휘인알라뷰!!!!!! 마마무 인성 모음은 천천히 멤버별로 만들 생각입니다 :) 거의 입덕영상 수준이라.......ㅋㅋㅋㅋㅋ...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/W7SBbGafiT4/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/W7SBbGafiT4/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/W7SBbGafiT4/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '무무옥탑방', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/U3bOFVfqRS3Sf4lmNuoC2jHl5Jk\"', 'id': {'kind': 'youtube#video', 'videoId': 'ebDS1ypOVUs'}, 'snippet': {'publishedAt': '2015-07-11T03:46:33.000Z', 'channelId': 'UClPdMyn66Q70keC3Ax_v5ww', 'title': '마마무 휘인 빠져들 수밖에 없는 씹덕생물체', 'description': '님들 정휘인 안하고 뭐하심?', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/ebDS1ypOVUs/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/ebDS1ypOVUs/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/ebDS1ypOVUs/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '정휘인사랑해', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/Gt0CG3WSrtGUnAcEUrA_LLgpsG8\"', 'id': {'kind': 'youtube#video', 'videoId': 'FmZ48RjgoDM'}, 'snippet': {'publishedAt': '2017-12-10T11:02:00.000Z', 'channelId': 'UCq0pZFhnpKAEISB6mH03YeA', 'title': '[마마무 솔라,휘인] 살짝 과장이 많다는 에인절 모음 (현웃주의)', 'description': '빨리 올리고 싶은 영상이 있어서 한 주 미리 올립니다! 불펌금지, 2차가공금지, 재업로드금지입니다. (소중한 영상을 흔쾌히 빌려주신 모든 직캠러분...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/FmZ48RjgoDM/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/FmZ48RjgoDM/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/FmZ48RjgoDM/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '무무옥탑방', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/s_nTnH9pOadk0IqsgwTADf8pMNw\"', 'id': {'kind': 'youtube#video', 'videoId': 'Ig0nxRWJ4Ig'}, 'snippet': {'publishedAt': '2017-10-06T09:10:54.000Z', 'channelId': 'UCrwx6JRz13tkmfNvc9iLhMw', 'title': 'Whee In(휘인)(MAMAMOO(마마무)) - Shadow(그림자) Yellow OST part 1', 'description': 'Title : Shadow(그림자) Artist : Whee In(휘인)(MAMAMOO(마마무)) Release Date : 2017-10-06 Genre : OST Label : 로엔 엔터테인먼트 Language : Korean Links ...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/Ig0nxRWJ4Ig/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/Ig0nxRWJ4Ig/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/Ig0nxRWJ4Ig/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'kDrama Freaks', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/CuEYzQ5FbeSsMpNsFV3MjhDMGO8\"', 'id': {'kind': 'youtube#video', 'videoId': 'T70IMOOhoyc'}, 'snippet': {'publishedAt': '2017-12-09T11:46:27.000Z', 'channelId': 'UCO29k0SA-qS4cacBDC2NToA', 'title': '[휘인,화사] 마마무간 실친의 통화내용', 'description': '', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/T70IMOOhoyc/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/T70IMOOhoyc/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/T70IMOOhoyc/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '맘무익힘책', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/F9Bwf0Bb0K6CYLPwHqmcQ3ZXkXw\"', 'id': {'kind': 'youtube#video', 'videoId': '8gu9assnkt0'}, 'snippet': {'publishedAt': '2015-08-04T11:23:19.000Z', 'channelId': 'UCuhAUMLzJxlP1W7mEk0_6lA', 'title': '[MMMTV] 제10화 - 마마무(MAMAMOO) 납량 특집 몰래카메라(휘인&문별)편', 'description': '[MMMTV] 제10화 - 마마무(MAMAMOO) 납량 특집 몰래카메라(휘인&문별)편 이번 MMMTV 10화는 9화에서 예고되었듯이 지난 몰래카메라 후편이 이어집니다!!...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/8gu9assnkt0/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/8gu9assnkt0/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/8gu9assnkt0/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'MAMAMOO', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/xq0x1km5M7XHEIHkPO6ZMbyrY5E\"', 'id': {'kind': 'youtube#video', 'videoId': 'huDKT309k34'}, 'snippet': {'publishedAt': '2015-10-30T08:16:28.000Z', 'channelId': 'UCHIRvVjNIcw6JCreixbgokQ', 'title': '루이(긱스)- 그냥가요(Feat. 휘인 of 마마무)', 'description': '변함없이 난 잘 지내 네가 없는 날들도 늘 그래 왔었기에 뭐 네게 이젠 아무런 할 말도 없고 감정도 없어 이 노랜 그냥 가요일 뿐이야 이 부분엔...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/huDKT309k34/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/huDKT309k34/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/huDKT309k34/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'eji Ok', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/eBg0Q_uHi8MSYvJV--Uu0Ynz4iI\"', 'id': {'kind': 'youtube#video', 'videoId': '0rmV0KXkpfo'}, 'snippet': {'publishedAt': '2017-01-26T12:39:12.000Z', 'channelId': 'UC6r6C2NTZbVeHvvCahIEVHQ', 'title': '[휘인] 정휘인 도른미 모음 / 입덕영상', 'description': '1. 마마무 티비 차안에서 노는법 2. 마마무 티비 1존심 뮤비촬영 비하인드 3. 마마무 티비 납량특집 4. 마마무 티비 아재크러쉬 5. 마마무 티비 아재크...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/0rmV0KXkpfo/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/0rmV0KXkpfo/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/0rmV0KXkpfo/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '그래안말려', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/U16iMVNdwjh0FnEJETYbCBlx98k\"', 'id': {'kind': 'youtube#video', 'videoId': 'JvYcoafZzjI'}, 'snippet': {'publishedAt': '2016-08-31T03:00:12.000Z', 'channelId': 'UCuhAUMLzJxlP1W7mEk0_6lA', 'title': '[MV] 마마무(MAMAMOO) - ANGEL (솔라&휘인)', 'description': '[MV] 마마무(MAMAMOO) - ANGEL (솔라&휘인)', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/JvYcoafZzjI/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/JvYcoafZzjI/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/JvYcoafZzjI/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'MAMAMOO', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/6cqOCgyn5bVcFSkziqGaLkxT3U4\"', 'id': {'kind': 'youtube#video', 'videoId': '2UEYJyx66rw'}, 'snippet': {'publishedAt': '2018-02-16T10:30:00.000Z', 'channelId': 'UCq0pZFhnpKAEISB6mH03YeA', 'title': '[마마무 솔라,문별] 3일동안 준비한 휘인 몰카', 'description': '불펌 금지, 2차가공 금지, 재업로드 금지입니다. DO NOT CROP the LOGO & EDIT & REUPLOAD. If you send me a subtitle, I will add it to the video.', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/2UEYJyx66rw/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/2UEYJyx66rw/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/2UEYJyx66rw/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '무무옥탑방', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/rbk0G1TiSzAov1Ejmy03BJrCc2c\"', 'id': {'kind': 'youtube#video', 'videoId': 'Zq60d03Vxgs'}, 'snippet': {'publishedAt': '2015-10-02T12:45:56.000Z', 'channelId': 'UCU5Dawq8QwGc7QoX373ynMA', 'title': '휘인! 오빠차 태풍 머리날려 댄스! 풀영상 ㅋ', 'description': '구김스.컴퍼니 의류 화보.촬영장에서 마마무 맴버 휘인이 보여준 태풍 댄스!! 인스타그램에서 15초밖에 못보여드렸는데.풀영상.올려요.', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/Zq60d03Vxgs/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/Zq60d03Vxgs/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/Zq60d03Vxgs/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'felix kim', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/nI-xHIO2qiHXygxojqEaCz2_XUE\"', 'id': {'kind': 'youtube#video', 'videoId': 'sxASSvXly8g'}, 'snippet': {'publishedAt': '2017-01-02T15:30:15.000Z', 'channelId': 'UCuw1hxBo5mDVUhgMzRDk3aw', 'title': '솔라와 휘인, 처음 밝히는 불화설의 진실은? [아이돌잔치] 6회 20170102', 'description': '[아이돌잔치 6회] 마마무에도 어색한 멤버들이 존재한다?! [아이돌잔치 6회] [Ch.19] 세상에 없는 TV 이제 시작합니다. 공식 홈페이지 : http://tvchosun.com/e...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/sxASSvXly8g/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/sxASSvXly8g/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/sxASSvXly8g/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'TVCHOSUN', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/jyf6mZH7hcBTxEEEKnqu1YWX7OE\"', 'id': {'kind': 'youtube#video', 'videoId': 'R8K6Xhir0Bc'}, 'snippet': {'publishedAt': '2017-01-02T15:30:00.000Z', 'channelId': 'UCuw1hxBo5mDVUhgMzRDk3aw', 'title': '휘인의 최초 고백! ‘이특의 아버지와 1촌이다‘ [아이돌잔치] 6회 20170102', 'description': '[아이돌잔치 6회] 슈퍼주니어의 팬이었던 휘인! 이특의 아버님과 1촌? [아이돌잔치 6회] [Ch.19] 세상에 없는 TV 이제 시작합니다. 공식 홈페이지...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/R8K6Xhir0Bc/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/R8K6Xhir0Bc/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/R8K6Xhir0Bc/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'TVCHOSUN', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/dCPJr7BYSusjxvehNPE1q7I2D0E\"', 'id': {'kind': 'youtube#video', 'videoId': 'oZe-QKcVPvE'}, 'snippet': {'publishedAt': '2018-05-23T15:03:24.000Z', 'channelId': 'UCv6colBP34LrC9xxf0lzQag', 'title': '마마무 휘인의 터치에 깜짝 놀란 솔라 : 음오아예 프리스타일 Ver.', 'description': '180523 마마무 MAMAMOO 솔라 - 음오아예 Um Oh Ah Yeh 프리스타일 Ver. (고려대축제) 직캠 fancam by zam.', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/oZe-QKcVPvE/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/oZe-QKcVPvE/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/oZe-QKcVPvE/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'z a m', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/QTzlR9vjXJ-CGRu5NKQPusFZJBA\"', 'id': {'kind': 'youtube#video', 'videoId': 'idP0WABHtaU'}, 'snippet': {'publishedAt': '2015-03-22T09:07:42.000Z', 'channelId': 'UCqmcAFHSP1DOZFb88-CuiAA', 'title': '마마무 휘인이 목소리(부제: 휜륨가스)', 'description': '귀여워...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/idP0WABHtaU/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/idP0WABHtaU/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/idP0WABHtaU/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'NohDaeRi', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/W43jH-CLsZC1saID4UuZ_YlF-ss\"', 'id': {'kind': 'youtube#video', 'videoId': 'NvhlcizIFXA'}, 'snippet': {'publishedAt': '2016-11-07T14:45:41.000Z', 'channelId': 'UCOcsML50NYKDMX_VKXNocjw', 'title': '마마무 실친라인 휘인X화사_ 투영걸 (재업', 'description': '내가 보고싶어서 올리는 재업로드! // 전 계정에 있었던 영상 재업로드 입니다 새롭거나 빠진 컷들은 없습니다!', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/NvhlcizIFXA/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/NvhlcizIFXA/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/NvhlcizIFXA/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '마수필라미', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/pZbadBOnmDPpjLET25Q1zTiBjP8\"', 'id': {'kind': 'youtube#video', 'videoId': 'kzIMLEK0ab0'}, 'snippet': {'publishedAt': '2017-09-08T08:00:04.000Z', 'channelId': 'UCq0pZFhnpKAEISB6mH03YeA', 'title': '[마마무 화사,휘인] 타팬도 현웃터지는 실친라인 모음 2탄', 'description': 'https://twitter.com/rooftop619 제가 다음주는 잠시 쉴 수도 있습니다ㅠㅠ 영상 소식은 트위터로 말씀드릴게요! 영상 불펌 금지, 2차가공 금지, 유튜브&타사...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/kzIMLEK0ab0/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/kzIMLEK0ab0/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/kzIMLEK0ab0/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '무무옥탑방', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/5T3vk_fU6pZdVtWSkNVUf63sJ8A\"', 'id': {'kind': 'youtube#video', 'videoId': '3zjuq3b89Io'}, 'snippet': {'publishedAt': '2017-01-02T15:30:00.000Z', 'channelId': 'UCuw1hxBo5mDVUhgMzRDk3aw', 'title': '중학교 친구 휘인과 화사, 조금 오글거리는 첫 만남은? [아이돌잔치] 6회 20170102', 'description': '[아이돌잔치 6회] 서로의 마음을 사로잡은 운명적인 첫 만남은? [아이돌잔치 6회] [Ch.19] 세상에 없는 TV 이제 시작합니다. 공식 홈페이지 : http://tvchosu...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/3zjuq3b89Io/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/3zjuq3b89Io/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/3zjuq3b89Io/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'TVCHOSUN', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/GiAV3wYa9zRk3Ihs4Q4WfwdDZy4\"', 'id': {'kind': 'youtube#video', 'videoId': '27aACOlSD40'}, 'snippet': {'publishedAt': '2017-06-15T15:00:00.000Z', 'channelId': 'UC_pwIXKXNm5KGhdEVzmY60A', 'title': \"마마무 (MAMAMOO) - '나로 말할 것 같으면' 2차 Teaser (휘인/화사)\", 'description': \"마마무 (MAMAMOO) - '나로 말할 것 같으면' 2차 Teaser (휘인/화사) 아이튠즈, 애플뮤직 : https://itun.es/kr/IxuLkb 스포티파이 : http://spoti.fi/2eSVuEJ CJ E&M Music ...\", 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/27aACOlSD40/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/27aACOlSD40/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/27aACOlSD40/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'Stone Music Entertainment', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/J2pKbyu4qyNTyADNv_YulMBJgM4\"', 'id': {'kind': 'youtube#video', 'videoId': 'lFWCkR6KPrc'}, 'snippet': {'publishedAt': '2017-09-07T09:00:05.000Z', 'channelId': 'UCWF5X9tjebglHCixhNrOLfA', 'title': '바람기억 라이브 비교 (성시경, 이영현, 휘인, 한동근 최효인, 케이윌, 반하나, 은가은, 민혜, 강균성, 양파, 박지민, 이은미, 한영애, 나얼)', 'description': '0:03 성시경 https://www.youtube.com/watch?v=LW1MEGZMb1c 1:10 이영현 https://www.youtube.com/watch?v=Cev-inecJHg 2:11 마마무 휘인 ...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/lFWCkR6KPrc/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/lFWCkR6KPrc/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/lFWCkR6KPrc/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '감나무사운드 Kamnamoo', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/9nFSxMpBI-QvAec12CNuARDpGdw\"', 'id': {'kind': 'youtube#video', 'videoId': 'fa_nBGwKf1M'}, 'snippet': {'publishedAt': '2017-01-02T15:30:00.000Z', 'channelId': 'UCuw1hxBo5mDVUhgMzRDk3aw', 'title': '휘인, 돌아가신 할머니 생각에 울컥... [아이돌잔치] 6회 20170102', 'description': '[아이돌잔치 6회] 휘인의 시그널 토크 “할머니에게 죄송...” [아이돌잔치 6회] [Ch.19] 세상에 없는 TV 이제 시작합니다. 공식 홈페이지 : http://tvchosun....', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/fa_nBGwKf1M/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/fa_nBGwKf1M/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/fa_nBGwKf1M/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'TVCHOSUN', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/qsX01osE5L3im52CVoBj-zw0wP0\"', 'id': {'kind': 'youtube#video', 'videoId': 'wJrOYsJovdo'}, 'snippet': {'publishedAt': '2017-07-15T15:40:07.000Z', 'channelId': 'UCrHEYpoaF1xKaPfThZVDQUA', 'title': '마마무 휘인의 쌈마이웨이 김지원 애교 패러디', 'description': '마마무 퍼플 다이어리에서...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/wJrOYsJovdo/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/wJrOYsJovdo/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/wJrOYsJovdo/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '마로말', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/Fy2AISZUq6AZ6aSNByXMY-mhJcI\"', 'id': {'kind': 'youtube#video', 'videoId': 'kt_6wafC9Ac'}, 'snippet': {'publishedAt': '2017-10-06T09:00:04.000Z', 'channelId': 'UCid83oPnsL-4ZEo8CyQr6Rg', 'title': '[옐로우 시즌1 OST part.1] 휘인 - 그림자', 'description': '옐로우 첫 OST 휘인 \"그림자\" 발매! 휘인-그림자 : goo.gl/TWMcxc 프로 짝사랑러들이라면 격공할 노래 너의 뒤에서 우연을 기다린대ㅠㅠㅠㅠ...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/kt_6wafC9Ac/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/kt_6wafC9Ac/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/kt_6wafC9Ac/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '플레이리스트', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/nNjcXtTwHEQo9wWm91VpzE9SpT4\"', 'id': {'kind': 'youtube#video', 'videoId': 'ydsuc1vOavw'}, 'snippet': {'publishedAt': '2018-05-08T10:00:37.000Z', 'channelId': 'UCuhAUMLzJxlP1W7mEk0_6lA', 'title': '[Special] 휘인(Whee In) - EASY 어쿠스틱ver. 새벽 라이브', 'description': '[Special] 휘인(Whee In) - EASY 어쿠스틱ver. 새벽 라이브 - MAMAMOO https://twitter.com/RBW_MAMAMOO https://www.facebook.com/RBW.MAMAMOO ...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/ydsuc1vOavw/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/ydsuc1vOavw/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/ydsuc1vOavw/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'MAMAMOO', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/n0aW3L5H7kmv5l5bZ-4uNIlXNcQ\"', 'id': {'kind': 'youtube#video', 'videoId': 'qDPWU4scOYg'}, 'snippet': {'publishedAt': '2015-05-05T17:57:55.000Z', 'channelId': 'UCm-QZB2Icsw7rZD5HhAx7Ew', 'title': '마마무 = 휘인 + 휜덕 3명', 'description': '휜덕들 휘니 우쭈쭈 모음.', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/qDPWU4scOYg/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/qDPWU4scOYg/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/qDPWU4scOYg/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '김용선 (26)', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/U0j_tvoqvGYnj_zHlJXQ2sJMa9M\"', 'id': {'kind': 'youtube#video', 'videoId': 's4ZdNXEsqLw'}, 'snippet': {'publishedAt': '2016-04-08T13:00:02.000Z', 'channelId': 'UCWvKpZEp5dFsKZiBDxkXHvw', 'title': \"#마마무 '휘인'편, 봄에 어울리는 상큼한 '속쌍 코랄메이크업'_(바비브라운 신상립스틱/눈매 바꾸기ㅋㅋ)ㅣ#킴닥스커버메이크업\", 'description': \"[KIMDAX X BOBBI BROWN] : 이 영상은 바비브라운과 함께 합니다. 안녕하세요 여러분♡ 불금에도 킴닥스랑 함께 ^3^ 주말에는 '전설의 그 동생(ㅋㅋ)'과...\", 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/s4ZdNXEsqLw/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/s4ZdNXEsqLw/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/s4ZdNXEsqLw/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'KIMDAX킴닥스', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/FojzEz7207TUCvGqG-RN_PWC9TI\"', 'id': {'kind': 'youtube#video', 'videoId': 'WXythMq7dy0'}, 'snippet': {'publishedAt': '2015-04-13T12:13:15.000Z', 'channelId': 'UCgGl-tG-O6wD2Feob8Qkn1g', 'title': '[마마무] 팬이 준 플래카드 챙기는 휘인이 (설렘주의)', 'description': '직캠ㅊㅊ:미키님.', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/WXythMq7dy0/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/WXythMq7dy0/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/WXythMq7dy0/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '쿠쿠하세요', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/aC0gW2FWaEukZwOloLBx3uMgwEg\"', 'id': {'kind': 'youtube#video', 'videoId': 'Cz8cBAE0hh8'}, 'snippet': {'publishedAt': '2018-04-17T10:46:36.000Z', 'channelId': 'UC7BgPzByO7vwVZO84pXxyUw', 'title': 'Whee In(휘인) – Easy (Feat. Sik-K) COLOR CODED HAN/ROM/ENG LYRICS', 'description': 'MAMAMOO (마마무) Whee In(휘인) – Easy (Feat. Sik-K) COLOR CODED HAN/ROM/ENG LYRICS OFFICIAL MV LINK: ...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/Cz8cBAE0hh8/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/Cz8cBAE0hh8/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/Cz8cBAE0hh8/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'setsu29', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/WkPbFbQPbmQ3HOn_mFk_plWUwEM\"', 'id': {'kind': 'youtube#video', 'videoId': 'hVuDl0LfL-s'}, 'snippet': {'publishedAt': '2018-05-11T12:01:17.000Z', 'channelId': 'UCphguQXWsUsUAk0UI19W5gQ', 'title': '마마무 휘인 (Mamamoo Wheein)- Easy 안무영상 (Dance Choreography)', 'description': '마마무 휘인- Easy 안무영상 Mamamoo Wheein- Easy Dance Choreography.', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/hVuDl0LfL-s/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/hVuDl0LfL-s/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/hVuDl0LfL-s/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'ᄋᄇ', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/rpsSnJ4k54Z8gTMVQbfURHWWdt0\"', 'id': {'kind': 'youtube#video', 'videoId': 'RB11bDZign4'}, 'snippet': {'publishedAt': '2018-03-11T09:02:42.000Z', 'channelId': 'UCjdcud5iOVcsiBqD_FGwbTA', 'title': '주간아! 마마무 휘인은 암기력 천재?!! 문별어머니는 정형돈 팬!! ［남녀사이］', 'description': '0. 마마무 휘인의 암기력 천재 입증 0. 문별 어머니는 정형돈과 휘인 팬!! 0. 별이 빛나는 밤 신곡 소개.', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/RB11bDZign4/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/RB11bDZign4/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/RB11bDZign4/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '남녀 사이', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/up5bKB9Dr0PDq_wbmhCQszh2KoA\"', 'id': {'kind': 'youtube#video', 'videoId': 'fbhVpzc4QzY'}, 'snippet': {'publishedAt': '2017-02-22T06:11:24.000Z', 'channelId': 'UCYRYVDkOijFfLyT4L0LmRJw', 'title': '170219 마리텔_휘인 잔망모음', 'description': '170219 마이리틀 텔레비전 (마리텔) MLT-44 마마무 (MAMAMOO) 정휘인 잔망 모음.', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/fbhVpzc4QzY/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/fbhVpzc4QzY/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/fbhVpzc4QzY/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '의무는맘무해', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/cD9HwyS5P-oayQ2CrRWKVJVmSgs\"', 'id': {'kind': 'youtube#video', 'videoId': 'fM5S69Gqf5c'}, 'snippet': {'publishedAt': '2017-05-14T13:04:03.000Z', 'channelId': 'UC6r6C2NTZbVeHvvCahIEVHQ', 'title': '[휘인] 진지하게 보면 잘추는 정휘인', 'description': '', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/fM5S69Gqf5c/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/fM5S69Gqf5c/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/fM5S69Gqf5c/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '그래안말려', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/D4dXoqYsVbEFxCZp526m-5s8mJU\"', 'id': {'kind': 'youtube#video', 'videoId': '84j0Tls8pe0'}, 'snippet': {'publishedAt': '2018-07-11T15:00:00.000Z', 'channelId': 'UCuhAUMLzJxlP1W7mEk0_6lA', 'title': \"[TEASER] 마마무(MAMAMOO) - '너나 해(Egotistic)' 휘인 티저\", 'description': \"[TEASER] 마마무(MAMAMOO) - '너나 해(Egotistic)' 휘인 티저 2018.07.16 6PM 'RED MOON' 발매 - MAMAMOO https://twitter.com/RBW_MAMAMOO ...\", 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/84j0Tls8pe0/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/84j0Tls8pe0/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/84j0Tls8pe0/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': 'MAMAMOO', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/KoXkVWhJYLL5_-AozpAcJ5F8uJQ\"', 'id': {'kind': 'youtube#video', 'videoId': 'JD62IOKzRo8'}, 'snippet': {'publishedAt': '2018-07-06T11:00:04.000Z', 'channelId': 'UCq0pZFhnpKAEISB6mH03YeA', 'title': '[마마무 휘인] 센스 모음', 'description': '행복하자 오래가자 DO NOT CROP the LOGO & EDIT & REUPLOAD. 불펌 금지, 2차가공 금지, 재업로드 금지입니다. 많은 무무분들이 제보해주신 영상으로 ...', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/JD62IOKzRo8/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/JD62IOKzRo8/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/JD62IOKzRo8/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '무무옥탑방', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/cQJyBJF0PRlZRyeB7Y92BE-q-38\"', 'id': {'kind': 'youtube#video', 'videoId': 'V8IosfHR_CA'}, 'snippet': {'publishedAt': '2015-07-21T03:04:04.000Z', 'channelId': 'UCnpn9pPIxINzxJCVfWS6lhw', 'title': '마마무 휘인 솔라 데뷔전 영상', 'description': '출처 http://w3.sbs.co.kr/tvview/tvviewEndPage.do?filename=et0599f0037200&mode=bill.', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/V8IosfHR_CA/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/V8IosfHR_CA/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/V8IosfHR_CA/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '꾸무무', 'liveBroadcastContent': 'none'}}\n{'kind': 'youtube#searchResult', 'etag': '\"XI7nbFXulYBIpL0ayR_gDh3eu1k/EuJDhQI3D_l3U-Av1rYCFIyRb4k\"', 'id': {'kind': 'youtube#video', 'videoId': 'Zg_sVU-_c2s'}, 'snippet': {'publishedAt': '2015-07-28T12:07:56.000Z', 'channelId': 'UCgGl-tG-O6wD2Feob8Qkn1g', 'title': '[마마무] 휘인이의 대처능력 bb', 'description': '휘여워.......', 'thumbnails': {'default': {'url': 'https://i.ytimg.com/vi/Zg_sVU-_c2s/default.jpg', 'width': 120, 'height': 90}, 'medium': {'url': 'https://i.ytimg.com/vi/Zg_sVU-_c2s/mqdefault.jpg', 'width': 320, 'height': 180}, 'high': {'url': 'https://i.ytimg.com/vi/Zg_sVU-_c2s/hqdefault.jpg', 'width': 480, 'height': 360}}, 'channelTitle': '쿠쿠하세요', 'liveBroadcastContent': 'none'}}\n" ], [ "import pandas as pd\ndf = pd.DataFrame(data=items)\ndf.head()\n# id안에 또 들어있음. 다시 전처리를 해 주어야 함.", "_____no_output_____" ], [ "videoId = items[0]['id']['videoId']\nvideoId", "_____no_output_____" ], [ "# item에서 특정 정보들만 가져오기 snippet에 제목, 설명, 썸네일 정보 등이 담겨있음.\n\nfor item in items:\n print(item['snippet']['thumbnails']['default']['url'])\n \n # 이미지 150개를 가져 온 결과물", "https://i.ytimg.com/vi/wxQZ48TgqZk/default.jpg\nhttps://i.ytimg.com/vi/KNWLQI5_JLc/default.jpg\nhttps://i.ytimg.com/vi/ZpK9KkOk8fU/default.jpg\nhttps://i.ytimg.com/vi/W16RU4e_Kf0/default.jpg\nhttps://i.ytimg.com/vi/TZlewwpQ0Uw/default.jpg\nhttps://i.ytimg.com/vi/KYrj7cs8oyE/default.jpg\nhttps://i.ytimg.com/vi/9kL8yGPCQKY/default.jpg\nhttps://i.ytimg.com/vi/NNNckYgmcA4/default.jpg\nhttps://i.ytimg.com/vi/pj5zLSX8OII/default.jpg\nhttps://i.ytimg.com/vi/6pl3BZ-3WBs/default.jpg\nhttps://i.ytimg.com/vi/rnxgV6dxbag/default.jpg\nhttps://i.ytimg.com/vi/U_tQ-GIC0O8/default.jpg\nhttps://i.ytimg.com/vi/JmJ8LUjWjt8/default.jpg\nhttps://i.ytimg.com/vi/9to5Fts8oKU/default.jpg\nhttps://i.ytimg.com/vi/W7SBbGafiT4/default.jpg\nhttps://i.ytimg.com/vi/ebDS1ypOVUs/default.jpg\nhttps://i.ytimg.com/vi/FmZ48RjgoDM/default.jpg\nhttps://i.ytimg.com/vi/Ig0nxRWJ4Ig/default.jpg\nhttps://i.ytimg.com/vi/T70IMOOhoyc/default.jpg\nhttps://i.ytimg.com/vi/8gu9assnkt0/default.jpg\nhttps://i.ytimg.com/vi/huDKT309k34/default.jpg\nhttps://i.ytimg.com/vi/0rmV0KXkpfo/default.jpg\nhttps://i.ytimg.com/vi/JvYcoafZzjI/default.jpg\nhttps://i.ytimg.com/vi/2UEYJyx66rw/default.jpg\nhttps://i.ytimg.com/vi/Zq60d03Vxgs/default.jpg\nhttps://i.ytimg.com/vi/sxASSvXly8g/default.jpg\nhttps://i.ytimg.com/vi/R8K6Xhir0Bc/default.jpg\nhttps://i.ytimg.com/vi/oZe-QKcVPvE/default.jpg\nhttps://i.ytimg.com/vi/idP0WABHtaU/default.jpg\nhttps://i.ytimg.com/vi/NvhlcizIFXA/default.jpg\nhttps://i.ytimg.com/vi/kzIMLEK0ab0/default.jpg\nhttps://i.ytimg.com/vi/3zjuq3b89Io/default.jpg\nhttps://i.ytimg.com/vi/27aACOlSD40/default.jpg\nhttps://i.ytimg.com/vi/lFWCkR6KPrc/default.jpg\nhttps://i.ytimg.com/vi/fa_nBGwKf1M/default.jpg\nhttps://i.ytimg.com/vi/wJrOYsJovdo/default.jpg\nhttps://i.ytimg.com/vi/kt_6wafC9Ac/default.jpg\nhttps://i.ytimg.com/vi/ydsuc1vOavw/default.jpg\nhttps://i.ytimg.com/vi/qDPWU4scOYg/default.jpg\nhttps://i.ytimg.com/vi/s4ZdNXEsqLw/default.jpg\nhttps://i.ytimg.com/vi/WXythMq7dy0/default.jpg\nhttps://i.ytimg.com/vi/Cz8cBAE0hh8/default.jpg\nhttps://i.ytimg.com/vi/hVuDl0LfL-s/default.jpg\nhttps://i.ytimg.com/vi/RB11bDZign4/default.jpg\nhttps://i.ytimg.com/vi/fbhVpzc4QzY/default.jpg\nhttps://i.ytimg.com/vi/fM5S69Gqf5c/default.jpg\nhttps://i.ytimg.com/vi/84j0Tls8pe0/default.jpg\nhttps://i.ytimg.com/vi/JD62IOKzRo8/default.jpg\nhttps://i.ytimg.com/vi/V8IosfHR_CA/default.jpg\nhttps://i.ytimg.com/vi/Zg_sVU-_c2s/default.jpg\n" ], [ "items[0]['snippet']", "_____no_output_____" ] ], [ [ "## Todo\n- ~~제목 만들기~~\n- ~~ul li 태그 만들기~~\n- 조회수\n- 게시날짜\n- 영상시간\n- ~~게시자명~~", "_____no_output_____" ] ], [ [ "items[0]['id']", "_____no_output_____" ], [ "# 조회 수를 가져오기 위해 이번에는 video url을 불러옵니다.\ndef get_video_info(videoId):\n url = 'https://www.googleapis.com/youtube/v3/videos'\n\n params = dict(\n part='snippet,contentDetails,statistics',\n id=videoId,\n key='AIzaSyBK7aaQgSA9mYbdQiJ5Eg6uu4oC_VdmD_s'\n )\n\n response = requests.get(url=url, params=params)\n return response.json()\n\nvideo = get_video_info(videoId)\nvideo", "_____no_output_____" ], [ "# 첫 번째 영상의 조회 수 보기\nvideo['items'][0]['statistics']['viewCount']", "_____no_output_____" ], [ "# 게시 날짜 가져오기\nvideo['items'][0]['snippet']['publishedAt']", "_____no_output_____" ], [ "# 영상 시간\nvideo['items'][0]['contentDetails']['duration']", "_____no_output_____" ], [ "html = []\nfor item in items:\n videoId = item['id']['videoId']\n videoInfo = get_video_info(videoId)\n viewCount = videoInfo['items'][0]['statistics']['viewCount']\n duratiton = videoInfo['items'][0]['contentDetails']['duration']\n \n html.append('<li>')\n# html.append('<a href=\"https://www.youtube.com/watch?v={}\">'.format(item['id']['videoId']))\n html.append('<iframe src=\"https://www.youtube.com/embed/{}\" allowfullscreen/>'.format(item['id']['videoId']))\n html.append('<img src=\"{}\">'.format(item['snippet']['thumbnails']['default']['url']))\n html.append('</iframe>')\n html.append('<h1 style=\"font-size:18px;\">{}</h1>'.format(item['snippet']['title']))\n html.append('<span>날짜: {}</span>'.format(item['snippet']['publishedAt']))\n html.append('<span>조회 수: {}</span>'.format(viewCount))\n html.append('<span>영상길이: {}</span>'.format(duratiton))\n html.append('<span>채널제목: {}</span>'.format(item['snippet']['channelTitle']))\n html.append('</li>')\n\nhtml[:10]", "_____no_output_____" ], [ "# 이미지 태그에 개행문자를 붙여 이어주기\ntemp = \"\\n\".join(html) # 리스트를 \\n을 넣은 문자열로 변경해줌.\n\ncontent = \"<ul>{}</ul>\".format(temp)\ncontent[:1000]", "_____no_output_____" ], [ "import webbrowser\nimport os\n\nfile_path = 'youtube.html'\nf = open(file_path,'w')\n\ntemplate = \"\"\"\n<html>\n<head>\n <meta charset=\"UTF-8\">\n <title>{{ title }}</title>\n <link rel=\"stylesheet\" href=\"style.css\"> \n</head>\n<body>\n <header>\n <h1>{{ title }}</h1>\n </header>\n <section>\n {{ content }}\n </section>\n <footer>ⓒ 장난감프로젝트</footer> \n</body>\n</html>\n\"\"\"\n\ntemplate = template.replace('{{ title }}', \"'휘인' Youtube API 수집 영상\")\ntemplate = template.replace('{{ content }}', content)\n\n\nf.write(template)\nf.close()", "_____no_output_____" ], [ "# 브라우저로 해당 파일 열어보기\nfilename = 'file:///'+os.getcwd()+'/' + file_path\nwebbrowser.open_new_tab(filename)", "_____no_output_____" ] ] ]

[ "code", "markdown", "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7b17c5e9c996242a9be690b1624313601666aa

768,645

ipynb

Jupyter Notebook

time_series/fft/fft.ipynb

AnnamalaiKathirkamanathan/machine-learning

628c92c3354e2ed36789793f6e47582b5df7846d

[ "MIT" ]

2

2019-10-18T14:48:04.000Z

2021-11-21T14:53:17.000Z

time_series/fft/fft.ipynb

AnnamalaiKathirkamanathan/machine-learning

628c92c3354e2ed36789793f6e47582b5df7846d

[ "MIT" ]

null

time_series/fft/fft.ipynb

AnnamalaiKathirkamanathan/machine-learning

628c92c3354e2ed36789793f6e47582b5df7846d

[ "MIT" ]

null

660.347938

261,832

0.942828

[ [ [ "<h1>Table of Contents<span class=\"tocSkip\"></span></h1>\n<div class=\"toc\"><ul class=\"toc-item\"><li><span><a href=\"#First-Foray-Into-Discrete/Fast-Fourier-Transformation\" data-toc-modified-id=\"First-Foray-Into-Discrete/Fast-Fourier-Transformation-1\"><span class=\"toc-item-num\">1  </span>First Foray Into Discrete/Fast Fourier Transformation</a></span><ul class=\"toc-item\"><li><span><a href=\"#Correlation\" data-toc-modified-id=\"Correlation-1.1\"><span class=\"toc-item-num\">1.1  </span>Correlation</a></span></li><li><span><a href=\"#Fourier-Transformation\" data-toc-modified-id=\"Fourier-Transformation-1.2\"><span class=\"toc-item-num\">1.2  </span>Fourier Transformation</a></span></li><li><span><a href=\"#DFT-In-Action\" data-toc-modified-id=\"DFT-In-Action-1.3\"><span class=\"toc-item-num\">1.3  </span>DFT In Action</a></span></li><li><span><a href=\"#Fast-Fourier-Transformation-(FFT)\" data-toc-modified-id=\"Fast-Fourier-Transformation-(FFT)-1.4\"><span class=\"toc-item-num\">1.4  </span>Fast Fourier Transformation (FFT)</a></span></li></ul></li><li><span><a href=\"#Reference\" data-toc-modified-id=\"Reference-2\"><span class=\"toc-item-num\">2  </span>Reference</a></span></li></ul></div>", "_____no_output_____" ] ], [ [ "# code for loading the format for the notebook\nimport os\n\n# path : store the current path to convert back to it later\npath = os.getcwd()\nos.chdir(os.path.join('..', '..', 'notebook_format'))\n\nfrom formats import load_style\nload_style(css_style='custom2.css', plot_style=False)", "_____no_output_____" ], [ "os.chdir(path)\n\n# 1. magic for inline plot\n# 2. magic to print version\n# 3. magic so that the notebook will reload external python modules\n# 4. magic to enable retina (high resolution) plots\n# https://gist.github.com/minrk/3301035\n%matplotlib inline\n%load_ext watermark\n%load_ext autoreload\n%autoreload 2\n%config InlineBackend.figure_format='retina'\n\nimport numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\nfrom sklearn.model_selection import train_test_split\n\n%watermark -a 'Ethen' -d -t -v -p numpy,pandas,sklearn,matplotlib", "Ethen 2018-12-13 16:10:17 \n\nCPython 3.6.4\nIPython 6.4.0\n\nnumpy 1.14.2\npandas 0.23.0\nsklearn 0.19.1\nmatplotlib 2.2.2\n" ] ], [ [ "# First Foray Into Discrete/Fast Fourier Transformation", "_____no_output_____" ], [ "In many real-world applications, signals are typically represented as a sequence of numbers that are time dependent. For example, digital audio signal would be one common example, or the hourly temperature in California would be another one. In order to extract meaningful characteristics from these kind of data, many different transformation techniques have been developed to decompose it into simpler individual pieces that are much easier and compact to reason with.\n\n**Discrete Fourier Transformation (DFT)** is one of these algorithms that takes a signal as an input and breaks it down into many individual frequency components. Giving us, the end-user, easier pieces to work with. For the digital audio signal, applying DFT gives us what tones are represented in the sount and at what energies.\n\nSome basics of digital signal processing is assumed. The following link contains an excellent primer to get people up to speed. I feverishly recommend going through all of it if the reader is not pressed with time. [Blog: Seeing Circles, Sines, And Signals a Compact Primer On Digital Signal Processing](https://jackschaedler.github.io/circles-sines-signals/index.html)", "_____no_output_____" ], [ "## Correlation", "_____no_output_____" ], [ "Correlation is a widely used concept in signal processing. It must be noted that the definition of correlation here is slightly different from the definition we encounter in statistics. In the context of signal processing, correlation measures how similar two signals are by computing the dot product between the two. i.e. given two signals $x$ and $y$, the correlation of the two signal can be computed using:\n\n\\begin{align}\n\\sum_{n=0}^N x_n \\cdot y_n\n\\end{align}\n\nThe intuition behind this is that if the two signals are indeed similar, then whenever $x_n$ is positive/negative then $y_n$ should also be positive/negative. Hence when two signals' sign often matches, the resulting correlation number will also be large, indicating that the two signals are similar to one another. It is worth noting that correlation can also take on negative values, a large negative correlation means that the signal is also similar to each other, but one is inverted with respect to the other.", "_____no_output_____" ] ], [ [ "# create examples of two signals that are dissimilar\n# and two that are similar to illustrate the concept\n\n\ndef create_signal(sample_duration, sample_freq, signal_type, signal_freq):\n \"\"\"\n Create some signals to work with, e.g. if we were to sample at 100 Hz\n (100 times per second) and collect the data for 10 seconds, resulting\n in 1000 samples in total. Then we would specify sample_duration = 10,\n sample_freq = 100.\n \n Apart from that, we will also give the option of generating sine or cosine\n wave and the frequencies of these signals\n \"\"\"\n raw_value = 2 * np.pi * signal_freq * np.arange(0, sample_duration, 1. / sample_freq)\n if signal_type == 'cos':\n return np.cos(raw_value)\n elif signal_type == 'sin':\n return np.sin(raw_value)", "_____no_output_____" ], [ "# change default style figure and font size\nplt.rcParams['figure.figsize'] = 8, 6\nplt.rcParams['font.size'] = 12\nplt.style.use('fivethirtyeight')\n\n# dissimilar signals have low correlation\nsignal1 = create_signal(10, 100, 'sin', 0.1)\nsignal2 = create_signal(10, 100, 'cos', 0.1)\nplt.plot(signal1, label='Sine')\nplt.plot(signal2, label='Cosine')\nplt.title('Correlation={:.1f}'.format(np.dot(signal1, signal2)))\nplt.legend()\nplt.show()", "_____no_output_____" ], [ "# similar signals have high correlation\nsignal1 = create_signal(10, 100, 'sin', 0.1)\nsignal2 = create_signal(10, 100, 'sin', 0.1)\nplt.plot(signal1, label='Sine 1')\nplt.plot(signal2, label='Sine 2', linestyle='--')\nplt.title('Correlation={}'.format(np.dot(signal1, signal2)))\nplt.legend()\nplt.show()", "_____no_output_____" ] ], [ [ "Correlation is one of the key concept behind DFT, because as we'll soon see, in DFT, our goal is to find frequencies that gives a high correlation with the signal at hand and a high amplitude of this correlation indicates the presence of this frequency in our signal.", "_____no_output_____" ], [ "## Fourier Transformation", "_____no_output_____" ], [ "Fourier Transformation takes a time-based signal as an input, measures every possible cycle and returns the overall cycle components (by cycle, we're essentially preferring to circles). Each cycle components stores information such as for each cycle:\n\n- **Amplitude:** how big is the circle?\n- **Frequency:** How fast is it moving? The faster the cycle component is moving, the higher the frequency of the wave.\n- **Phase:** Where does it start, or what angle does it start?\n\nThis cycle component is also referred to as **phasor**. The following gif aims to make this seemingly abstract description into a concrete process that we can visualize.\n\n<img src=\"img/fft_decompose.gif\">\n\nAfter applying DFT to our signal shown on the right, we realized that it can be decomposed into five different phasors. Here, the center of the first phasor/cycle component is placed at the origin, and the center of each subsequent phasor is \"attached\" to the tip of the previous phasor. Once the chain of phasors is built, we begin rotating the phasor. We can then reconstruct the time domain signal by tracing the vertical distance from the origin to the tip of the last phasor.", "_____no_output_____" ], [ "Let's now take a look at DFT's formula:\n\n\\begin{align}\nX_k = \\sum_{n=0}^{N-1} x_n \\cdot e^{ -\\varphi \\mathrm{i} }\n\\end{align}\n\n- $x_n$: The signal's value at time $n$.\n- $e^{-\\varphi\\mathrm{i}}$: Is a compact way of describing a pair of sine and cosine waves.\n- $\\varphi = \\frac{n}{N} 2\\pi k$: Records phase and frequency of our cycle components. Where $N$ is the number of samples we have. $n$ the current sample we're considering. $k$ the currenct frequency we're considering. The $2\\pi k$ part represents the cycle component's speed measured in radians and $n / N$ measures the percentage of time that our cycle component has traveled.\n- $X_k$ Amount of cycle component with frequency $k$.\n\n> Side Note: If the readers are a bit rusty with trigonometry (related to sine and cosine) and complex numbers. e.g. There're already many excellent materials out there that covers these concepts. [Blog: Trigonometry Review](https://jackschaedler.github.io/circles-sines-signals/trig_review.html) and [Blog: Complex Numbers](https://jackschaedler.github.io/circles-sines-signals/complex.html)\n\nFrom the formula, we notice that it's taking the dot product between the original signal $x_n$ and $e^{ -\\varphi \\mathrm{i} }$. If we expand $e^{ -\\varphi \\mathrm{i} }$ using the Euler's formula. $e^{ -\\varphi \\mathrm{i} } = cos(\\varphi) - sin(\\varphi)i$, we end up with the formula:\n\n\\begin{align}\nX_k &= \\sum_{n=0}^{N-1} x_n \\cdot \\big( cos(\\varphi) - sin(\\varphi)i \\big) \\\\\n &= \\sum_{n=0}^{N-1} x_n \\cdot cos(\\varphi) - i \\sum_{n=0}^{N-1} x_n \\cdot sin(\\varphi)\n\\end{align}\n\nBy breaking down the formula a little bit, we can see that underneath the hood, what fourier transformation is doing is taking the input signal and doing 2 correlation calculations, one with the sine wave (it will give us the y coordinates of the circle) and one with the cosine wave (which will give us the x coordinates or the circle). And the following succinct one-sentence colour-coded explanation is also a great reference that we can use for quick reference.\n\n<img src=\"img/fft_one_sentence.png\" width=\"50%\" height=\"50%\">", "_____no_output_____" ], [ "## DFT In Action", "_____no_output_____" ], [ "To see DFT in action, we will create a dummy signal that will be composed of four sinusoidal waves of different frequencies. 0, 10, 2 and 0.5 Hz respectively.", "_____no_output_____" ] ], [ [ "# reminder:\n# sample_duration means we're collecting the data for x seconds\n# sample_freq means we're sampling x times per second\nsample_duration = 10\nsample_freq = 100\nsignal_type = 'sin'\nnum_samples = sample_freq * sample_duration\nnum_components = 4\n\ncomponents = np.zeros((num_components, num_samples))\ncomponents[0] = np.ones(num_samples)\ncomponents[1] = create_signal(sample_duration, sample_freq, signal_type, 10)\ncomponents[2] = create_signal(sample_duration, sample_freq, signal_type, 2)\ncomponents[3] = create_signal(sample_duration, sample_freq, signal_type, 0.5)\n\nfig, ax = plt.subplots(nrows=num_components, sharex=True, figsize=(12,8))\nfor i in range(num_components):\n ax[i].plot(components[i])\n ax[i].set_ylim((-1.1, 1.1))\n ax[i].set_title('Component {}'.format(i))\n ax[i].set_ylabel('Amplitude')\n\nax[num_components - 1].set_xlabel('Samples')\nplt.tight_layout()", "_____no_output_____" ] ], [ [ "Then we will combine these individual signals together with some weights assigned to each signal.", "_____no_output_____" ] ], [ [ "signal = -0.5 * components[0] + 0.1 * components[1] + 0.2 * components[2] - 0.6 * components[3]\n\nplt.plot(signal)\nplt.xlabel('Samples')\nplt.ylabel('Amplitude')\nplt.show()", "_____no_output_____" ] ], [ [ "By looking at the dummy signal we've created visually, we might be able to notice the presence of a signal which shows 5 periods in the sampling duration of 10 seconds. In other words, after applying DFT to our signal, we should expect the presence a signal with the frequency of 0.5 HZ.\n\nHere, we will leverage numpy's implementation to check whether the result makes intuitive sense or not. The implementation is called `fft`, but let's not worry about that for the moment.", "_____no_output_____" ] ], [ [ "fft_result = np.fft.fft(signal)\nprint('length of fft result: ', len(fft_result))\nfft_result[:5]", "length of fft result: 1000\n" ] ], [ [ "The `fft` routine returns an array of length 1000 which is equivalent to the number of samples. If we look at each individual element in the array, we'll notice that these are the DFT coefficients. It has two components, the real number corresponds to the cosine waves and the imaginary number that comes from the sine waves. In general though, we don't really care if there's a cosine or sine wave present, as we are only concerned which frequency pattern has a higher correlation with our original signal. This can be done by considering the absolute value of these coefficients.", "_____no_output_____" ] ], [ [ "plt.plot(np.abs(fft_result))\nplt.xlim((-5, 120)) # notice that we limited the x-axis to 120 to focus on the interesting part\nplt.ylim((-5, 520))\nplt.xlabel('K')\nplt.ylabel('|DFT(K)|')\nplt.show()", "_____no_output_____" ] ], [ [ "If we plot the absolute values of the `fft` result, we can clearly see a spike at K=0, 5, 20, 100 in the graph above. However, we are often times more interested in the energy of of each frequency. **Frequency Resolution** is the distance in Hz between two adjacent data points in DFT, which is defined as:\n\n\\begin{align}\n\\Delta f = \\frac{f_s}{N}\n\\end{align}\n\nWhere $f_s$ is the sampling rate and $N$ is the number of data points. The denominator can be expressed in terms of sampling rate and time, $N = f_s \\cdot t$. Looking closely at the formula, it is telling us the only thing that increases frequency resolution is time.\n\nIn our case, the `sample_duration` we've specified above was 10, thus the frequencies corresponding to these K are: 0 Hz, 0.5 Hz, 2 Hz and 10 Hz respectively (remember that these frequencies were the components that was used in the dummy signal that we've created). And based on the graph depicted below, we can see that by passing our signal to a DFT, we were able to retrieve its underlying frequency information.", "_____no_output_____" ] ], [ [ "t = np.linspace(0, sample_freq, len(fft_result))\nplt.plot(t, np.abs(fft_result))\nplt.xlim((-1, 15))\nplt.ylim((-5, 520))\nplt.xlabel('K')\nplt.ylabel('|DFT(K)|')\nplt.show()", "_____no_output_____" ] ], [ [ "## Fast Fourier Transformation (FFT)", "_____no_output_____" ], [ "Recall that the formula for Discrete Fourier Transformation was:\n\n\\begin{align}\nX_k = \\sum_{n=0}^{N-1} x_n \\cdot e^{ -\\frac{n}{N} 2\\pi k \\mathrm{i} }\n\\end{align}\n\nSince we now know that it's computing the dot product between the original signal and a cycle component at every frequency, we can implement this ourselves.", "_____no_output_____" ] ], [ [ "def dft(x):\n \"\"\"Compute the Discrete Fourier Transform of the 1d ndarray x.\"\"\"\n N = x.size\n n = np.arange(N)\n k = n.reshape((N, 1))\n \n # complex number in python are denoted by the j symbol,\n # instead of i that we're showing in the formula\n e = np.exp(-2j * np.pi * k * n / N)\n return np.dot(e, x)", "_____no_output_____" ], [ "# apply dft to our original signal and confirm\n# the results looks the same\ndft_result = dft(signal)\nprint('result matches:', np.allclose(dft_result, fft_result))\n\nplt.plot(np.abs(dft_result))\nplt.xlim((-5, 120))\nplt.ylim((-5, 520))\nplt.xlabel('K')\nplt.ylabel('|DFT(K)|')\nplt.show()", "result matches: True\n" ] ], [ [ "However, if we compare the timing between our simplistic implementation versus the one from numpy, we can see a dramatic time difference.", "_____no_output_____" ] ], [ [ "%timeit dft(signal)\n%timeit np.fft.fft(signal)", "34.1 ms ± 898 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)\n24.8 µs ± 535 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)\n" ] ], [ [ "If we leave aside the fact that one is implemented using Python's numpy and one is most likely implemented in optimized C++, the time difference actually comes from the fact that in practice, people uses a more optimized version of Fourier Transformation called **Fast Fourier Transformation** (how unexpected ...) to perform the calculation. The algorithm accomplish significant speedup by exploiting symmetry property. i.e. if we devise a hypothetical algorithm which can decompose a 1024-point DFT into two 512-point DFTs, then we are essentially halving our computational cost. Let's take a look at how we can achieve this by looking at an example with 8 data points.\n\n\\begin{align}\nX_k = x_0 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 0 } + x_1 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 1 } + \\dots + x_7 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 7 }\n\\end{align}\n\n\nOur goal is to examine the possibility of rewriting this eight-point DFT in terms of two DFTs of smaller length. Let's first examine choosing all the terms with an even sample index, i.e. $x_0$, $x_2$, $x_4$, and $x_6$. Giving us:\n\n\\begin{align}\nG_k &= x_0 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 0 } + x_2 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 2 } + x_4 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 4 } + x_6 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 6 } \\\\\n &= x_0 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{4} k ~\\times~ 0 } + x_2 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{4} k ~\\times~ 1 } + x_4 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{4} k ~\\times~ 2 } + x_6 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{4} k ~\\times~ 3 }\n\\end{align}\n\nAfter plugging the values for the even sample index and simplifying the fractions in the complex exponentials, we can observe that our $G_k$ is a 4 samples DFT with $x_0$, $x_2$, $x_4$, $x_6$ as our input signal. Now that we've shown that we can decompose the even index samples, let's see if we can simplify the remaining terms, the odd-index samples, are given by:\n\n\\begin{align}\nQ_k &= x_1 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 1 } + x_3 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 3 } + x_5 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 5 } + x_7 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 7 } \\\\\n &= e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 1 } \\cdot \\big( x_1 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 0 } + x_3 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 2 } + x_5 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 4 } + x_7 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 6 } \\big) \\\\\n &= e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 1 } \\cdot \\big( x_1 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{4} k ~\\times~ 0 } + x_3 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{4} k ~\\times~ 1 } + x_5 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{4} k ~\\times~ 2 } + x_7 \\cdot e^{ -\\mathrm{i} \\frac{2\\pi}{4} k ~\\times~ 3 } \\big) \\\\\n &= e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 1 } \\cdot H_k\n\\end{align}\n\nAfter the derivation, we can see our $Q_k$ is obtained by multiplying $e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 1 }$ by the four point DFT with the odd index samples of $x_1$, $x_3$, $x_5$, $x_7$, which we'll denote as $H_k$. Hence, we have achieved the goal of decomposing an eight-point DFT into two four-point ones:\n\n\\begin{align}\nX_k &= G_k + e^{ -\\mathrm{i} \\frac{2\\pi}{8} k ~\\times~ 1 } \\cdot H_k\n\\end{align}\n\nWe have only worked through rearranging the terms a bit, next we'll introduce a symmetric trick that allows us to compute the sub-result only once and save computational cost.\n\nThe question that we'll be asking ourselves is what is the value of $X_{N+k}$ is. From our above expression:\n\n\n\\begin{align}\nX_{N + k} &= \\sum_{n=0}^{N-1} x_n \\cdot e^{-i~2\\pi~(N + k)~n~/~N}\\\\\n &= \\sum_{n=0}^{N-1} x_n \\cdot e^{- i~2\\pi~n} \\cdot e^{-i~2\\pi~k~n~/~N}\\\\\n &= \\sum_{n=0}^{N-1} x_n \\cdot e^{-i~2\\pi~k~n~/~N}\n\\end{align}\n\nHere we've used the property that $exp[2\\pi~i~n] = 1$ for any integer $n$, since $exp[2\\pi~i]$ means that we're going 1 full circle, and multiplying that number by any integer $n$ means we're spinning for $n$ circles. The last line shows a nice symmetry property of the DFT: $X_{N+k}=X_k$. This means when we break our eight-point DFT into two four-point DFTs, it allows us to re-use a lot of the results for both $X_k$ and $X_{k + 4}$ and significantly reduce the number of calculations through the symmetric property:\n\n\\begin{align}\nX_{k + 4} &= G_{k + 4} + e^{ -\\mathrm{i} \\frac{2\\pi}{8} (k + 4) ~\\times~ 1 } \\cdot H_{k + 4} \\\\\n &= G_k + e^{ -\\mathrm{i} \\frac{2\\pi}{8} (k + 4) ~\\times~ 1 } \\cdot H_k\n\\end{align}\n\nWe saw that the starting point of the algorithm was that the DFT length $N$ was even and we were able to decrease the computation by splitting it into two DFTS of length $N/2$, following this procedure we can again decompose each of the $N/2$ DFTs into two $N/4$ DFTs. This property turns the original $\\mathcal{O}[N^2]$ DFT computation into a $\\mathcal{O}[N\\log N]$ algorithm to compute DFT.", "_____no_output_____" ] ], [ [ "def fft(x):\n N = x.shape[0]\n \n if N % 2 > 0:\n raise ValueError('size of x must be a power of 2')\n elif N <= 32: # this cutoff should be enough to start using the non-recursive version\n return dft(x)\n else:\n fft_even = fft(x[0::2])\n fft_odd = fft(x[1::2])\n factor = np.exp(-2j * np.pi * np.arange(N) / N)\n return np.concatenate([fft_even + factor[:N // 2] * fft_odd,\n fft_even + factor[N // 2:] * fft_odd])", "_____no_output_____" ], [ "# here, we assume the input data length is a power of two\n# if it doesn't, we can choose to zero-pad the input signal\nx = np.random.random(1024)\nnp.allclose(fft(x), np.fft.fft(x))", "_____no_output_____" ], [ "%timeit dft(x)\n%timeit fft(x)\n%timeit np.fft.fft(x)", "36.7 ms ± 1.71 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)\n1.62 ms ± 32.9 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n27.6 µs ± 711 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)\n" ] ], [ [ "By re-vising our algorithm that we use to compute DFT, we have improved our implementation by an order of magnitude! The algorithm we've implemented here is also referred to as the [radix-2 Cooley-Tukey FFT](https://en.wikipedia.org/wiki/Cooley%E2%80%93Tukey_FFT_algorithm).\n\nThough the pure-Python functions are probably not useful in practice, as due to the importance of the FFT in so many applications, Both NumPy, `numpy.fft`, and SciPy, `scipy.fftpack` have wrappers of the extremely well-tested FFTPACK library. But I still believe that by spending the time to get an intuition of what's happening underneath the hood of these tools, the better practitioners we'll be.\n\nApplication of DFT in the context of equipment condition monitoring. [Blog: Fast Fourier Transforms in Python](https://ericstrong.org/fast-fourier-transforms-in-python/)", "_____no_output_____" ], [ "# Reference", "_____no_output_____" ], [ "- [Blog: Understanding the FFT Algorithm](https://jakevdp.github.io/blog/2013/08/28/understanding-the-fft/)\n- [Blog: Dummies guide to Fourier Transform](https://nipunbatra.github.io/blog/2016/FT.html)\n- [Blog: An Introduction to the Fast Fourier Transform](https://www.allaboutcircuits.com/technical-articles/an-introduction-to-the-fast-fourier-transform/)\n- [Blog: An Intuitive Discrete Fourier Transform Tutorial](http://practicalcryptography.com/miscellaneous/machine-learning/intuitive-guide-discrete-fourier-transform/)\n- [Blog: Understanding the Fourier Transform by example](https://www.ritchievink.com/blog/2017/04/23/understanding-the-fourier-transform-by-example/)\n- [Blog: The Fourier Transform, explained in one sentence](https://blog.revolutionanalytics.com/2014/01/the-fourier-transform-explained-in-one-sentence.html)\n- [Blog: Seeing Circles, Sines, And Signals a Compact Primer On Digital Signal Processing](https://jackschaedler.github.io/circles-sines-signals/index.html)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown", "markdown", "markdown" ], [ "code", "code", "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown", "markdown", "markdown" ] ]

ec7b1a304cad4b9899406d3413082da4e111932d

24,880

ipynb

Jupyter Notebook

Desafio 2/cloud-park-project-ptbr-2/assets/notebook/notebook_parte-2_XvWXXx6YQ.ipynb

elladarte/Maratona_Behind_the_Code_2020

b92fa06a285263a27bd3d157a466f334b41833be

[ "Apache-2.0" ]

3

2020-12-30T06:40:07.000Z

2021-04-06T02:19:14.000Z

Desafio 2/cloud-park-project-ptbr-2/assets/notebook/notebook_parte-2_XvWXXx6YQ.ipynb

elladarte/Maratona_Behind_the_Code_2020

b92fa06a285263a27bd3d157a466f334b41833be

[ "Apache-2.0" ]

null

Desafio 2/cloud-park-project-ptbr-2/assets/notebook/notebook_parte-2_XvWXXx6YQ.ipynb

elladarte/Maratona_Behind_the_Code_2020

b92fa06a285263a27bd3d157a466f334b41833be

[ "Apache-2.0" ]

null

24,880

0.728497

[ [ [ "# MARATONA BEHIND THE CODE 2020\n\n## DESAFIO 2: PARTE 2", "_____no_output_____" ], [ "### Introdução", "_____no_output_____" ], [ "Na parte 1 deste desafio, você realizou o pré-processamento e o treinamento de um modelo a partir de um conjunto de dados base fornecido. Nesta segunda etapa você irá integrar todas as transformações e eventos de treinamento criados anteriormente em uma Pipeline completa para *deploy* no **Watson Machine Learning**!", "_____no_output_____" ], [ "### Preparação do Notebook", "_____no_output_____" ], [ "Primeiro realizaremos a instalação do scikit-learn e a importação das mesmas bibliotecas utilizadas anteriormente", "_____no_output_____" ] ], [ [ "!pip install scikit-learn==0.20.0 --upgrade", "_____no_output_____" ], [ "import json\nimport requests\nimport pandas as pd\nimport numpy as np\nimport xgboost as xgb\nfrom sklearn.impute import SimpleImputer\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.tree import DecisionTreeClassifier\nfrom sklearn.pipeline import Pipeline\nfrom sklearn.model_selection import KFold, cross_validate", "_____no_output_____" ] ], [ [ "É necessário inserir o conjunto de dados base novamente como um dataframe pandas, seguindo as instruções\n\n![alt text](https://i.imgur.com/K1DwL9I.png \"importing-csv-as-df\")\n\nApós a seleção da opção **\"Insert to code\"**, a célula abaixo será preenchida com o código necessário para importação e leitura dos dados no arquivo .csv como um DataFrame Pandas.", "_____no_output_____" ] ], [ [ "\n<< INSIRA O DATASET COMO UM DATAFRAME PANDAS NESTA CÉLULA >>\n", "_____no_output_____" ] ], [ [ "### Construção da Pipeline completa para encapsulamento no WML", "_____no_output_____" ], [ "#### Preparando transformações personalizadas para carregamento no WML", "_____no_output_____" ], [ "Na etapa anterior, foi mostrado como criar uma transformação personalizada, através da declaração de uma classe Python com os métodos ``fit`` e ``transform``.\n\n - Código da transformação personalizada DropColumns():\n \n from sklearn.base import BaseEstimator, TransformerMixin\n # All sklearn Transforms must have the `transform` and `fit` methods\n class DropColumns(BaseEstimator, TransformerMixin):\n def __init__(self, columns):\n self.columns = columns\n def fit(self, X, y=None):\n return self\n def transform(self, X):\n # Primeiro realizamos a cópia do dataframe 'X' de entrada\n data = X.copy()\n # Retornamos um novo dataframe sem as colunas indesejadas\n return data.drop(labels=self.columns, axis='columns')\n\nPara integrar esses tipos de transformações personalizadas nas Pipelines do Watson Machine Learning, é necessário primeiramente empacotar seu código personalizado como uma biblioteca Python. Isso pode ser feito facilmente com o uso da ferramenta *setuptools*.\n\nNo seguinte repositório git: https://github.com/vnderlev/sklearn_transforms temos todos os arquivos necessários para a criação de um pacote Python, nomeado **my_custom_sklearn_transforms**.\nEsse pacote possui a seguinte estrutura de arquivos:\n\n /my_custom_sklearn_transforms.egg-info\n dependency_links.txt\n not-zip-safe\n PKG-INFO\n SOURCES.txt\n top_level.txt\n /my_custom_sklearn_transforms\n __init__.py\n sklearn_transformers.py\n PKG-INFO\n README.md\n setup.cfg\n setup.py\n \nO arquivo principal, que irá conter o código das nossas transformadas personalizadas, é o arquivo **/my_custom_sklearn_transforms/sklearn_transformers.py**. Se você acessá-lo no repositório, irá notar que ele contém exatamente o mesmo código declarado na primeira etapa (a classe DropColumns).\n\nCaso você tenha declarado transformações próprias (além da DropColumn fornecida), você deverá adicionar todas as classes dessas transformadas criadas por você nesse mesmo arquivo. Para tal, você deve realizar o fork desse repositório (isso pode ser feito na própria interface Web do Github, clicando no botão conforme a imagem abaixo), e adicionar suas classes personalizadas no arquivo **sklearn_transformers.py**.\n\n![alt text](https://i.imgur.com/D81E1uM.png \"forking-a-repo\")\n\nSe você somente fez o uso da transformação fornecida (DropColumns), pode ignorar essa etapa de fork, e seguir utilizando o pacote base fornecido! :)\n\nApós a preparação do seu pacote Python com as suas transformadas personalizadas, substitua o link do repositório git na célula abaixo e execute-a. Caso você não tenha preparado nenhuma nova transformada, execute a célula com o link do repositório já fornecido. \n\n<hr>\n \n**OBSERVAÇÃO**\n\nCaso a execução da célula abaixo retorne um erro de que o repositório já existe, execute:\n\n**!rm -r -f sklearn_transforms**", "_____no_output_____" ] ], [ [ "# substitua o link abaixo pelo link do seu repositório git (se for o caso)\n!git clone https://github.com/vnderlev/sklearn_transforms.git", "_____no_output_____" ], [ "!cd sklearn_transforms\n!ls -ltr", "_____no_output_____" ] ], [ [ "Para subir o código no WML, precisamos enviar um arquivo .zip com todo o código fonte, então iremos zipar o diretório clonado em seguida:", "_____no_output_____" ] ], [ [ "!zip -r sklearn_transforms.zip sklearn_transforms", "_____no_output_____" ] ], [ [ "Com o arquivo zip do nosso pacote carregado no Kernel deste notebook, podemos utilizar a ferramenta pip para instalá-lo, conforme a célula abaixo:", "_____no_output_____" ] ], [ [ "!pip install sklearn_transforms.zip", "_____no_output_____" ] ], [ [ "Podemos agora realizar a importação do nosso pacote personalizado em nosso notabook!\n\nIremos importar a transformação DropColumns. Se você possui outras transformações personalizadas, não se esqueça de importá-las!", "_____no_output_____" ] ], [ [ "from my_custom_sklearn_transforms.sklearn_transformers import DropColumns", "_____no_output_____" ] ], [ [ "#### Declarando a Pipeline\n\nApós a importação das transformações personalizadas como um pacote Python, podemos partir para a declaração da nossa Pipeline.\n\nO processo é bem semelhante ao realizado na primeira etapa, porém com algumas diferenças importantes, então preste bem atenção!\n\nA Pipeline exemplo possui três estágios: \n\n - remover a coluna \"NOME\"\n - imputar \"zeros\" em todos os valores faltantes\n - inserir os dados pré-processados como entrada em um modelo treinado\n \nRelembrando, a entrada desta Pipeline será o conjunto cru de dados fornecido exceto a coluna \"LABELS\" (variável-alvo a ser determinada pelo modelo).\n\nTeremos então 17 valores de entrada **na PIPELINE** (no modelo serão 16 entradas, pois a coluna NAME será removida no primeiro estágio após a transformação DropColumn).\n\n MATRICULA - número de quatro algarismos único para cada estudante\n NOME - nome completo do estudante\n FALTAS_DE - número de faltas na disciplina de ``Direito Empresarial``\n FALTAS_EM - número de faltas na disciplina de ``Empreendedorismo``\n FALTAS_MF - número de faltas na disciplina de ``Matemática Financeira``\n MEDIA_DE - média simples das notas do aluno na disciplina de ``Direito Empresarial`` (0-10)\n MEDIA_EM - média simples das notas do aluno na disciplina de ``Empreendedorismo`` (0-10)\n MEDIA_MF - média simples das notas do aluno na disciplina de ``Matemática Financeira`` (0-10)\n HRS_ESTUDO_DE - horas de estudo particular na disciplina de ``Direito Empresarial``\n HRS_ESTUDO_EM - horas de estudo particular na disciplina de ``Empreendedorismo``\n HRS_ESTUDO_MF - horas de estudo particular na disciplina de ``Matemática Financeira``\n REPROVACOES_DE - número de reprovações na disciplina de ``Direito Empresarial``\n REPROVACOES_EM - número de reprovações na disciplina de ``Empreendedorismo``\n REPROVACOES_MF - número de reprovações na disciplina de ``Matemática Financeira``\n LIVROS_TEXTO - quantidade de livros e textos acessados pelo aluno no sistema da universidade\n AULAS_AO_VIVO - horas de aulas ao vivo presenciadas pelo aluno (total em todas as disciplinas)\n EXERCICIOS - número de exercícios realizados pelo estudante (total em todas as disciplinas) no sistema da universidade\n\nA saída da Pipeline será um valor estimado para a coluna \"LABELS\".", "_____no_output_____" ] ], [ [ "# Criação de uma Transform personalizada ``DropColumns``\n\nrm_columns = DropColumns(\n columns=[\"NOME\"]\n)", "_____no_output_____" ], [ "# Criação de um objeto ``SimpleImputer``\n\nsi = SimpleImputer(\n missing_values=np.nan, # os valores faltantes são do tipo ``np.nan`` (padrão Pandas)\n strategy='constant', # a estratégia escolhida é a alteração do valor faltante por uma constante\n fill_value=0, # a constante que será usada para preenchimento dos valores faltantes é um int64=0.\n verbose=0,\n copy=True\n)", "_____no_output_____" ], [ "# Definição das colunas que serão features (nota-se que a coluna NOME não está presente)\nfeatures = [\n \"MATRICULA\", \"NOME\", 'REPROVACOES_DE', 'REPROVACOES_EM', \"REPROVACOES_MF\", \"REPROVACOES_GO\",\n \"NOTA_DE\", \"NOTA_EM\", \"NOTA_MF\", \"NOTA_GO\",\n \"INGLES\", \"H_AULA_PRES\", \"TAREFAS_ONLINE\", \"FALTAS\", \n]\n\n# Definição da variável-alvo\ntarget = [\"PERFIL\"]\n\n# Preparação dos argumentos para os métodos da biblioteca ``scikit-learn``\nX = df_data_1[features]\ny = df_data_1[target]", "_____no_output_____" ] ], [ [ "**ATENÇÃO!!**\n\nA célula acima, embora muito parecida com a definição de features na primeira etapa deste desafio, possui uma grande diferença!\n\nNela está presente a coluna \"NOME\" como uma feature! Isso ocorre pois neste caso essas são as entradas da *PIPELINE*, e não do modelo.", "_____no_output_____" ] ], [ [ "# Separação dos dados em um conjunto de treino e um conjunto de teste\nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=337)", "_____no_output_____" ] ], [ [ "Na célula abaixo é realizada a declaração de um objeto **Pipeline** do scikit-learn, onde é declarado o parâmetro *steps*, que nada mais é do que uma lista com as etapas da nossa pipeline:\n\n 'remove_cols' - transformação personalizada DropColumns\n 'imputer' - transformação embutida do scikit-learn para imputação de valores faltantes\n 'dtc' - um classificador via árvore de decisão\n \nNote que passamos como passos as transformadas instanciadas anteriormente, sob nome `rm_columns` e `si`.", "_____no_output_____" ] ], [ [ "# Criação da nossa pipeline para armazenamento no Watson Machine Learning:\nmy_pipeline = Pipeline(\n steps=[\n ('remove_cols', rm_columns),\n ('imputer', si),\n ('dtc', DecisionTreeClassifier()),\n ]\n)", "_____no_output_____" ] ], [ [ "Em seguida iremos executar o método `fit()` da Pipeline, realizando o pré-processamento e o treinamento do modelo de uma só vez.", "_____no_output_____" ] ], [ [ "# Inicialização da Pipeline (pré-processamento e realização do treinamento do modelo)\nmy_pipeline.fit(X_train, y_train)", "_____no_output_____" ] ], [ [ "Agora que temos uma pipeline completa, com etapas de pré-processamento configuradas e também um modelo por árvore de decisão já treinado, podemos realizar a integração com o Watson Machine Learning!", "_____no_output_____" ], [ "### Encapsulando uma Pipeline personalizada no Watson Machine Learning", "_____no_output_____" ], [ "#### Estabelecendo conexão entre o cliente Python do WML e a sua instância do serviço na nuvem", "_____no_output_____" ] ], [ [ "# Biblioteca Python com implementação de um cliente HTTP para a API do WML\nfrom watson_machine_learning_client import WatsonMachineLearningAPIClient", "_____no_output_____" ] ], [ [ "As próximas células irão realizar o deploy da pipeline declarada neste notebook no WML. Só prossiga se você já está satisfeito com seu modelo e acha que já é a hora de fazer o deploy da sua solução.\n\nCole as credenciais de sua instância do Watson Machine Learning na variável na célula abaixo.\n\nÉ importante que a variável que contém os valores tenha o nome de ``wml_credentials`` para que as próximas células deste notebook executem corretamente.", "_____no_output_____" ] ], [ [ "wml_credentials = {\n \"apikey\": \"\",\n \"iam_apikey_description\": \"\",\n \"iam_apikey_name\": \"\",\n \"iam_role_crn\": \"\",\n \"iam_serviceid_crn\": \"\",\n \"instance_id\": \"\",\n \"url\": \"\"\n}", "_____no_output_____" ], [ "# Instanciando um objeto cliente do Watson Machine Learning a partir das credenciais fornecidas\n\nclientWML = WatsonMachineLearningAPIClient(wml_credentials)", "_____no_output_____" ], [ "# Extraindo detalhes da sua instância do Watson Machine Learning\n\ninstance_details = clientWML.service_instance.get_details()\nprint(json.dumps(instance_details, indent=4))", "_____no_output_____" ] ], [ [ "**ATENÇÃO!!**\n\nFique atento para os limites de consumo de sua instância do Watson Machine Learning!\n\nCaso você expire a camada grátis, não será possível avaliar seu modelo (pois é necessária a realização de algumas chamadas de API que consomem predições!)", "_____no_output_____" ], [ "#### Listando todos os artefatos armazenados no seu WML", "_____no_output_____" ], [ "Para listar todos os artefatos armazenados em seu Watson Machine Learning, você pode usar a seguinte função:\n\n clientWML.repository.list()", "_____no_output_____" ] ], [ [ "# Listando todos os artefatos atualmente armazenados na sua instância do WML\n\nclientWML.repository.list()", "_____no_output_____" ] ], [ [ "No plano LITE do Watson Machine Learning só é permitido o deploy de um único modelo por vez. Se for o caso de você já possuir um modelo online na sua instância, você pode apagá-lo utilizando o método clientWML.repository.delete():\n\n artifact_guid = \"359c8951-d2fe-4063-8706-cc06b32d5e0d\"\n clientWML.repository.delete(artifact_guid)", "_____no_output_____" ], [ "#### Criando uma nova definição de pacote Python personalizado no WML", "_____no_output_____" ], [ "O primeiro passo para realizar seu deploy é armazenar o código das transformações personalizadas criadas por você.\n\nPara essa etapa precisamos apenas do arquivo .zip do pacote criado (que já possuimos carregado no Kernel!)", "_____no_output_____" ] ], [ [ "# Definição de metadados do nosso pacote com as Transforms personalizadas\npkg_meta = {\n clientWML.runtimes.LibraryMetaNames.NAME: \"my_custom_sklearn_transform_1\",\n clientWML.runtimes.LibraryMetaNames.DESCRIPTION: \"A custom sklearn transform\",\n clientWML.runtimes.LibraryMetaNames.FILEPATH: \"sklearn_transforms.zip\", # Note que estamos utilizando o .zip criado anteriormente!\n clientWML.runtimes.LibraryMetaNames.VERSION: \"1.0\",\n clientWML.runtimes.LibraryMetaNames.PLATFORM: { \"name\": \"python\", \"versions\": [\"3.6\"] }\n}\ncustom_package_details = clientWML.runtimes.store_library( pkg_meta )\ncustom_package_uid = clientWML.runtimes.get_library_uid( custom_package_details )\n\nprint(\"\\n Lista de artefatos de runtime armazenados no WML:\")\nclientWML.repository.list()", "_____no_output_____" ] ], [ [ "#### Criando uma nova definição de runtime Python personalizado no WML\n\nO segundo passo é armazenar uma definição de runtime Python para utilizar a nossa biblioteca personalizada.\n\nIsso pode ser feito da seguinte forma:", "_____no_output_____" ] ], [ [ "runtime_meta = {\n clientWML.runtimes.ConfigurationMetaNames.NAME: \"my_custom_wml_runtime_1\",\n clientWML.runtimes.ConfigurationMetaNames.DESCRIPTION: \"A Python runtime with custom sklearn Transforms\",\n clientWML.runtimes.ConfigurationMetaNames.PLATFORM: {\n \"name\": \"python\",\n \"version\": \"3.6\"\n },\n clientWML.runtimes.ConfigurationMetaNames.LIBRARIES_UIDS: [ custom_package_uid ]\n}\nruntime_details = clientWML.runtimes.store( runtime_meta )\ncustom_runtime_uid = clientWML.runtimes.get_uid( runtime_details )\n\nprint(\"\\n Detalhes do runtime armazenado:\")\nprint(json.dumps(runtime_details, indent=4))", "_____no_output_____" ], [ "# Listando todos runtimes armazenados no seu WML:\nclientWML.runtimes.list()", "_____no_output_____" ] ], [ [ "#### Criando uma nova definição de Pipeline personalizada no WML\n\nFinalmente iremos criar uma definição (metadados) para a nossa Pipeline ser hospedada no WML.\n\nDefinimos como parâmetros um nome para o artefato e o ID do runtime criado anteriormente.", "_____no_output_____" ] ], [ [ "model_meta = {\n clientWML.repository.ModelMetaNames.NAME: 'desafio-2-mbtc2020-pipeline-1',\n clientWML.repository.ModelMetaNames.DESCRIPTION: \"my pipeline for submission\",\n clientWML.repository.ModelMetaNames.RUNTIME_UID: custom_runtime_uid\n}", "_____no_output_____" ] ], [ [ "Em seguida chamamos o método para armazenar a nova definição:", "_____no_output_____" ] ], [ [ "# Função para armazenar uma definição de Pipeline no WML\nstored_model_details = clientWML.repository.store_model(\n model=my_pipeline, # `my_pipeline` é a variável criada anteriormente e contém nossa Pipeline já treinada :)\n meta_props=model_meta, # Metadados definidos na célula anterior\n training_data=None # Não altere esse parâmetro\n)\n\nprint(\"\\n Lista de artefatos armazenados no WML:\")\nclientWML.repository.list()\n\n# Detalhes do modelo hospedado no Watson Machine Learning\nprint(\"\\n Metadados do modelo armazenado:\")\nprint(json.dumps(stored_model_details, indent=4))", "_____no_output_____" ] ], [ [ "#### Realizando o deployment do seu modelo para consumo imediato por outras aplicações", "_____no_output_____" ] ], [ [ "# O deployment do modelo é finalmente realizado por meio do método ``deployments.create()``\n\nmodel_deployment_details = clientWML.deployments.create(\n artifact_uid=stored_model_details[\"metadata\"][\"guid\"], # Não altere esse parâmetro\n name=\"desafio-2-mbtc2020-deployment-1\",\n description=\"Solução do desafio 2 - MBTC\",\n asynchronous=False, # Não altere esse parâmetro\n deployment_type='online', # Não altere esse parâmetro\n deployment_format='Core ML', # Não altere esse parâmetro\n meta_props=model_meta # Não altere esse parâmetro\n)", "_____no_output_____" ] ], [ [ "#### Testando um modelo hospedado no Watson Machine Learning", "_____no_output_____" ] ], [ [ "# Recuperando a URL endpoint do modelo hospedado na célula anterior\n\nmodel_endpoint_url = clientWML.deployments.get_scoring_url(model_deployment_details)\nprint(\"A URL de chamada da sua API é: {}\".format(model_endpoint_url))", "_____no_output_____" ], [ "# Detalhes do deployment realizado\n\ndeployment_details = clientWML.deployments.get_details(\n deployment_uid=model_deployment_details[\"metadata\"][\"guid\"] # esse é o ID do seu deployment!\n)\n\nprint(\"Metadados do deployment realizado: \\n\")\nprint(json.dumps(deployment_details, indent=4))", "_____no_output_____" ], [ "scoring_payload = {\n 'fields': [\n \"MATRICULA\", \"NOME\", 'REPROVACOES_DE', 'REPROVACOES_EM', \"REPROVACOES_MF\", \"REPROVACOES_GO\",\n \"NOTA_DE\", \"NOTA_EM\", \"NOTA_MF\", \"NOTA_GO\",\n \"INGLES\", \"H_AULA_PRES\", \"TAREFAS_ONLINE\", \"FALTAS\", \n ],\n 'values': [\n [\n 513949,\"Marli Quésia de Oliveira\",1,1,1,1,4.3,4.0,3.1,4.9,0,3,4,3,\n ]\n ]\n}\n\nprint(\"\\n Payload de dados a ser classificada:\")\nprint(json.dumps(scoring_payload, indent=4))", "_____no_output_____" ], [ "result = clientWML.deployments.score(\n model_endpoint_url,\n scoring_payload\n)\n\nprint(\"\\n Resultados:\")\nprint(json.dumps(result, indent=4))", "_____no_output_____" ] ], [ [ "<hr>\n\n## Parabéns! \n\nSe tudo foi executado sem erros, você já tem um classificador baseado em machine learning encapsulado como uma API REST!\n\nPara testar a sua solução integrada com um assistente virtual e realizar a submissão, acesse a página:\n\nhttps://uninassau.maratona.dev\n\nVocê irá precisar da endpoint url do seu modelo e das credenciais do WML :)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ] ]

ec7b22ff0028e6d129250f0a19d3b8be5ea6ad43

46,856

ipynb

Jupyter Notebook

docs/machine_learning/preprocessing_images/remove_backgrounds.ipynb

revgizmo-forks/ds_notes

ffc73d06b07fb2b137e7e679d3c99dab53580afa

[ "CC0-1.0" ]

1

2020-03-18T21:13:25.000Z

docs/machine_learning/preprocessing_images/remove_backgrounds.ipynb

revgizmo-forks/ds_notes

ffc73d06b07fb2b137e7e679d3c99dab53580afa

[ "CC0-1.0" ]

null

docs/machine_learning/preprocessing_images/remove_backgrounds.ipynb

revgizmo-forks/ds_notes

ffc73d06b07fb2b137e7e679d3c99dab53580afa

[ "CC0-1.0" ]

1

2020-10-17T22:00:42.000Z

241.525773

42,758

0.917919

[ [ [ "---\ntitle: \"Remove Backgrounds\"\nauthor: \"Chris Albon\"\ndate: 2017-12-20T11:53:49-07:00\ndescription: \"How to remove the backgrounds in images using OpenCV in Python.\"\ntype: technical_note\ndraft: false\n---", "_____no_output_____" ] ], [ [ "<a alt=\"grabcut\" href=\"https://machinelearningflashcards.com\">\n <img src=\"/images/machine_learning_flashcards/Grabcut_print.png\" class=\"flashcard center-block\">\n</a>", "_____no_output_____" ], [ "## Preliminaries", "_____no_output_____" ] ], [ [ "# Load image\nimport cv2\nimport numpy as np\nfrom matplotlib import pyplot as plt", "_____no_output_____" ] ], [ [ "## Load Image", "_____no_output_____" ] ], [ [ "# Load image\nimage_bgr = cv2.imread('images/plane_256x256.jpg')", "_____no_output_____" ] ], [ [ "## Convert To RGB", "_____no_output_____" ] ], [ [ "# Convert to RGB\nimage_rgb = cv2.cvtColor(image_bgr, cv2.COLOR_BGR2RGB)", "_____no_output_____" ] ], [ [ "## Draw Rectangle Around Foreground", "_____no_output_____" ] ], [ [ "# Rectange values: start x, start y, width, height\nrectangle = (0, 56, 256, 150)", "_____no_output_____" ] ], [ [ "## Apply GrabCut", "_____no_output_____" ] ], [ [ "# Create initial mask\nmask = np.zeros(image_rgb.shape[:2], np.uint8)\n\n# Create temporary arrays used by grabCut\nbgdModel = np.zeros((1, 65), np.float64)\nfgdModel = np.zeros((1, 65), np.float64)\n\n# Run grabCut\ncv2.grabCut(image_rgb, # Our image\n mask, # The Mask\n rectangle, # Our rectangle\n bgdModel, # Temporary array for background\n fgdModel, # Temporary array for background\n 5, # Number of iterations\n cv2.GC_INIT_WITH_RECT) # Initiative using our rectangle\n\n# Create mask where sure and likely backgrounds set to 0, otherwise 1\nmask_2 = np.where((mask==2) | (mask==0), 0, 1).astype('uint8')\n\n# Multiply image with new mask to subtract background\nimage_rgb_nobg = image_rgb * mask_2[:, :, np.newaxis]", "_____no_output_____" ] ], [ [ "## Show image", "_____no_output_____" ] ], [ [ "# Show image\nplt.imshow(image_rgb_nobg), plt.axis(\"off\")\nplt.show()", "_____no_output_____" ] ] ]

[ "raw", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "raw" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

ec7b255ce13d60930f85728cd6b83bab24e51ea6

48,732

ipynb

Jupyter Notebook

notebooks/1_model_based.ipynb

bachsh/interpretability-implementations-demos

8c03c535d19445d27073702080072f8c28852a36

[ "MIT" ]

null

notebooks/1_model_based.ipynb

bachsh/interpretability-implementations-demos

8c03c535d19445d27073702080072f8c28852a36

[ "MIT" ]

null

notebooks/1_model_based.ipynb

bachsh/interpretability-implementations-demos

8c03c535d19445d27073702080072f8c28852a36

[ "MIT" ]

null

96.690476

21,340

0.813244

[ [ [ "%load_ext autoreload\n%autoreload 2\nimport matplotlib.pyplot as plt\nimport numpy as np\nfrom copy import deepcopy\nfrom sklearn.datasets import fetch_openml\nimport pandas as pd\nimport time, sys, os\n\n# some helpful funcs for getting data\ndef get_reg_boston_data():\n boston_data = pd.read_csv(\"../data/boston.csv\", index_col=0)\n y = boston_data.medv.values\n X = boston_data.drop(\"medv\", axis=1)\n features = X.columns\n X = X.values\n return X, y, features", "_____no_output_____" ] ], [ [ "### short decision tree", "_____no_output_____" ] ], [ [ "from sklearn.tree import DecisionTreeRegressor, plot_tree\n\n# load in some data\nX, y, features = get_reg_boston_data()\n\n# specify a decision tree with a maximum depth\ndt = DecisionTreeRegressor(max_depth=3)\ndt.fit(X, y)\n\n# calculate mse on the training data\npreds = dt.predict(X)\nprint(f'train mse: {np.mean(np.square(preds-y)):0.2f}')\n\nplot_tree(dt)\nplt.savefig('tree.pdf')\nplt.show()", "train mse: 15.38\n" ] ], [ [ "### integer linear models", "_____no_output_____" ] ], [ [ "from imodels import SLIM", "_____no_output_____" ], [ "np.random.seed(123)\n\n# generate X and y\nn, p = 1000, 10\nX = np.random.randn(n, p)\ny = X[:, 0] + 2 * X[:, 1] - 1 * X[:, 2] + np.random.randn(n)\n\n# fit linear models with different regularization parameters\nmodel = SLIM()\nfor lambda_reg in [0, 1e-2, 5e-2, 1e-1, 1, 2]:\n model.fit(X, y, lambda_reg)\n mse = np.mean(np.square(y - model.predict(X)))\n print(f'lambda: {lambda_reg}\\tmse: {mse: 0.2f}\\tweights: {model.model.coef_}')", "lambda: 0\tmse: 2.87\tweights: [0 2 0 0 0 0 0 0 0 0]\nlambda: 0.01\tmse: 2.87\tweights: [0 2 0 0 0 0 0 0 0 0]\nlambda: 0.05\tmse: 2.87\tweights: [0 2 0 0 0 0 0 0 0 0]\nlambda: 0.1\tmse: 2.09\tweights: [ 1 1 -1 0 0 0 0 0 0 0]\nlambda: 1\tmse: 2.87\tweights: [0 2 0 0 0 0 0 0 0 0]\nlambda: 2\tmse: 2.87\tweights: [0 2 0 0 0 0 0 0 0 0]\n" ] ], [ [ "### rulefit\n- installed with: `pip install pip install git+https://github.com/csinva/interpretability-implementations-demos`\n- [documentation](https://github.com/christophM/rulefit) and the [original paper](http://statweb.stanford.edu/~jhf/ftp/RuleFit.pdf)", "_____no_output_____" ] ], [ [ "from imodels import RuleFit\n\n# load some data\nX, y, features = get_reg_boston_data()\n\n# fit a rulefit model\nrf = RuleFit()\nrf.fit(X, y, feature_names=features)\n\n# calculate mse on the training data\npreds = rf.predict(X)\nprint(f'train mse: {np.mean(np.square(preds-y)):0.2f}')", "train mse: 1.86\n" ] ], [ [ "now, let's inspect the rules", "_____no_output_____" ] ], [ [ "rules = rf.get_rules()\n\nrules = rules[rules.coef != 0].sort_values(\"support\", ascending=False)\n\n# 'rule' is how the feature is constructed\n# 'coef' is its weight in the final linear model\n# 'support' is how many points it applies to\nrules[['rule', 'coef', 'support']].head().style.background_gradient(cmap='viridis')", "_____no_output_____" ] ], [ [ "### greedy rule lists\n**like a decision tree that only ever splits going left**", "_____no_output_____" ] ], [ [ "from imodels import GreedyRuleList\n\n# load some data\ndata = fetch_openml(\"diabetes\") # get dataset\ny = (data.target == 'tested_positive').astype(np.int) # labels 0-1\nX = data.data\nprint('shapes', X.shape, y.shape)\n\n# fit a rulefit model\nm = GreedyRuleList()\nm.fit(X, y=y, feature_names=data.feature_names) # stores into m.rules_\nprobs = m.predict_proba(X)\nprint(m)\n\n# look at prediction breakdown\nplt.hist(probs[:, 1][y==0], label='class 0')\nplt.hist(probs[:, 1][y==1], label='class 1', alpha=0.8)\nplt.ylabel('Count')\nplt.xlabel('Predicted probability of class 1')\nplt.legend()\nplt.show()", "shapes (768, 8) (768,)\nmean 0.3489583333333333\nmean 0.19381443298969073\nmean 0.08487084870848709\nmean 0.013245033112582781\nmean 0.006666666666666667\n" ] ], [ [ "### scalable bayesian rule lists\n- docs are [here](https://github.com/csinva/interpretability-workshop/tree/master/models/bayesian_rule_lists)", "_____no_output_____" ] ], [ [ "from imodels import RuleListClassifier\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.ensemble import RandomForestClassifier\n\n\nfeature_labels = [\"#Pregnant\",\"Glucose concentration test\",\"Blood pressure(mmHg)\",\"Triceps skin fold thickness(mm)\",\n \"2-Hour serum insulin (mu U/ml)\",\"Body mass index\",\"Diabetes pedigree function\",\"Age (years)\"]", "_____no_output_____" ], [ "np.random.seed(13)\ndata = fetch_openml(\"diabetes\") # get dataset\ny = (data.target == 'tested_positive').astype(np.int) # labels 0-1\n\nXtrain, Xtest, ytrain, ytest = train_test_split(data.data, y, test_size=0.75) # split\n\n# train classifier (allow more iterations for better accuracy; use BigDataRuleListClassifier for large datasets)\nprint('training...')\nmodel = RuleListClassifier(max_iter=10000, class1label=\"diabetes\", verbose=False)\nmodel.fit(Xtrain, ytrain, feature_labels=feature_labels)\n\nprint(\"RuleListClassifier Accuracy:\", model.score(Xtest, ytest), \"Learned interpretable model:\\n\", model)\n# print(\"RandomForestClassifier Accuracy:\", RandomForestClassifier(n_estimators=10).fit(Xtrain, ytrain).score(Xtest, ytest))", "training...\nRuleListClassifier Accuracy: 0.7065972222222222 Learned interpretable model:\n Trained RuleListClassifier for detecting diabetes\n==================================================\nIF #Pregnant : 6.5_to_inf THEN probability of diabetes: 65.7% (49.5%-80.3%)\nELSE IF Glucose concentration test : -inf_to_122.5 THEN probability of diabetes: 9.9% (4.9%-16.4%)\nELSE IF Body mass index : 30.9_to_inf THEN probability of diabetes: 69.2% (54.1%-82.5%)\nELSE IF Age (years) : -inf_to_26.5 THEN probability of diabetes: 22.2% (3.2%-52.7%)\nELSE probability of diabetes: 44.4% (23.0%-67.1%)\n=================================================\n\n" ] ], [ [ "### optimal classification tree\n- docs [here](https://github.com/csinva/interpretability-workshop/tree/master/models/optimal_classification_tree)\n- note: this implementation is still somewhat unstable, and can be made faster by installing either `cplex` or `gurobi`", "_____no_output_____" ] ], [ [ "sys.path.append('../imodels/optimal_classification_tree/pyoptree')\nsys.path.append('../imodels/optimal_classification_tree/')", "_____no_output_____" ], [ "from optree import OptimalTreeModel\nfeature_names = np.array([\"x1\", \"x2\"])\n\nX = np.array([[1, 2, 2, 2, 3], [1, 2, 1, 0, 1]]).T\ny = np.array([1, 1, 0, 0, 0]).reshape(-1, 1)\nX_test = np.array([[1, 1, 2, 2, 2, 3, 3], [1, 2, 2, 1, 0, 1, 0]]).T\ny_test = np.array([1, 1, 1, 0, 0, 0, 0])\n\nnp.random.seed(13)\nmodel = OptimalTreeModel(tree_depth=3, N_min=1, alpha=0.1) #, solver_name='baron'\nmodel.fit(X_test, y_test) # this method is currently using the fast, but not optimal solver\npreds = model.predict(X_test)\n\n# fit on the bigger diabetes dset from above\n# model.fit(Xtrain, ytrain) # this method is currently using the fast, but not optimal solver\n# preds = model.predict(Xtest)\n\nprint('acc', np.mean(preds == y_test))", "_____no_output_____" ], [ "model.print_tree(feature_names)", "depth 0:\n\t x2 > 0.8114524236945737\n\ndepth 1:\n\tnode 2 undefined\n\tnode 3 undefined\n\ndepth 2:\n\tnode 4 undefined\n\t x1 > 0.01086684288089712\n\t x2 > 0.9159532769401844\n\tnode 7 undefined\n\ndepth 3:\n\tnode 8 undefined\n\tnode 9 undefined\n\tnode 10 undefined\n\tnode 11 undefined\n\tnode 12 undefined\n\tnode 13 undefined\n\tnode 14 undefined\n\tnode 15 undefined\n\n" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code" ] ]

ec7b2f362eb5aad9a3c57ca492f42799c5f8f956

320,377

ipynb

Jupyter Notebook

FINAL-TF2-FILES/TF_2_Notebooks_and_Data/05-RNNs/01-RNN-Time-Series-Example.ipynb

tanuja333/Tensorflow_Keras

e29464da56666c675667b491b12d625ffaefddd9

[ "Apache-2.0" ]

2

2020-08-14T13:42:03.000Z

2020-08-19T20:32:29.000Z

FINAL-TF2-FILES/TF_2_Notebooks_and_Data/05-RNNs/01-RNN-Time-Series-Example.ipynb

tanuja333/Tensorflow_Keras

e29464da56666c675667b491b12d625ffaefddd9

[ "Apache-2.0" ]

9

2020-09-25T21:54:00.000Z

2022-02-10T01:39:05.000Z

FINAL-TF2-FILES/TF_2_Notebooks_and_Data/05-RNNs/01-RNN-Time-Series-Example.ipynb

tanuja333/Tensorflow_Keras

e29464da56666c675667b491b12d625ffaefddd9

[ "Apache-2.0" ]

null

189.124557

85,508

0.895289

[ [ [ "___\n\n<a href='http://www.pieriandata.com'><img src='../Pierian_Data_Logo.png'/></a>\n___\n<center><em>Copyright Pierian Data</em></center>\n<center><em>For more information, visit us at <a href='http://www.pieriandata.com'>www.pieriandata.com</a></em></center>\n\n# RNN Example for Time Series", "_____no_output_____" ] ], [ [ "import pandas as pd\nimport numpy as np\n\n%matplotlib inline\nimport matplotlib.pyplot as plt", "_____no_output_____" ] ], [ [ "## Data\n\nRelease: Advance Monthly Sales for Retail and Food Services \nUnits: Millions of Dollars, Not Seasonally Adjusted\n\nFrequency: Monthly\n\nThe value for the most recent month is an advance estimate that is based on data from a subsample of firms from the larger Monthly Retail Trade Survey. The advance estimate will be superseded in following months by revised estimates derived from the larger Monthly Retail Trade Survey. The associated series from the Monthly Retail Trade Survey is available at https://fred.stlouisfed.org/series/MRTSSM448USN\n\nInformation about the Advance Monthly Retail Sales Survey can be found on the Census website at https://www.census.gov/retail/marts/about_the_surveys.html\n\nSuggested Citation:\nU.S. Census Bureau, Advance Retail Sales: Clothing and Clothing Accessory Stores [RSCCASN], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/RSCCASN, November 16, 2019.\n\nhttps://fred.stlouisfed.org/series/RSCCASN", "_____no_output_____" ] ], [ [ "df = pd.read_csv('../Data/RSCCASN.csv',index_col='DATE',parse_dates=True)", "_____no_output_____" ], [ "df.head()", "_____no_output_____" ], [ "df.columns = ['Sales']", "_____no_output_____" ], [ "df.plot(figsize=(12,8))", "_____no_output_____" ] ], [ [ "## Train Test Split", "_____no_output_____" ] ], [ [ "len(df)", "_____no_output_____" ] ], [ [ "Data is monthly, let's forecast 1.5 years into the future.", "_____no_output_____" ] ], [ [ "len(df)- 18", "_____no_output_____" ], [ "test_size = 18", "_____no_output_____" ], [ "test_ind = len(df)- test_size", "_____no_output_____" ], [ "train = df.iloc[:test_ind]\ntest = df.iloc[test_ind:]", "_____no_output_____" ], [ "train", "_____no_output_____" ], [ "test", "_____no_output_____" ] ], [ [ "## Scale Data", "_____no_output_____" ] ], [ [ "from sklearn.preprocessing import MinMaxScaler", "_____no_output_____" ], [ "scaler = MinMaxScaler()", "_____no_output_____" ], [ "# IGNORE WARNING ITS JUST CONVERTING TO FLOATS\n# WE ONLY FIT TO TRAININ DATA, OTHERWISE WE ARE CHEATING ASSUMING INFO ABOUT TEST SET\nscaler.fit(train)", "_____no_output_____" ], [ "scaled_train = scaler.transform(train)\nscaled_test = scaler.transform(test)", "_____no_output_____" ] ], [ [ "# Time Series Generator\n\nThis class takes in a sequence of data-points gathered at\nequal intervals, along with time series parameters such as\nstride, length of history, etc., to produce batches for\ntraining/validation.\n\n#### Arguments\n data: Indexable generator (such as list or Numpy array)\n containing consecutive data points (timesteps).\n The data should be at 2D, and axis 0 is expected\n to be the time dimension.\n targets: Targets corresponding to timesteps in `data`.\n It should have same length as `data`.\n length: Length of the output sequences (in number of timesteps).\n sampling_rate: Period between successive individual timesteps\n within sequences. For rate `r`, timesteps\n `data[i]`, `data[i-r]`, ... `data[i - length]`\n are used for create a sample sequence.\n stride: Period between successive output sequences.\n For stride `s`, consecutive output samples would\n be centered around `data[i]`, `data[i+s]`, `data[i+2*s]`, etc.\n start_index: Data points earlier than `start_index` will not be used\n in the output sequences. This is useful to reserve part of the\n data for test or validation.\n end_index: Data points later than `end_index` will not be used\n in the output sequences. This is useful to reserve part of the\n data for test or validation.\n shuffle: Whether to shuffle output samples,\n or instead draw them in chronological order.\n reverse: Boolean: if `true`, timesteps in each output sample will be\n in reverse chronological order.\n batch_size: Number of timeseries samples in each batch\n (except maybe the last one).", "_____no_output_____" ] ], [ [ "from tensorflow.keras.preprocessing.sequence import TimeseriesGenerator", "_____no_output_____" ], [ "# Let's redefine to get 12 months back and then predict the next month out\nlength = 12\ngenerator = TimeseriesGenerator(scaled_train, scaled_train, length=length, batch_size=1)", "_____no_output_____" ], [ "# What does the first batch look like?\nX,y = generator[0]", "_____no_output_____" ], [ "print(f'Given the Array: \\n{X.flatten()}')\nprint(f'Predict this y: \\n {y}')", "Given the Array: \n[0. 0.02127505 0.05580163 0.08942056 0.09512053 0.08146965\n 0.07860151 0.12979233 0.09566512 0.1203892 0.15426227 0.41595266]\nPredict this y: \n [[0.02047633]]\n" ] ], [ [ "### Create the Model", "_____no_output_____" ] ], [ [ "from tensorflow.keras.models import Sequential\nfrom tensorflow.keras.layers import Dense\nfrom tensorflow.keras.layers import LSTM", "_____no_output_____" ], [ "# We're only using one feature in our time series\nn_features = 1", "_____no_output_____" ], [ "# define model\nmodel = Sequential()\nmodel.add(LSTM(100, activation='relu', input_shape=(length, n_features)))\nmodel.add(Dense(1))\nmodel.compile(optimizer='adam', loss='mse')", "_____no_output_____" ], [ "model.summary()", "Model: \"sequential_4\"\n_________________________________________________________________\nLayer (type) Output Shape Param # \n=================================================================\nlstm_3 (LSTM) (None, 100) 40800 \n_________________________________________________________________\ndense_3 (Dense) (None, 1) 101 \n=================================================================\nTotal params: 40,901\nTrainable params: 40,901\nNon-trainable params: 0\n_________________________________________________________________\n" ] ], [ [ "### EarlyStopping and creating a Validation Generator\n\nNOTE: The scaled_test dataset size MUST be greater than your length chosen for your batches. Review video for more info on this.", "_____no_output_____" ] ], [ [ "from tensorflow.keras.callbacks import EarlyStopping", "_____no_output_____" ], [ "early_stop = EarlyStopping(monitor='val_loss',patience=2)", "_____no_output_____" ], [ "validation_generator = TimeseriesGenerator(scaled_test,scaled_test, length=length, batch_size=1)", "_____no_output_____" ], [ "# fit model\nmodel.fit_generator(generator,epochs=20,\n validation_data=validation_generator,\n callbacks=[early_stop])", "Epoch 1/20\n304/304 [==============================] - 10s 34ms/step - loss: 0.0273 - val_loss: 0.0351\nEpoch 2/20\n304/304 [==============================] - 10s 34ms/step - loss: 0.0153 - val_loss: 0.0360\nEpoch 3/20\n304/304 [==============================] - 10s 34ms/step - loss: 0.0130 - val_loss: 0.0049\nEpoch 4/20\n304/304 [==============================] - 10s 34ms/step - loss: 0.0055 - val_loss: 0.0170\nEpoch 5/20\n304/304 [==============================] - 10s 33ms/step - loss: 0.0052 - val_loss: 0.0010\nEpoch 6/20\n304/304 [==============================] - 10s 34ms/step - loss: 0.0021 - val_loss: 4.6310e-04\nEpoch 7/20\n304/304 [==============================] - 10s 34ms/step - loss: 0.0012 - val_loss: 0.0029\nEpoch 8/20\n304/304 [==============================] - 10s 34ms/step - loss: 0.0021 - val_loss: 0.0011\n" ], [ "losses = pd.DataFrame(model.history.history)", "_____no_output_____" ], [ "losses.plot()", "_____no_output_____" ] ], [ [ "## Evaluate on Test Data", "_____no_output_____" ] ], [ [ "first_eval_batch = scaled_train[-length:]", "_____no_output_____" ], [ "first_eval_batch = first_eval_batch.reshape((1, n_input, n_features))", "_____no_output_____" ], [ "model.predict(first_eval_batch)", "_____no_output_____" ], [ "scaled_test[0]", "_____no_output_____" ] ], [ [ "Now let's put this logic in a for loop to predict into the future for the entire test range.\n\n----", "_____no_output_____" ], [ "**NOTE: PAY CLOSE ATTENTION HERE TO WHAT IS BEING OUTPUTED AND IN WHAT DIMENSIONS. ADD YOUR OWN PRINT() STATEMENTS TO SEE WHAT IS TRULY GOING ON!!**", "_____no_output_____" ] ], [ [ "test_predictions = []\n\nfirst_eval_batch = scaled_train[-length:]\ncurrent_batch = first_eval_batch.reshape((1, length, n_features))\n\nfor i in range(len(test)):\n \n # get prediction 1 time stamp ahead ([0] is for grabbing just the number instead of [array])\n current_pred = model.predict(current_batch)[0]\n \n # store prediction\n test_predictions.append(current_pred) \n \n # update batch to now include prediction and drop first value\n current_batch = np.append(current_batch[:,1:,:],[[current_pred]],axis=1)", "_____no_output_____" ] ], [ [ "## Inverse Transformations and Compare", "_____no_output_____" ] ], [ [ "true_predictions = scaler.inverse_transform(test_predictions)", "_____no_output_____" ], [ "# IGNORE WARNINGS\ntest['Predictions'] = true_predictions", "C:\\Users\\Marcial\\Anaconda3\\envs\\tf2gpu\\lib\\site-packages\\ipykernel_launcher.py:2: SettingWithCopyWarning: \nA value is trying to be set on a copy of a slice from a DataFrame.\nTry using .loc[row_indexer,col_indexer] = value instead\n\nSee the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n \n" ], [ "test", "_____no_output_____" ], [ "test.plot(figsize=(12,8))", "_____no_output_____" ] ], [ [ "# Retrain and Forecasting", "_____no_output_____" ] ], [ [ "full_scaler = MinMaxScaler()\nscaled_full_data = full_scaler.fit_transform(df)", "_____no_output_____" ], [ "length = 12 # Length of the output sequences (in number of timesteps)\ngenerator = TimeseriesGenerator(scaled_full_data, scaled_full_data, length=length, batch_size=1)", "_____no_output_____" ], [ "model = Sequential()\nmodel.add(LSTM(100, activation='relu', input_shape=(length, n_features)))\nmodel.add(Dense(1))\nmodel.compile(optimizer='adam', loss='mse')\n\n\n# fit model\nmodel.fit_generator(generator,epochs=8)", "Epoch 1/8\n322/322 [==============================] - 11s 33ms/step - loss: 0.0242\nEpoch 2/8\n322/322 [==============================] - 11s 33ms/step - loss: 0.0151\nEpoch 3/8\n322/322 [==============================] - 11s 35ms/step - loss: 0.0087\nEpoch 4/8\n322/322 [==============================] - 11s 33ms/step - loss: 0.0047\nEpoch 5/8\n322/322 [==============================] - 11s 33ms/step - loss: 0.0036\nEpoch 6/8\n322/322 [==============================] - 11s 35ms/step - loss: 0.0020\nEpoch 7/8\n322/322 [==============================] - 11s 34ms/step - loss: 0.0019\nEpoch 8/8\n322/322 [==============================] - 11s 34ms/step - loss: 0.0013\n" ], [ "forecast = []\n# Replace periods with whatever forecast length you want\nperiods = 12\n\nfirst_eval_batch = scaled_full_data[-length:]\ncurrent_batch = first_eval_batch.reshape((1, length, n_features))\n\nfor i in range(periods):\n \n # get prediction 1 time stamp ahead ([0] is for grabbing just the number instead of [array])\n current_pred = model.predict(current_batch)[0]\n \n # store prediction\n forecast.append(current_pred) \n \n # update batch to now include prediction and drop first value\n current_batch = np.append(current_batch[:,1:,:],[[current_pred]],axis=1)", "_____no_output_____" ], [ "forecast = scaler.inverse_transform(forecast)", "_____no_output_____" ] ], [ [ "### Creating new timestamp index with pandas.", "_____no_output_____" ] ], [ [ "df", "_____no_output_____" ], [ "forecast_index = pd.date_range(start='2019-11-01',periods=periods,freq='MS')", "_____no_output_____" ], [ "forecast_df = pd.DataFrame(data=forecast,index=forecast_index,\n columns=['Forecast'])", "_____no_output_____" ], [ "forecast_df", "_____no_output_____" ], [ "df.plot()\nforecast_df.plot()", "_____no_output_____" ] ], [ [ "### Joining pandas plots\n\nhttps://stackoverflow.com/questions/13872533/plot-different-dataframes-in-the-same-figure", "_____no_output_____" ] ], [ [ "ax = df.plot()\nforecast_df.plot(ax=ax)", "_____no_output_____" ], [ "ax = df.plot()\nforecast_df.plot(ax=ax)\nplt.xlim('2018-01-01','2020-12-01')", "_____no_output_____" ] ], [ [ "# Great Job!", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ] ]

ec7b2fb1a0d21a7072244c37367c3946cb2b304c

26,521

ipynb

Jupyter Notebook

hw4/cifar10_Target_Propagation.ipynb

haohaoqian/NMDA

6b7993fbe3a1bb28bd15cec8d60a7c13076705b4

[ "MIT" ]

null

hw4/cifar10_Target_Propagation.ipynb

haohaoqian/NMDA

6b7993fbe3a1bb28bd15cec8d60a7c13076705b4

[ "MIT" ]

null

hw4/cifar10_Target_Propagation.ipynb

haohaoqian/NMDA

6b7993fbe3a1bb28bd15cec8d60a7c13076705b4

[ "MIT" ]

null

47.190391

523

0.557822

[ [ [ "# Difference Target Propagation\n---\nIn this notebook, we will use an interesting neural network that learns via **target propagation** to classify images from the CIFAR-10 database. Note that we provide a number of instructions (including code and descriptions) for the ease of your learning. Be free to change the provided code if you want but in this case you should explain your motivation in the submitted report.\n\n\nOur aim is to reproduce the results of the paper entitled \"Difference Target Propagation\" （ https://arxiv.org/abs/1412.7525 ） on the CIFAR-10 database. Before completing this work, you are required to carefully read the paper and understand its basic idea about the proposed learning strategy. \n\n\n![](model.png)\n\nFigure 1. The Schematic overview of the target propagation\n", "_____no_output_____" ], [ "### Test for CUDA\n\nSince these are larger (32x32x3) images, it may prove useful to speed up your training time by using a GPU. CUDA is a parallel computing platform and CUDA Tensors are the same as typical Tensors, only they utilize GPU's for computation.", "_____no_output_____" ] ], [ [ "import torch\nimport numpy as np\n\n# check if CUDA is available\ntrain_on_gpu = torch.cuda.is_available()\ndevice=torch.device('cuda' if torch.cuda.is_available() else 'cpu')\nif not train_on_gpu:\n print('CUDA is not available. Training on CPU ...')\nelse:\n print('CUDA is available! Training on GPU ...')", "CUDA is available! Training on GPU ...\n" ] ], [ [ "---\n## Load the Data\n\nIf you're not familiar with the Cifar-10, you may find it useful to look at: http://www.cs.toronto.edu/~kriz/cifar.html . \n\nA copy of the data is also placed on the class website. \n\n#### TODO: Load the data", "_____no_output_____" ] ], [ [ "#############################################################################\n# TODO: load the data #\n#############################################################################\nimport torchvision\nimport torchvision.transforms as transforms\n\nbatch_size=1024\n\ntransform=transforms.Compose(\n [transforms.ToTensor(),\n transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])])\n\nfull_dataset = torchvision.datasets.CIFAR10(\n root='./data',\n train=True,\n download=True,\n transform=transform)\ntrain_length = int(full_dataset.__len__()*0.8)\nval_length = full_dataset.__len__() - train_length\ntrain_dataset, val_dataset = torch.utils.data.random_split(full_dataset, [train_length, val_length])\ntest_dataset = torchvision.datasets.CIFAR10(\n root='./data',\n train=False,\n download=True,\n transform=transform)\n\ntrainloader = torch.utils.data.DataLoader(\n train_dataset,\n batch_size=batch_size,\n shuffle=True,\n num_workers=0)\nvalloader = torch.utils.data.DataLoader(\n val_dataset,\n batch_size=batch_size,\n shuffle=True,\n num_workers=0)\ntestloader = torch.utils.data.DataLoader(\n test_dataset,\n batch_size=1,\n shuffle=False,\n num_workers=0)\nclasses = ['plane', 'car', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck']\n#############################################################################\n# END OF YOUR CODE #\n#############################################################################", "Files already downloaded and verified\nFiles already downloaded and verified\n" ] ], [ [ "---\n## Define the Network Architecture\n\nHere, you'll define a neural network named DTPNet, whose architecture resembles multiple layer perceptron (MLP) but adopts target propagation for parameter updating. You may use the following Pytorch functions to build it.\n\n* [Linear transformation layer](https://pytorch.org/docs/stable/generated/torch.nn.Linear.html)\n* [Non-linear activations](https://pytorch.org/docs/stable/generated/torch.tanh.html)\n* [Gradient computing operation](https://pytorch.org/docs/stable/autograd.html?highlight=torch%20autograd%20grad#torch.autograd.grad)\n\n\n### TODO: \n#### 1) Define DTPNet: Completing the __init__ function\n**Sizes of the hidden layers**: As the MLP takes a vector as input (ignoring the dimension of the batch), we should transform the image to a vector, whose dimension should be 3072 ($ 32\\times 32\\times 3 $)。 We suggest that the network architecture was 3072-1000-1000-1000-10. In each layer, the network uses the hyperbolic tangent as the activation function. You can also design these hyperparameters on your own.\n\n HINT: because we will compute the loss and gradient for each layer instead of chain rule, you might want to build a separate computational graph for each layer. Think of how to do this.\n#### 2) Define the forward path of DTPNet: Completing the \"forward\" function\nThe forward path involves computing unit values for all layers, that is,\n\n\\begin{align}\n\\text{for } i&= \\text{1 to } M \\\\\n & h_i \\leftarrow f_i(h_{i-1})\n\\end{align}\nwhere $f_i$ stands for the transformation layer of the DTPNet. Please refer to the original paper for more information.\n\n#### 3) Define the backward path of DTPNet: Completing the \"backward\" function\nThe computational details of the backward path are described in Algorithm 1 of the paper. It involves computing the targets, calculating the loss and calculate the gradients for each of the layers in the neural network in a top-down manner. \n\nTo make the code more readable, you need first define the \"backward\" function will call ``compute_target`` and ``reconstruction`` function.\n\n**Important:** The basic idea of the strategy to update the parameter for each layer are presented in the following section. You may need it for completing this function.\n\n##### 3.1) Calculate the targets for each layer\nFor the target of the highest layer, it is computed by\n$\\hat{\\mathbf{h}}_{M-1} \\leftarrow \\mathbf{h}_{M-1} - \\hat{\\eta} \\frac {\\partial L} {\\partial \\mathbf{h}_{M-1}}$, \\; ($L$ is the global loss)\n\nFor the targets of the lower layers, they are computed by\n\\begin{align}\n\\text{for } i&= M-1 \\text{ to } 2 \\\\\n& \\hat{\\mathbf{h}}_{i-1} \\leftarrow \\mathbf{h}_{i-1} - g_i(\\mathbf{h}_{i}) + g_i(\\hat{\\mathbf{h}}_{i})\n\\end{align}\n\n\n##### 3.2) Implement the reconstruction function \nThe reconstruction function involves $g_i(f_i(\\mathbf{h}_{i-1}))$.", "_____no_output_____" ] ], [ [ "import torch\nimport torch.nn as nn\nimport torch.optim as optim\n\n#Define a MLP\nclass DTPNet(nn.Module):\n def __init__(self,hidden_sizes=None):\n super(DTPNet, self).__init__()\n # construct the network\n #############################################################################\n self.num_layers=len(hidden_sizes)-1\n self.criterion = torch.nn.MSELoss()\n \n self.forward_model=dict()\n self.forward_optim=dict()\n for i in range(self.num_layers):\n self.forward_model['F{}'.format(i+1)]=nn.Linear(in_features=hidden_sizes[i],out_features=hidden_sizes[i+1]).to(device)\n self.forward_optim['F{}'.format(i+1)]=optim.RMSprop(params=self.forward_model['F{}'.format(i+1)].parameters(), lr=0.01)\n \n self.backward_model=dict()\n self.backward_optim=dict()\n for i in range(1,self.num_layers-1):\n self.backward_model['G{}'.format(i+1)]=nn.Linear(in_features=hidden_sizes[i+1],out_features=hidden_sizes[i]).to(device)\n self.backward_optim['G{}'.format(i+1)]=optim.RMSprop(params=self.backward_model['G{}'.format(i+1)].parameters(), lr=0.01)\n #############################################################################\n\n def forward(self, x):\n #Input: a batch of images, the size is [batchsize, 3072]\n #Output:the values of each layer in the network\n values = dict()\n #############################################################################\n for i,k in enumerate(self.forward_model):\n if i==self.num_layers-1:\n x=torch.softmax(self.forward_model[k](x),dim=-1)\n else:\n x=torch.tanh(self.forward_model[k](x))\n values['H{}'.format(i+1)]=x \n #############################################################################\n return values\n \n #以下函数定义有修改，详见实验报告\n def backward(self, values, global_loss):\n targets=self.compute_target(values, global_loss)\n\n loss_inv=self.compute_loss_inv(values)\n loss=self.compute_loss(values,targets,global_loss)\n \n for key in loss_inv.keys():\n grad=torch.autograd.grad(loss_inv[key], self.backward_model[key].parameters(), retain_graph = True)\n self.backward_model[key].weight.grad=grad[0]\n self.backward_model[key].bias.grad=grad[1]\n for key in loss.keys():\n grad=torch.autograd.grad(loss[key], self.forward_model[key].parameters(), retain_graph = True)\n self.forward_model[key].weight.grad=grad[0]\n self.forward_model[key].bias.grad=grad[1]\n\n for key in loss_inv.keys():\n self.backward_optim[key].step()\n for key in loss.keys():\n self.forward_optim[key].step()\n\n def compute_target(self, values, global_loss):\n #Input: values=[value_1,value_2,...,value_N]: the values of each layer (totally N layers) in the network\n #Output: loss=[target_1,target_2,...,target_N]: the targets of each layer (totally N layers) in the network\n targets = dict() \n lr0=0.327736332653 \n targets['H{}_'.format(self.num_layers-1)]=values['H{}'.format(self.num_layers-1)]-lr0*torch.autograd.grad(global_loss, values['H{}'.format(self.num_layers-1)], retain_graph = True)[0]\n for i in range(self.num_layers-1,1,-1):\n targets['H{}_'.format(i-1)]=values['H{}'.format(i-1)]-self.backward_model['G{}'.format(i)](values['H{}'.format(i)])+self.backward_model['G{}'.format(i)](targets['H{}_'.format(i)]) \n \n return targets\n \n def compute_loss(self, values, targets, global_loss):\n loss=dict()\n\n for i in range(self.num_layers-1):\n loss['F{}'.format(i+1)]=self.criterion(values['H{}'.format(i+1)],targets['H{}_'.format(i+1)])\n loss['F{}'.format(self.num_layers)]=global_loss\n return loss\n \n def compute_loss_inv(self, values):\n #Input: values=[value_1,value_2,...,value_N]: the values of each layer (totally N layers) in the network\n #output: loss=[value_2,...,value_N-1]: the reconstructed values of each layer (totally N layers) in the network\n loss_inv = dict()\n c=0.359829566008\n for i in range(self.num_layers-1,1,-1):\n temp=torch.randn(values['H{}'.format(i-1)].shape).to(device)*(c**1/2)+values['H{}'.format(i-1)]\n loss_inv['G{}'.format(i)]=self.criterion(self.backward_model['G{}'.format(i)](self.forward_model['F{}'.format(i)](temp)),temp)\n\n return loss_inv\n \n def set_train(self):\n for key in self.forward_model.keys():\n self.forward_model[key].train()\n for key in self.backward_model.keys():\n self.backward_model[key].train() \n \n def set_eval(self):\n for key in self.forward_model.keys():\n self.forward_model[key].eval()\n for key in self.backward_model.keys():\n self.backward_model[key].eval() \n\nmodel = DTPNet([3072,1000,1000,1000,10])\nprint('forward_model:',model.forward_model)#用字典定义模型层，打印模型层\nprint('backward_model:',model.backward_model)#用字典定义模型层，打印模型层", "forward_model: {'F1': Linear(in_features=3072, out_features=1000, bias=True), 'F2': Linear(in_features=1000, out_features=1000, bias=True), 'F3': Linear(in_features=1000, out_features=1000, bias=True), 'F4': Linear(in_features=1000, out_features=10, bias=True)}\nbackward_model: {'G2': Linear(in_features=1000, out_features=1000, bias=True), 'G3': Linear(in_features=1000, out_features=1000, bias=True)}\n" ] ], [ [ "### Specify [Loss Function](http://pytorch.org/docs/stable/nn.html#loss-functions) \n\nBefore the training process, you need to first specify your the loss function. For example, we set the negative log-likelihood as the the global loss of the image classification task. \n\n#### TODO: Define the loss", "_____no_output_____" ] ], [ [ "#############################################################################\n# TODO: define the loss function #\n#############################################################################\nglobal_loss_fn=torch.nn.NLLLoss()\n#############################################################################\n# END OF YOUR CODE #\n#############################################################################", "_____no_output_____" ] ], [ [ "---\n## Train the Network\nThe details of the training phrase of DTPNet can be found in Algorithm 1 of the paper, which is also posted here for your convenience. Please note that the training process of DTPNet is dramatically different from the traditional neural network. The main difference is that back-propagation uses chain rule to update parameters and the DTP calculates the targets and updates parameters layer by layer. Thus, so you can use the \"detach\" function in Pytorch to truncate the gradients to avoid auto-differetiation.\n\nHint: Instead using the backpropagation algorithm, we calculate the loss and then calculate the gradients for each layer, then we can use the gradients for each layer to update the parameters for each layer.\n## Examples to show how to update parameters of a single layer by Pytorch\nThere are two approaches to derive the gradients for the parameters of a neural network, namely, the optimizer module and autograd.grad() function. You can choose one of them or both of them to accomplish this work. Below, we will elaborate them in detail.\n### 1) Optimizer\nThe ``optimizer`` module controls the parameter updating for the neural network in Pytorch. We take a single layer of the multiple layer perceptron (MLP) as an example, which is defined by\n```\nsingle_layer = torch.nn.Linear(2,3)\n```\nWe can build an optimizer via Pytorch.\n```\noptimizer = torch.optim.RMSprop([{'params':single_layer.parameters(), 'lr': 1}])\n```\nThen we can update the parameters of single_layer by calling the ``step`` function of the optimizer, for example,\n```\nx = torch.randn(10,2)\ny = torch.randn(10,3)\npredict = single_layer(x)\nloss = ((predict-y)**2).sum() \nloss.backward()\noptimizer.step()\n```\n\n\n\n### 2) autograd.grad()\nThis function provides a way for the users to manually obtain the gradients for each of the parameters.\n```\nx = torch.randn(10,2)\ny = torch.randn(10,3)\npredict = MLP(x)\nloss = ((predict-y)**2).sum()\ngrad_weight = torch.autograd.grad(loss, MLP.weight, retain_graph = True)[0]\n\n```\n\n\n\n<img src=\"algo.png\" alt=\"drawing\" width=\"800\"/>\n\n\n\nRemember to look at how the training and validation loss decreases over time and print them.", "_____no_output_____" ] ], [ [ "#############################################################################\n# TODO: train and validation #\n#############################################################################\nfor epoch in range(30):\n train_loss = 0.0\n train_acc = 0.0\n train_correct_count = 0\n model.set_train()\n for i, data in enumerate(trainloader, 0):\n inputs, labels = data\n inputs=inputs.to(device)\n inputs=inputs.view([inputs.shape[0],3072])\n labels=labels.to(device)\n\n values = model(inputs)\n outputs = values['H{}'.format(model.num_layers)]\n train_correct_count += (torch.argmax(outputs,dim=1)==labels).sum().cpu().item()\n global_loss = global_loss_fn(outputs, labels)\n model.backward(values,global_loss)\n train_loss += global_loss.cpu().item()\n\n train_loss=train_loss/(i+1)\n train_acc=train_correct_count/(i+1)/batch_size\n\n val_loss = 0.0\n val_acc = 0.0\n val_correct_count = 0\n model.set_eval()\n for i, data in enumerate(valloader, 0):\n inputs, labels = data\n inputs=inputs.to(device).view(inputs.shape[0],3072)\n labels=labels.to(device)\n\n outputs = model(inputs)['H{}'.format(model.num_layers)]\n val_correct_count += (torch.argmax(outputs,dim=1)==labels).sum().cpu().item()\n global_loss = global_loss_fn(outputs, labels)\n val_loss += global_loss.cpu().item()\n val_loss=val_loss/(i+1)\n val_acc=val_correct_count/(i+1)/batch_size\n\n print('Epoch %d|Train_loss:%.3f Eval_loss:%.3f Train_acc:%.3f Eval_acc:%.3f'%(epoch+1,train_loss,val_loss,train_acc,val_acc))\n#############################################################################\n# END OF YOUR CODE #\n#############################################################################", "Epoch 1|Train_loss:-0.331 Eval_loss:-0.339 Train_acc:0.335 Eval_acc:0.341\nEpoch 2|Train_loss:-0.373 Eval_loss:-0.350 Train_acc:0.378 Eval_acc:0.352\nEpoch 3|Train_loss:-0.384 Eval_loss:-0.355 Train_acc:0.388 Eval_acc:0.356\nEpoch 4|Train_loss:-0.391 Eval_loss:-0.360 Train_acc:0.396 Eval_acc:0.362\nEpoch 5|Train_loss:-0.396 Eval_loss:-0.375 Train_acc:0.402 Eval_acc:0.378\nEpoch 6|Train_loss:-0.400 Eval_loss:-0.375 Train_acc:0.406 Eval_acc:0.375\nEpoch 7|Train_loss:-0.405 Eval_loss:-0.363 Train_acc:0.411 Eval_acc:0.364\nEpoch 8|Train_loss:-0.407 Eval_loss:-0.371 Train_acc:0.413 Eval_acc:0.372\nEpoch 9|Train_loss:-0.409 Eval_loss:-0.381 Train_acc:0.416 Eval_acc:0.383\nEpoch 10|Train_loss:-0.413 Eval_loss:-0.380 Train_acc:0.419 Eval_acc:0.378\nEpoch 11|Train_loss:-0.414 Eval_loss:-0.383 Train_acc:0.421 Eval_acc:0.381\nEpoch 12|Train_loss:-0.417 Eval_loss:-0.376 Train_acc:0.422 Eval_acc:0.374\nEpoch 13|Train_loss:-0.418 Eval_loss:-0.380 Train_acc:0.424 Eval_acc:0.381\nEpoch 14|Train_loss:-0.419 Eval_loss:-0.383 Train_acc:0.426 Eval_acc:0.384\nEpoch 15|Train_loss:-0.423 Eval_loss:-0.384 Train_acc:0.427 Eval_acc:0.383\nEpoch 16|Train_loss:-0.424 Eval_loss:-0.384 Train_acc:0.429 Eval_acc:0.384\nEpoch 17|Train_loss:-0.425 Eval_loss:-0.386 Train_acc:0.431 Eval_acc:0.384\nEpoch 18|Train_loss:-0.427 Eval_loss:-0.387 Train_acc:0.433 Eval_acc:0.385\nEpoch 19|Train_loss:-0.429 Eval_loss:-0.381 Train_acc:0.433 Eval_acc:0.378\nEpoch 20|Train_loss:-0.427 Eval_loss:-0.381 Train_acc:0.434 Eval_acc:0.378\nEpoch 21|Train_loss:-0.427 Eval_loss:-0.380 Train_acc:0.436 Eval_acc:0.378\nEpoch 22|Train_loss:-0.431 Eval_loss:-0.385 Train_acc:0.436 Eval_acc:0.385\nEpoch 23|Train_loss:-0.432 Eval_loss:-0.378 Train_acc:0.437 Eval_acc:0.374\nEpoch 24|Train_loss:-0.433 Eval_loss:-0.388 Train_acc:0.438 Eval_acc:0.385\nEpoch 25|Train_loss:-0.435 Eval_loss:-0.376 Train_acc:0.439 Eval_acc:0.371\nEpoch 26|Train_loss:-0.432 Eval_loss:-0.386 Train_acc:0.440 Eval_acc:0.383\nEpoch 27|Train_loss:-0.435 Eval_loss:-0.385 Train_acc:0.441 Eval_acc:0.382\nEpoch 28|Train_loss:-0.435 Eval_loss:-0.385 Train_acc:0.443 Eval_acc:0.381\nEpoch 29|Train_loss:-0.440 Eval_loss:-0.384 Train_acc:0.443 Eval_acc:0.382\nEpoch 30|Train_loss:-0.437 Eval_loss:-0.384 Train_acc:0.445 Eval_acc:0.381\n" ] ], [ [ "---\n## Test the Trained Network\n\nTest your trained model on previously unseen data and print the test accuracy of each class and the whole! Try your best to get a better accuracy.", "_____no_output_____" ] ], [ [ "#############################################################################\n# TODO: test the trained network #\n#############################################################################\nclass_count=np.zeros(10,dtype=int)\ncorrect_count=np.zeros(10,dtype=int)\nmodel.set_eval()\nfor i, data in enumerate(testloader, 0):\n inputs, labels = data\n inputs=inputs.to(device)\n inputs=inputs.view([inputs.shape[0],3072])\n outputs = torch.argmax(model(inputs)['H{}'.format(model.num_layers)],dim=1).cpu().item()\n class_count[labels]+=1\n if outputs==labels:\n correct_count[labels]+=1\nfor i in range(10):\n print('Test|Class{}('.format(i+1)+classes[i]+')-acc={}'.format(correct_count[i]/class_count[i]))\nprint('\\nTest|Overall-acc={}'.format(np.sum(correct_count/10000)))\n#############################################################################\n# END OF YOUR CODE #\n#############################################################################", "Test|Class1(plane)-acc=0.48\nTest|Class2(car)-acc=0.453\nTest|Class3(bird)-acc=0.246\nTest|Class4(cat)-acc=0.171\nTest|Class5(deer)-acc=0.311\nTest|Class6(dog)-acc=0.339\nTest|Class7(frog)-acc=0.464\nTest|Class8(horse)-acc=0.403\nTest|Class9(ship)-acc=0.561\nTest|Class10(truck)-acc=0.496\n\nTest|Overall-acc=0.39239999999999997\n" ] ], [ [ "### Question: What are your model's weaknesses during your experiment and how might they be improved?\n+ 本实验只使用了较为简单的MLP网络，由于网络结构过于简单，最终分类结果较差，仅用作DTP算法是示意。实际上，MLP网络输入时先将图像展开为列向量，这种操作不易获得图像上邻近位置的信息，因而分类效果较差。使用更加复杂的卷积神经网络应当可以达到更好的效果，将卷积神经网络中的一个卷积层或一个卷积块看做一个广义的前向计算函数${f_i}$即可使用DTP算法进行训练。由于任务一中已经验证了ResNet的分类表现，此处不再重复实现。\n+ DTP算法中涉及较多的超参数选择，如各层局部参数更新的学习率、随机高斯噪声的功率谱密度以及全局损失函数、优化器的选择。在这些超参数上进行进一步调优可能在相同的网络结构上实现更好的测试表现。\n", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ] ]

ec7b366144bddbb77799e165d584ae9c925192d2

7,950

ipynb

Jupyter Notebook

SuperdenseCoding/SuperdenseCoding.ipynb

gviscoli/QuantumKatas

99eda654938d1e983c7232f876dedcb623d4af70

[ "MIT" ]

1

2020-07-17T16:37:25.000Z

SuperdenseCoding/SuperdenseCoding.ipynb

mftsang/QuantumKatas

dec9fa891535350e1d5868a86f0cf10be645fe70

[ "MIT" ]

null

SuperdenseCoding/SuperdenseCoding.ipynb

mftsang/QuantumKatas

dec9fa891535350e1d5868a86f0cf10be645fe70

[ "MIT" ]

null

37.5

296

0.606164

[ [ [ "empty" ] ] ]

[ "empty" ]

[ [ "empty" ] ]

ec7b3a019e2e3f668319c43bed17e9a88d0caa7b

555,966

ipynb

Jupyter Notebook

tutorials/Certification_Trainings/Public/13.NLU_crashcourse_every_Spark_NLP_Model_in_one_line.ipynb

hatrungduc/spark-nlp-workshop

4a4ec0195d1d3d847261df9ef2df7aa5f95bbaec

[ "Apache-2.0" ]

687

2018-09-07T03:45:39.000Z

2022-03-20T17:11:20.000Z

tutorials/Certification_Trainings/Public/13.NLU_crashcourse_every_Spark_NLP_Model_in_one_line.ipynb

hatrungduc/spark-nlp-workshop

4a4ec0195d1d3d847261df9ef2df7aa5f95bbaec

[ "Apache-2.0" ]

89

2018-09-18T02:04:42.000Z

2022-02-24T18:22:27.000Z

tutorials/Certification_Trainings/Public/13.NLU_crashcourse_every_Spark_NLP_Model_in_one_line.ipynb

hatrungduc/spark-nlp-workshop

4a4ec0195d1d3d847261df9ef2df7aa5f95bbaec

[ "Apache-2.0" ]

407

2018-09-07T03:45:44.000Z

2022-03-20T05:12:25.000Z

117.292405

74,482

0.770351

[ [ [ "![JohnSnowLabs](https://nlp.johnsnowlabs.com/assets/images/logo.png)\n\n[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/JohnSnowLabs/spark-nlp-workshop/blob/master/tutorials/Certification_Trainings/Public/13.NLU_crashcourse_every_Spark_NLP_Model_in_one_line.ipynb)\n\n\n# NLU 20 Minutes Crashcourse - the fast & easy Data Science route\n\n\n## Spark NLP vs NLU, whats the difference?\n[NLU](https://nlu.johnsnowlabs.com/) is a Python wrapper around Spark NLP. It gives you all of Spark NLPs features in 1 line of code and supports all the common Pythonic Data Structures like Pandas and Modin Dataframes. It's the ultimate tool to swifitly explore the models in Spark NLP and evaluate them for different use cases. With NLU you can : \n\n- Use any Model in Spark NLP in 1 line of code, by leveragin NLU's automatic pipeline generation\n- Predicts on most common Python data structure like strings and Pandas array\n- Transforms the Spark NLP Dataframe structure into a pretty Pandas Dataframe structure, which can be configured to `document`, `sentence`, `chunk`, and`token` output levels.\n- Enables you to visualize outputs of models in 1 line of code using the [viz methods](https://nlu.johnsnowlabs.com/docs/en/viz_examples)\n- Use a [Powerful Streamlit Dashboard and Buildingblocks](https://nlu.johnsnowlabs.com/docs/en/streamlit_viz_examples) that enable to you to test out any model in 0 lines of code using a GUI. In addition, you can compare embeddings using various Manifold and Matrix Decomposition visualizations \n\n\n\nUnder the hood, NLU automagically generates a Spark NLP pipeline for you, based on the model name you put in `nlu.load()`. All the NLP data transformations and predictions are still beeing performed by Spark NLP, NLU just gives you the most simple API possible for all of the features. \n\n\n\nNLU's core processing peformed on the data returned by Spark NLP is currently happening via the Numpy engine and will not be distributed by default, this means NLU is slower and takes up more memory than Spark NLP, because there is additional computation performed on your data. \nYou have the option to set **.predict(return_spark_df =True)**. With this setting, NLU all computations will be **distributed** but NLU will not peform further data processing on the datafame, so you will get the standard Spark NLP Dataframe structure. \n\n\n\nThis short notebook will teach you a lot of things!\n- Sentiment classification, binary, multi class and regressive\n- Extract Parts of Speech (POS)\n- Extract Named Entities (NER)\n- Extract Keywords (YAKE!)\n- Answer Open and Closed book questions with T5\n- Summarize text and more with Multi task T5\n- Translate text with Microsofts Marian Model\n- Train a Multi Lingual Classifier for 100+ languages from a dataset with just one language\n\n## NLU Webinars and Video Tutorials\n- [NLU & Streamlit Tutorial](https://vimeo.com/579508034#)\n- [Crash course of the 50 + Medical Domains and the 200+ Healtchare models in NLU](https://www.youtube.com/watch?v=gGDsZXt1SF8)\n- [Multi Lingual NLU Webinar - Tutorial on Chinese News dataset](https://www.youtube.com/watch?v=ftAOqJuxnV4)\n- [John Snow Labs NLU: Become a Data Science Superhero with One Line of Python code](https://events.johnsnowlabs.com/john-snow-labs-nlu-become-a-data-science-superhero-with-one-line-of-python-code?hsCtaTracking=c659363c-2188-4c86-945f-5cfb7b42fcfc%7C8b2b188b-92a3-48ba-ad7e-073b384425b0)\n- [Python Web Def Conf - Python's NLU library: 1,000+ Models, 200+ Languages, State of the Art Accuracy, 1 Line of Code](https://2021.pythonwebconf.com/presentations/john-snow-labs-nlu-the-simplicity-of-python-the-power-of-spark-nlp)\n- [NYC/DC NLP Meetup with NLU](https://youtu.be/hJR9m3NYnwk?t=2155)\n\n\n## More ressources \n- [Join our Slack](https://join.slack.com/t/spark-nlp/shared_invite/zt-lutct9gm-kuUazcyFKhuGY3_0AMkxqA)\n- [NLU Website](https://nlu.johnsnowlabs.com/)\n- [NLU Github](https://github.com/JohnSnowLabs/nlu)\n- [Many more NLU example tutorials](https://github.com/JohnSnowLabs/nlu/tree/master/examples)\n- [Overview of every powerful nlu 1-liner](https://nlu.johnsnowlabs.com/docs/en/examples)\n- [Checkout the Modelshub for an overview of all models](https://nlp.johnsnowlabs.com/models) \n- [Checkout the NLU Namespace where you can find every model as a tabel](https://nlu.johnsnowlabs.com/docs/en/spellbook)\n- [Intro to NLU article](https://medium.com/spark-nlp/1-line-of-code-350-nlp-models-with-john-snow-labs-nlu-in-python-2f1c55bba619)\n- [Indepth and easy Sentence Similarity Tutorial, with StackOverflow Questions using BERTology embeddings](https://medium.com/spark-nlp/easy-sentence-similarity-with-bert-sentence-embeddings-using-john-snow-labs-nlu-ea078deb6ebf)\n- [1 line of Python code for BERT, ALBERT, ELMO, ELECTRA, XLNET, GLOVE, Part of Speech with NLU and t-SNE](https://medium.com/spark-nlp/1-line-of-code-for-bert-albert-elmo-electra-xlnet-glove-part-of-speech-with-nlu-and-t-sne-9ebcd5379cd)", "_____no_output_____" ], [ "# Install NLU\nYou need Java8, Pyspark and Spark-NLP installed, [see the installation guide for instructions](https://nlu.johnsnowlabs.com/docs/en/install). If you need help or run into troubles, [ping us on slack :)](https://join.slack.com/t/spark-nlp/shared_invite/zt-lutct9gm-kuUazcyFKhuGY3_0AMkxqA) ", "_____no_output_____" ] ], [ [ "!wget https://setup.johnsnowlabs.com/nlu/colab.sh -O - | bash\n\nimport nlu", "_____no_output_____" ] ], [ [ "# Simple NLU basics on Strings", "_____no_output_____" ], [ "## Context based spell Checking in 1 line\n\n![Spell Check](https://i.imgflip.com/52wb7w.jpg)", "_____no_output_____" ] ], [ [ "nlu.load('spell').predict('I also liek to live dangertus')", "spellcheck_dl download started this may take some time.\nApproximate size to download 111.4 MB\n[OK!]\nsentence_detector_dl download started this may take some time.\nApproximate size to download 354.6 KB\n[OK!]\n" ] ], [ [ "## Binary Sentiment classification in 1 Line\n![Binary Sentiment](https://cdn.pixabay.com/photo/2015/11/13/10/07/smiley-1041796_960_720.jpg)\n", "_____no_output_____" ] ], [ [ "nlu.load('sentiment').predict('I love NLU and rainy days!')", "sentimentdl_glove_imdb download started this may take some time.\nApproximate size to download 8.7 MB\n[OK!]\nglove_100d download started this may take some time.\nApproximate size to download 145.3 MB\n[OK!]\nglove_100d download started this may take some time.\nApproximate size to download 145.3 MB\n[OK!]\nsentence_detector_dl download started this may take some time.\nApproximate size to download 354.6 KB\n[OK!]\n" ] ], [ [ "## Part of Speech (POS) in 1 line\n![Parts of Speech](https://image.shutterstock.com/image-photo/blackboard-background-written-colorful-chalk-600w-1166166529.jpg)\n\n|Tag |Description | Example|\n|------|------------|------|\n|CC| Coordinating conjunction | This batch of mushroom stew is savory **and** delicious |\n|CD| Cardinal number | Here are **five** coins |\n|DT| Determiner | **The** bunny went home |\n|EX| Existential there | **There** is a storm coming |\n|FW| Foreign word | I'm having a **déjà vu** |\n|IN| Preposition or subordinating conjunction | He is cleverer **than** I am |\n|JJ| Adjective | She wore a **beautiful** dress |\n|JJR| Adjective, comparative | My house is **bigger** than yours |\n|JJS| Adjective, superlative | I am the **shortest** person in my family |\n|LS| List item marker | A number of things need to be considered before starting a business **,** such as premises **,** finance **,** product demand **,** staffing and access to customers |\n|MD| Modal | You **must** stop when the traffic lights turn red |\n|NN| Noun, singular or mass | The **dog** likes to run |\n|NNS| Noun, plural | The **cars** are fast |\n|NNP| Proper noun, singular | I ordered the chair from **Amazon** |\n|NNPS| Proper noun, plural | We visted the **Kennedys** |\n|PDT| Predeterminer | **Both** the children had a toy |\n|POS| Possessive ending | I built the dog'**s** house |\n|PRP| Personal pronoun | **You** need to stop |\n|PRP$| Possessive pronoun | Remember not to judge a book by **its** cover |\n|RB| Adverb | The dog barks **loudly** |\n|RBR| Adverb, comparative | Could you sing more **quietly** please? |\n|RBS| Adverb, superlative | Everyone in the race ran fast, but John ran **the fastest** of all |\n|RP| Particle | He ate **up** all his dinner |\n|SYM| Symbol | What are you doing **?** |\n|TO| to | Please send it back **to** me |\n|UH| Interjection | **Wow!** You look gorgeous |\n|VB| Verb, base form | We **play** soccer |\n|VBD| Verb, past tense | I **worked** at a restaurant |\n|VBG| Verb, gerund or present participle | **Smoking** kills people |\n|VBN| Verb, past participle | She has **done** her homework |\n|VBP| Verb, non-3rd person singular present | You **flit** from place to place |\n|VBZ| Verb, 3rd person singular present | He never **calls** me |\n|WDT| Wh-determiner | The store honored the complaints, **which** were less than 25 days old |\n|WP| Wh-pronoun | **Who** can help me? |\n|WP\\$| Possessive wh-pronoun | **Whose** fault is it? |\n|WRB| Wh-adverb | **Where** are you going? |", "_____no_output_____" ] ], [ [ "nlu.load('pos').predict('POS assigns each token in a sentence a grammatical label')", "pos_anc download started this may take some time.\nApproximate size to download 3.9 MB\n[OK!]\nsentence_detector_dl download started this may take some time.\nApproximate size to download 354.6 KB\n[OK!]\n" ] ], [ [ "## Named Entity Recognition (NER) in 1 line\n\n![NER](http://ckl-it.de/wp-content/uploads/2021/02/ner-1.png)\n\n|Type | \tDescription |\n|------|--------------|\n| PERSON | \tPeople, including fictional like **Harry Potter** |\n| NORP | \tNationalities or religious or political groups like the **Germans** |\n| FAC | \tBuildings, airports, highways, bridges, etc. like **New York Airport** |\n| ORG | \tCompanies, agencies, institutions, etc. like **Microsoft** |\n| GPE | \tCountries, cities, states. like **Germany** |\n| LOC | \tNon-GPE locations, mountain ranges, bodies of water. Like the **Sahara desert**|\n| PRODUCT | \tObjects, vehicles, foods, etc. (Not services.) like **playstation** |\n| EVENT | \tNamed hurricanes, battles, wars, sports events, etc. like **hurricane Katrina**|\n| WORK_OF_ART | \tTitles of books, songs, etc. Like **Mona Lisa** |\n| LAW | \tNamed documents made into laws. Like : **Declaration of Independence** |\n| LANGUAGE | \tAny named language. Like **Turkish**|\n| DATE | \tAbsolute or relative dates or periods. Like every second **friday**|\n| TIME | \tTimes smaller than a day. Like **every minute**|\n| PERCENT | \tPercentage, including ”%“. Like **55%** of workers enjoy their work |\n| MONEY | \tMonetary values, including unit. Like **50$** for those pants |\n| QUANTITY | \tMeasurements, as of weight or distance. Like this person weights **50kg** |\n| ORDINAL | \t“first”, “second”, etc. Like David placed **first** in the tournament |\n| CARDINAL | \tNumerals that do not fall under another type. Like **hundreds** of models are avaiable in NLU |\n", "_____no_output_____" ] ], [ [ "nlu.load('ner').predict(\"John Snow Labs congratulates the Amarican John Biden to winning the American election!\", output_level='chunk')", "onto_recognize_entities_sm download started this may take some time.\nApprox size to download 160.1 MB\n[OK!]\n" ] ], [ [ "# Let's apply NLU to a dataset!\n\n<div>\n<img src=\"http://ckl-it.de/wp-content/uploads/2021/02/crypto.jpeg \" width=\"400\" height=\"250\" >\n</div>\n", "_____no_output_____" ] ], [ [ "import pandas as pd \nimport nlu\n!wget http://ckl-it.de/wp-content/uploads/2020/12/small_btc.csv \ndf = pd.read_csv('/content/small_btc.csv').iloc[0:5000].title\ndf\n\n", "--2021-10-13 10:31:22-- http://ckl-it.de/wp-content/uploads/2020/12/small_btc.csv\nResolving ckl-it.de (ckl-it.de)... 217.160.0.108, 2001:8d8:100f:f000::209\nConnecting to ckl-it.de (ckl-it.de)|217.160.0.108|:80... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 22244914 (21M) [text/csv]\nSaving to: ‘small_btc.csv’\n\nsmall_btc.csv 100%[===================>] 21.21M 13.3MB/s in 1.6s \n\n2021-10-13 10:31:24 (13.3 MB/s) - ‘small_btc.csv’ saved [22244914/22244914]\n\n" ] ], [ [ "## NER on a Crypto News dataset\n### The **NER** model which you can load via `nlu.load('ner')` recognizes 18 different classes in your dataset.\nWe set output level to chunk, so that we get 1 row per NER class.\n\n\n#### Predicted entities:\n\n\nNER is avaiable in many languages, which you can [find in the John Snow Labs Modelshub](https://nlp.johnsnowlabs.com/models)", "_____no_output_____" ], [ "", "_____no_output_____" ] ], [ [ "ner_df = nlu.load('ner').predict(df, output_level = 'chunk')\nner_df ", "onto_recognize_entities_sm download started this may take some time.\nApprox size to download 160.1 MB\n[OK!]\n" ] ], [ [ "### Top 50 Named Entities", "_____no_output_____" ] ], [ [ "ner_df.entities.value_counts()[:100].plot.barh(figsize = (16,20))", "_____no_output_____" ] ], [ [ "### Top 50 Named Entities which are PERSONS", "_____no_output_____" ] ], [ [ "ner_df[ner_df.entities_class == 'PERSON'].entities.value_counts()[:50].plot.barh(figsize=(18,20), title ='Top 50 Occuring Persons in the dataset')", "_____no_output_____" ] ], [ [ "### Top 50 Named Entities which are Countries/Cities/States", "_____no_output_____" ] ], [ [ "ner_df[ner_df.entities_class == 'GPE'].entities.value_counts()[:50].plot.barh(figsize=(18,20),title ='Top 50 Countries/Cities/States Occuring in the dataset')", "_____no_output_____" ] ], [ [ "### Top 50 Named Entities which are PRODUCTS ", "_____no_output_____" ] ], [ [ "ner_df[ner_df.entities_class == 'PRODUCT'].entities.value_counts()[:50].plot.barh(figsize=(18,20),title ='Top 50 products occuring in the dataset')", "_____no_output_____" ] ], [ [ "### Top 50 Named Entities which are ORGANIZATIONS", "_____no_output_____" ] ], [ [ "ner_df[ner_df.entities_class == 'ORG'].entities.value_counts()[:50].plot.barh(figsize=(18,20),title ='Top 50 products occuring in the dataset')", "_____no_output_____" ] ], [ [ "## YAKE on a Crypto News dataset\n### The **YAKE!** model (Yet Another Keyword Extractor) is a **unsupervised** keyword extraction algorithm.\nYou can load it via which you can load via `nlu.load('yake')`. It has no weights and is very fast.\nIt has various parameters that can be configured to influence which keywords are beeing extracted, [here for an more indepth YAKE guide](https://github.com/JohnSnowLabs/nlu/blob/master/examples/webinars_conferences_etc/multi_lingual_webinar/1_NLU_base_features_on_dataset_with_YAKE_Lemma_Stemm_classifiers_NER_.ipynb)", "_____no_output_____" ] ], [ [ "yake_df = nlu.load('yake').predict(df)\nyake_df", "sentence_detector_dl download started this may take some time.\nApproximate size to download 354.6 KB\n[OK!]\n" ] ], [ [ "### Top 50 extracted Keywords with YAKE!", "_____no_output_____" ] ], [ [ "yake_df.explode('keywords').keywords.value_counts()[0:50].plot.barh(figsize=(14,18))", "_____no_output_____" ] ], [ [ "## Binary Sentimental Analysis and Distribution on a dataset", "_____no_output_____" ] ], [ [ "sent_df = nlu.load('sentiment').predict(df)\nsent_df", "sentimentdl_glove_imdb download started this may take some time.\nApproximate size to download 8.7 MB\n[OK!]\nglove_100d download started this may take some time.\nApproximate size to download 145.3 MB\n[OK!]\nglove_100d download started this may take some time.\nApproximate size to download 145.3 MB\n[OK!]\nsentence_detector_dl download started this may take some time.\nApproximate size to download 354.6 KB\n[OK!]\n" ], [ "sent_df.sentiment.value_counts().plot.bar(title='Sentiment ')", "_____no_output_____" ] ], [ [ "## Emotional Analysis and Distribution of Headlines ", "_____no_output_____" ] ], [ [ "emo_df = nlu.load('emotion').predict(df)\nemo_df", "classifierdl_use_emotion download started this may take some time.\nApproximate size to download 21.3 MB\n[OK!]\ntfhub_use download started this may take some time.\nApproximate size to download 923.7 MB\n[OK!]\nsentence_detector_dl download started this may take some time.\nApproximate size to download 354.6 KB\n[OK!]\n" ], [ "emo_df.emotion.value_counts().plot.bar(title='Emotion Distribution')\n", "_____no_output_____" ] ], [ [ "**Make sure to restart your notebook again** before starting the next section", "_____no_output_____" ] ], [ [ "print(\"Please restart kernel if you are in google colab to free up RAM\")\n1+'wait'\n", "_____no_output_____" ] ], [ [ "# Answer **Closed Book** and Open **Book Questions** with Google's T5!\n\n\n![T5 GIF](https://1.bp.blogspot.com/-o4oiOExxq1s/Xk26XPC3haI/AAAAAAAAFU8/NBlvOWB84L0PTYy9TzZBaLf6fwPGJTR0QCLcBGAsYHQ/s1600/image3.gif)\n\nYou can load the **question answering** model with `nlu.load('en.t5')`", "_____no_output_____" ] ], [ [ "import nlu\n# Load question answering T5 model\nt5_closed_question = nlu.load('en.t5')", "google_t5_small_ssm_nq download started this may take some time.\nApproximate size to download 139 MB\n[OK!]\n" ] ], [ [ "## Answer **Closed Book Questions** \nClosed book means that no additional context is given and the model must answer the question with the knowledge stored in it's weights", "_____no_output_____" ] ], [ [ "t5_closed_question.predict(\"Who is president of Nigeria?\")", "_____no_output_____" ], [ "t5_closed_question.predict(\"What is the most common language in India?\")", "_____no_output_____" ], [ "t5_closed_question.predict(\"What is the capital of Germany?\")", "_____no_output_____" ] ], [ [ "## Answer **Open Book Questions** \nThese are questions where we give the model some additional context, that is used to answer the question", "_____no_output_____" ] ], [ [ "t5_open_book = nlu.load('answer_question')", "t5_base download started this may take some time.\nApproximate size to download 446 MB\n[OK!]\n" ], [ "context = 'Peters last week was terrible! He had an accident and broke his leg while skiing!'\nquestion1 = 'Why was peters week so bad?' \nquestion2 = 'How did peter broke his leg?' \n\nt5_open_book.predict([question1+context, question2 + context]) ", "_____no_output_____" ], [ "# Ask T5 questions in the context of a News Article\nquestion1 = 'Who is Jack ma?'\nquestion2 = 'Who is founder of Alibaba Group?'\nquestion3 = 'When did Jack Ma re-appear?'\nquestion4 = 'How did Alibaba stocks react?'\nquestion5 = 'Whom did Jack Ma meet?'\nquestion6 = 'Who did Jack Ma hide from?'\n\n\n# from https://www.bbc.com/news/business-55728338 \nnews_article_context = \"\"\" context:\nAlibaba Group founder Jack Ma has made his first appearance since Chinese regulators cracked down on his business empire.\nHis absence had fuelled speculation over his whereabouts amid increasing official scrutiny of his businesses.\nThe billionaire met 100 rural teachers in China via a video meeting on Wednesday, according to local government media.\nAlibaba shares surged 5% on Hong Kong's stock exchange on the news.\n\"\"\"\n\nquestions = [\n question1+ news_article_context,\n question2+ news_article_context,\n question3+ news_article_context,\n question4+ news_article_context,\n question5+ news_article_context,\n question6+ news_article_context,]\n\n", "_____no_output_____" ], [ "t5_open_book.predict(questions)", "_____no_output_____" ] ], [ [ "# Multi Problem T5 model for Summarization and more\nThe main T5 model was trained for over 20 tasks from the SQUAD/GLUE/SUPERGLUE datasets. See [this notebook](https://github.com/JohnSnowLabs/nlu/blob/master/examples/webinars_conferences_etc/multi_lingual_webinar/7_T5_SQUAD_GLUE_SUPER_GLUE_TASKS.ipynb) for a demo of all tasks \n\n\n# Overview of every task available with T5\n[The T5 model](https://arxiv.org/pdf/1910.10683.pdf) is trained on various datasets for 17 different tasks which fall into 8 categories.\n\n\n\n1. Text summarization\n2. Question answering\n3. Translation\n4. Sentiment analysis\n5. Natural Language inference\n6. Coreference resolution\n7. Sentence Completion\n8. Word sense disambiguation\n\n### Every T5 Task with explanation:\n|Task Name | Explanation | \n|----------|--------------|\n|[1.CoLA](https://nyu-mll.github.io/CoLA/) | Classify if a sentence is gramaticaly correct|\n|[2.RTE](https://dl.acm.org/doi/10.1007/11736790_9) | Classify whether if a statement can be deducted from a sentence|\n|[3.MNLI](https://arxiv.org/abs/1704.05426) | Classify for a hypothesis and premise whether they contradict or contradict each other or neither of both (3 class).|\n|[4.MRPC](https://www.aclweb.org/anthology/I05-5002.pdf) | Classify whether a pair of sentences is a re-phrasing of each other (semantically equivalent)|\n|[5.QNLI](https://arxiv.org/pdf/1804.07461.pdf) | Classify whether the answer to a question can be deducted from an answer candidate.|\n|[6.QQP](https://www.quora.com/q/quoradata/First-Quora-Dataset-Release-Question-Pairs) | Classify whether a pair of questions is a re-phrasing of each other (semantically equivalent)|\n|[7.SST2](https://www.aclweb.org/anthology/D13-1170.pdf) | Classify the sentiment of a sentence as positive or negative|\n|[8.STSB](https://www.aclweb.org/anthology/S17-2001/) | Classify the sentiment of a sentence on a scale from 1 to 5 (21 Sentiment classes)|\n|[9.CB](https://ojs.ub.uni-konstanz.de/sub/index.php/sub/article/view/601) | Classify for a premise and a hypothesis whether they contradict each other or not (binary).|\n|[10.COPA](https://www.aaai.org/ocs/index.php/SSS/SSS11/paper/view/2418/0) | Classify for a question, premise, and 2 choices which choice the correct choice is (binary).|\n|[11.MultiRc](https://www.aclweb.org/anthology/N18-1023.pdf) | Classify for a question, a paragraph of text, and an answer candidate, if the answer is correct (binary),|\n|[12.WiC](https://arxiv.org/abs/1808.09121) | Classify for a pair of sentences and a disambigous word if the word has the same meaning in both sentences.|\n|[13.WSC/DPR](https://www.aaai.org/ocs/index.php/KR/KR12/paper/view/4492/0) | Predict for an ambiguous pronoun in a sentence what it is referring to. |\n|[14.Summarization](https://arxiv.org/abs/1506.03340) | Summarize text into a shorter representation.|\n|[15.SQuAD](https://arxiv.org/abs/1606.05250) | Answer a question for a given context.|\n|[16.WMT1.](https://arxiv.org/abs/1706.03762) | Translate English to German|\n|[17.WMT2.](https://arxiv.org/abs/1706.03762) | Translate English to French|\n|[18.WMT3.](https://arxiv.org/abs/1706.03762) | Translate English to Romanian|\n\n", "_____no_output_____" ] ], [ [ "# Load the Multi Task Model T5\nt5_multi = nlu.load('en.t5.base')", "t5_base download started this may take some time.\nApproximate size to download 446 MB\n[OK!]\n" ], [ "# https://www.reuters.com/article/instant-article/idCAKBN2AA2WF\ntext = \"\"\"(Reuters) - Mastercard Inc said on Wednesday it was planning to offer support for some cryptocurrencies on its network this year, joining a string of big-ticket firms that have pledged similar support.\n\nThe credit-card giant’s announcement comes days after Elon Musk’s Tesla Inc revealed it had purchased $1.5 billion of bitcoin and would soon accept it as a form of payment.\n\nAsset manager BlackRock Inc and payments companies Square and PayPal have also recently backed cryptocurrencies.\n\nMastercard already offers customers cards that allow people to transact using their cryptocurrencies, although without going through its network.\n\n\"Doing this work will create a lot more possibilities for shoppers and merchants, allowing them to transact in an entirely new form of payment. This change may open merchants up to new customers who are already flocking to digital assets,\" Mastercard said. (mstr.cd/3tLaPZM)\n\nMastercard specified that not all cryptocurrencies will be supported on its network, adding that many of the hundreds of digital assets in circulation still need to tighten their compliance measures.\n\nMany cryptocurrencies have struggled to win the trust of mainstream investors and the general public due to their speculative nature and potential for money laundering.\n\"\"\"\nt5_multi['t5'].setTask('summarize ') \nshort = t5_multi.predict(text)\nshort", "_____no_output_____" ], [ "print(f\"Original Length {len(short.document.iloc[0])} Summarized Length : {len(short.t5.iloc[0])} \\n summarized text :{short.t5.iloc[0]} \")\n", "Original Length 1277 Summarized Length : 352 \n summarized text :mastercard said on Wednesday it was planning to offer support for some cryptocurrencies on its network this year . the credit-card giant’s announcement comes days after Elon Musk’s Tesla Inc revealed it had purchased $1.5 billion of bitcoin . asset manager blackrock and payments companies Square and PayPal have also recently backed cryptocurrencies . \n" ], [ "short.t5.iloc[0]", "_____no_output_____" ] ], [ [ "**Make sure to restart your notebook again** before starting the next section", "_____no_output_____" ] ], [ [ "print(\"Please restart kernel if you are in google colab and run next cell after the restart to configure java 8 back\")\n1+'wait'\n", "Please restart kernel if you are in google colab and run next cell after the restart to configure java 8 back\n" ] ], [ [ "# Translate between more than 200 Languages with [ Microsofts Marian Models](https://marian-nmt.github.io/publications/)\n\nMarian is an efficient, free Neural Machine Translation framework mainly being developed by the Microsoft Translator team (646+ pretrained models & pipelines in 192+ languages)\nYou need to specify the language your data is in as `start_language` and the language you want to translate to as `target_language`. \n The language references must be [ISO language codes](https://en.wikipedia.org/wiki/List_of_ISO_639-1_codes)\n\n`nlu.load('<start_language>.translate_to.<target_language>')` \n\n**Translate Turkish to English:** \n`nlu.load('tr.translate_to.en')`\n\n**Translate English to French:** \n`nlu.load('en.translate_to.fr')`\n\n\n**Translate French to Hebrew:** \n`nlu.load('fr.translate_to.he')`\n\n\n\n\n\n![Languages](https://camo.githubusercontent.com/b548abf3d1f9657d01fd74404354ec49fc11eea0/687474703a2f2f636b6c2d69742e64652f77702d636f6e74656e742f75706c6f6164732f323032312f30322f666c6167732e6a706567)", "_____no_output_____" ] ], [ [ "import nlu\nimport pandas as pd\n!wget http://ckl-it.de/wp-content/uploads/2020/12/small_btc.csv \ndf = pd.read_csv('/content/small_btc.csv').iloc[0:20].title", "--2021-10-13 10:44:10-- http://ckl-it.de/wp-content/uploads/2020/12/small_btc.csv\nResolving ckl-it.de (ckl-it.de)... 217.160.0.108, 2001:8d8:100f:f000::209\nConnecting to ckl-it.de (ckl-it.de)|217.160.0.108|:80... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 22244914 (21M) [text/csv]\nSaving to: ‘small_btc.csv.1’\n\nsmall_btc.csv.1 100%[===================>] 21.21M 13.4MB/s in 1.6s \n\n2021-10-13 10:44:12 (13.4 MB/s) - ‘small_btc.csv.1’ saved [22244914/22244914]\n\n" ] ], [ [ "## Translate to German", "_____no_output_____" ] ], [ [ "translate_pipe = nlu.load('en.translate_to.de')\ntranslate_pipe.predict(df)", "translate_en_de download started this may take some time.\nApprox size to download 268 MB\n[OK!]\n" ] ], [ [ "## Translate to Chinese", "_____no_output_____" ] ], [ [ "translate_pipe = nlu.load('en.translate_to.zh')\ntranslate_pipe.predict(df)", "translate_en_zh download started this may take some time.\nApprox size to download 280.9 MB\n[OK!]\n" ] ], [ [ "## Translate to Hindi", "_____no_output_____" ] ], [ [ "translate_pipe = nlu.load('en.translate_to.hi')\ntranslate_pipe.predict(df)", "translate_en_hi download started this may take some time.\nApprox size to download 275.1 MB\n[OK!]\n" ] ], [ [ "# Train a Multi Lingual Classifier for 100+ languages from a dataset with just one language\n\n[Leverage Language-agnostic BERT Sentence Embedding (LABSE) and acheive state of the art!](https://arxiv.org/abs/2007.01852) \n\nTraining a classifier with LABSE embeddings enables the knowledge to be transferred to 109 languages!\nWith the [SentimentDL model](https://nlp.johnsnowlabs.com/docs/en/annotators#sentimentdl-multi-class-sentiment-analysis-annotator) from Spark NLP you can achieve State Of the Art results on any binary class text classification problem.\n\n### Languages suppoted by LABSE\n![labse languages](http://ckl-it.de/wp-content/uploads/2021/02/LABSE.png)\n\n", "_____no_output_____" ], [ "**Make sure to restart your notebook again** before starting the next section", "_____no_output_____" ] ], [ [ "print(\"Please restart kernel if you are in google colab and run next cell after the restart to configure java 8 back\")\n1+'wait'\n", "_____no_output_____" ], [ "import nlu\n# Download French twitter Sentiment dataset https://www.kaggle.com/hbaflast/french-twitter-sentiment-analysis\n! wget http://ckl-it.de/wp-content/uploads/2021/02/french_tweets.csv\n\nimport pandas as pd\n\ntrain_path = '/content/french_tweets.csv'\n\ntrain_df = pd.read_csv(train_path)\n# the text data to use for classification should be in a column named 'text'\ncolumns=['text','y']\ntrain_df = train_df[columns]\ntrain_df = train_df.sample(frac=1).reset_index(drop=True)\ntrain_df", "--2021-10-13 10:56:02-- http://ckl-it.de/wp-content/uploads/2021/02/french_tweets.csv\nResolving ckl-it.de (ckl-it.de)... 217.160.0.108, 2001:8d8:100f:f000::209\nConnecting to ckl-it.de (ckl-it.de)|217.160.0.108|:80... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 10237264 (9.8M) [text/csv]\nSaving to: ‘french_tweets.csv’\n\nfrench_tweets.csv 100%[===================>] 9.76M 8.46MB/s in 1.2s \n\n2021-10-13 10:56:03 (8.46 MB/s) - ‘french_tweets.csv’ saved [10237264/10237264]\n\n" ] ], [ [ "## Train Deep Learning Classifier using `nlu.load('train.sentiment')`\n\nAl you need is a Pandas Dataframe with a label column named `y` and the column with text data should be named `text`\n\nWe are training on a french dataset and can then predict classes correct **in 100+ langauges**", "_____no_output_____" ] ], [ [ "from sklearn.metrics import classification_report\n# Train longer!\ntrainable_pipe = nlu.load('xx.embed_sentence.labse train.sentiment')\ntrainable_pipe['sentiment_dl'].setMaxEpochs(60) \ntrainable_pipe['sentiment_dl'].setLr(0.005) \nfitted_pipe = trainable_pipe.fit(train_df.iloc[:2000])\n# predict with the trainable pipeline on dataset and get predictions\npreds = fitted_pipe.predict(train_df.iloc[:2000],output_level='document')\n\n#sentence detector that is part of the pipe generates sone NaNs. lets drop them first\npreds.dropna(inplace=True)\nprint(classification_report(preds['y'], preds['trained_sentiment']))\n\npreds", " precision recall f1-score support\n\n negative 0.86 0.86 0.86 938\n neutral 0.00 0.00 0.00 0\n positive 0.88 0.87 0.87 1062\n\n accuracy 0.86 2000\n macro avg 0.58 0.58 0.58 2000\nweighted avg 0.87 0.86 0.87 2000\n\n" ], [ "", "_____no_output_____" ] ], [ [ "### Test the fitted pipe on new example", "_____no_output_____" ], [ "#### The Model understands Englsih\n![en](https://www.worldometers.info/img/flags/small/tn_nz-flag.gif)", "_____no_output_____" ] ], [ [ "fitted_pipe.predict(\"This was awful!\")", "_____no_output_____" ], [ "fitted_pipe.predict(\"This was great!\")", "_____no_output_____" ] ], [ [ "#### The Model understands German\n![de](https://www.worldometers.info/img/flags/small/tn_gm-flag.gif)", "_____no_output_____" ] ], [ [ "# German for:' this movie was great!'\nfitted_pipe.predict(\"Der Film war echt klasse!\")", "_____no_output_____" ], [ "# German for: 'This movie was really boring'\nfitted_pipe.predict(\"Der Film war echt langweilig!\")", "_____no_output_____" ] ], [ [ "#### The Model understands Chinese\n![zh](https://www.worldometers.info/img/flags/small/tn_ch-flag.gif)", "_____no_output_____" ] ], [ [ "# Chinese for: \"This model was awful!\"\nfitted_pipe.predict(\"这部电影太糟糕了！\")", "_____no_output_____" ], [ "# Chine for : \"This move was great!\"\nfitted_pipe.predict(\"此举很棒！\")\n", "_____no_output_____" ] ], [ [ "#### Model understanda Afrikaans\n\n![af](https://www.worldometers.info/img/flags/small/tn_sf-flag.gif)\n\n", "_____no_output_____" ] ], [ [ "# Afrikaans for 'This movie was amazing!'\nfitted_pipe.predict(\"Hierdie film was ongelooflik!\")\n", "_____no_output_____" ], [ "# Afrikaans for :'The movie made me fall asleep, it's awful!'\nfitted_pipe.predict('Die film het my aan die slaap laat raak, dit is verskriklik!')", "_____no_output_____" ] ], [ [ "#### The model understands Vietnamese\n![vi](https://www.worldometers.info/img/flags/small/tn_vm-flag.gif)", "_____no_output_____" ] ], [ [ "# Vietnamese for : 'The movie was painful to watch'\nfitted_pipe.predict('Phim đau điếng người xem')\n", "_____no_output_____" ], [ "\n# Vietnamese for : 'This was the best movie ever'\nfitted_pipe.predict('Đây là bộ phim hay nhất từ trước đến nay')", "_____no_output_____" ] ], [ [ "#### The model understands Japanese\n![ja](https://www.worldometers.info/img/flags/small/tn_ja-flag.gif)\n", "_____no_output_____" ] ], [ [ "\n# Japanese for : 'This is now my favorite movie!'\nfitted_pipe.predict('これが私のお気に入りの映画です！')", "_____no_output_____" ], [ "\n# Japanese for : 'I would rather kill myself than watch that movie again'\nfitted_pipe.predict('その映画をもう一度見るよりも自殺したい')", "_____no_output_____" ] ], [ [ "# There are many more models you can put to use in 1 line of code!\n## Checkout [the Modelshub](https://nlp.johnsnowlabs.com/models) and the [NLU Namespace](https://nlu.johnsnowlabs.com/docs/en/spellbook) for more models\n\n### NLU Webinars and Video Tutorials\n- [NLU & Streamlit Tutorial](https://vimeo.com/579508034#)\n- [Crash course of the 50 + Medical Domains and the 200+ Healtchare models in NLU](https://www.youtube.com/watch?v=gGDsZXt1SF8)\n- [Multi Lingual NLU Webinar - Tutorial on Chinese News dataset](https://www.youtube.com/watch?v=ftAOqJuxnV4)\n- [John Snow Labs NLU: Become a Data Science Superhero with One Line of Python code](https://events.johnsnowlabs.com/john-snow-labs-nlu-become-a-data-science-superhero-with-one-line-of-python-code?hsCtaTracking=c659363c-2188-4c86-945f-5cfb7b42fcfc%7C8b2b188b-92a3-48ba-ad7e-073b384425b0)\n- [Python Web Def Conf - Python's NLU library: 1,000+ Models, 200+ Languages, State of the Art Accuracy, 1 Line of Code](https://2021.pythonwebconf.com/presentations/john-snow-labs-nlu-the-simplicity-of-python-the-power-of-spark-nlp)\n- [NYC/DC NLP Meetup with NLU](https://youtu.be/hJR9m3NYnwk?t=2155)\n\n### More ressources \n- [Join our Slack](https://join.slack.com/t/spark-nlp/shared_invite/zt-lutct9gm-kuUazcyFKhuGY3_0AMkxqA)\n- [NLU Website](https://nlu.johnsnowlabs.com/)\n- [NLU Github](https://github.com/JohnSnowLabs/nlu)\n- [Many more NLU example tutorials](https://github.com/JohnSnowLabs/nlu/tree/master/examples)\n- [Overview of every powerful nlu 1-liner](https://nlu.johnsnowlabs.com/docs/en/examples)\n- [Checkout the Modelshub for an overview of all models](https://nlp.johnsnowlabs.com/models) \n- [Checkout the NLU Namespace where you can find every model as a tabel](https://nlu.johnsnowlabs.com/docs/en/spellbook)\n- [Intro to NLU article](https://medium.com/spark-nlp/1-line-of-code-350-nlp-models-with-john-snow-labs-nlu-in-python-2f1c55bba619)\n- [Indepth and easy Sentence Similarity Tutorial, with StackOverflow Questions using BERTology embeddings](https://medium.com/spark-nlp/easy-sentence-similarity-with-bert-sentence-embeddings-using-john-snow-labs-nlu-ea078deb6ebf)\n- [1 line of Python code for BERT, ALBERT, ELMO, ELECTRA, XLNET, GLOVE, Part of Speech with NLU and t-SNE](https://medium.com/spark-nlp/1-line-of-code-for-bert-albert-elmo-electra-xlnet-glove-part-of-speech-with-nlu-and-t-sne-9ebcd5379cd)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ] ]

ec7b43ac89092120c47e51cca8e3ecbaf1bce4a9

192,395

ipynb

Jupyter Notebook

Project.ipynb

abdullahmuaad9/arabic-classification-dl-ml

0419b82a55d7361d4e8367ef26e7fc28e2e74b7a

[ "MIT" ]

1

2020-05-20T14:08:35.000Z

Project.ipynb

malikziq/arabic-classification-dl-ml

0419b82a55d7361d4e8367ef26e7fc28e2e74b7a

[ "MIT" ]

null

Project.ipynb

malikziq/arabic-classification-dl-ml

0419b82a55d7361d4e8367ef26e7fc28e2e74b7a

[ "MIT" ]

1

2020-08-01T05:54:24.000Z

102.995182

19,798

0.799293

[ [ [ "<a href=\"https://colab.research.google.com/github/malikziq/arabic-classification-dl-ml/blob/master/Project.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>", "_____no_output_____" ], [ "# Connect to Drive & Libraries", "_____no_output_____" ] ], [ [ "import os\nimport re\nimport sys\nimport string\nimport math\nfrom decimal import Decimal\n\nimport numpy as np \nnp.random.seed(32)\nnp.set_printoptions(threshold=sys.maxsize)\n\nimport pandas as pd \nimport matplotlib.pyplot as plt\npd.set_option('display.max_colwidth',1000)\n\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import roc_auc_score\nfrom sklearn.manifold import TSNE\nfrom sklearn.preprocessing import LabelBinarizer, LabelEncoder\n\nfrom keras.preprocessing.text import Tokenizer\nfrom keras.preprocessing.sequence import pad_sequences\nfrom keras.layers import LSTM, Conv1D, MaxPooling1D, Dropout\nfrom keras.utils.np_utils import to_categorical\nfrom keras.callbacks import EarlyStopping\n\n# use natural language toolkit\nimport nltk\nfrom nltk.stem.isri import ISRIStemmer\n\n# Use GPU\nimport tensorflow as tf\ntf.test.gpu_device_name()\n\n# Connect to Drive\nfrom google.colab import drive\ndrive.mount('/content/gdrive')", "Using TensorFlow backend.\n" ] ], [ [ "# Read Data\n", "_____no_output_____" ] ], [ [ "# The data set path in drive\ndata_path = '/content/gdrive/My Drive/Colab Notebooks/IR_NLP/Project/arabic_dataset_classifiction.csv'\n# Read Data\ndata = pd.read_csv(data_path)\ndata.targe.value_counts()", "_____no_output_____" ], [ "# Drop empty records & rest Indexes \ndata=data.dropna() # drop NaN\ndata=data.reset_index(drop=True)\n\n# Get equal amount from each class. df.index.isin(['one','two'])]\ndata = data.groupby(['targe']).head(10000)\n#data = data.loc[data['targe'].isin([1, 3])].groupby(['targe']).head(10000)\n\ndata.targe.value_counts()", "_____no_output_____" ] ], [ [ "# Deep Learning", "_____no_output_____" ], [ "## Preprocessing \n\nIn the stage of preprocessing the data runs through:\n\n* Splitting into 80% traning 20% testing.\n* Tokanized, filtered from punctuations. The out put of the tokanizer is a one-hot encoded vector with the index of the words in the text.\n* Padding the text data (x train and test).\n* Encode y data labels to one-hot encoded represntaion.", "_____no_output_____" ] ], [ [ "# Split Data \ntrain_text, test_text, train_y, test_y = train_test_split(data['text'],data['targe'],test_size = 0.2, random_state=5)", "_____no_output_____" ], [ "MAX_NB_WORDS = 20000\n\n# get the raw text data\ntexts_train = train_text.astype(str)\ntexts_test = test_text.astype(str)\n\n# finally, vectorize the text samples into a 2D integer tensor\ntokenizer = Tokenizer(nb_words=MAX_NB_WORDS,\n char_level=False,\n filters='”،,.\":!\"~{#$%&()*+,-./:;<=>?@[\\\\]^_`{|}~\\t\\n',)\ntokenizer.fit_on_texts(texts_train)\nsequences = tokenizer.texts_to_sequences(texts_train)\nsequences_test = tokenizer.texts_to_sequences(texts_test)\n\nword_index = tokenizer.word_index\nprint('Found %s unique tokens.' % len(word_index))", "/usr/local/lib/python3.6/dist-packages/keras_preprocessing/text.py:178: UserWarning: The `nb_words` argument in `Tokenizer` has been renamed `num_words`.\n warnings.warn('The `nb_words` argument in `Tokenizer` '\n" ], [ "# Exmple on Tokanizer output:\nprint(tokenizer.texts_to_sequences(['من الذين على الأدبية يحل الشاعر']))", "[[2, 80, 3, 7919, 2942, 4076]]\n" ], [ "type(tokenizer.word_index), len(tokenizer.word_index)", "_____no_output_____" ], [ "index_to_word = dict((i, w) for w, i in tokenizer.word_index.items())", "_____no_output_____" ], [ "print(\" \".join([index_to_word[i] for i in sequences[0]]))\nprint(sequences[0])", "صغر السن يشك ِّ في الغالب عائقا في اندماج اللاعبين في أي بطولة خصوصا الأوروبية منها بيد أن أي لاعب لها القدرة على تجاوز هذه الصعوبات التي في بداياته وهو الأمر الحاصل للدولي المغربي سفيان أمرابط المنتقل حديثا إلى الدوري الهولندي بطل في الموسم الماضي شقيق الدولي نور الدين أمرابط لاعب واتفورد الإنجليزي سار بثبات نحو هدفه في الانتقال إلى فريق عريق مثل الذي انتقل إليه أخيرا التي لم تتجاوز بعد سن العشرين حتى يصنع لنفسه مسيرة عن الارتباط باسم شقيقه الذي في الدوريات الأوروبية وحمل قميص أندية مرجعية في تاريخ كرة القدم ِّ بذلك على مؤهلاته التي في أكاديمية نادي الهولندي هناك اللعبة وعمل بها إلى غاية الفريق الأول وقبلها في أحد فرق الهواة في مدينة الهولندية في سن السابعة عشرة قبل دون نقاش دعوة الناخب الوطني عبد الله الإدريسي مدرب المنتخب الوطني لأقل من سنة للمشاركة في تصفيات كأس العالم في الإمارات إلى جانب أسماء واعدة ِّ لم منها في منتخب الكبار إلا هو على بطولة جيدة كان فيها بشكل إيجابي في بلوغ دور الثمانية بعد احتلال الصدارة في مجموعته التي ضم َّ كلا من وكان من أهم الجديدة للمغرب لكل نصيب أمرابط من ما في البطولة العالمية هو ثقة فريقه الأم الذي قر َّ الارتباط به بشكل احترافي يمتد منذ ذلك الحين إلى غاية نهاية السنة الجارية قاطعا بذلك الطريق على العديد من الفرق الأوروبية الراغبة في ضم ِّ إلى صفوفها آنذاك وفتح له الباب الذي منذ اختياره احتراف كرة القدم المغربية الأصيلة بحكم أن عائلته الصغيرة من قبيلة آيت في إقليم الناظور قادته دائما العلاقة بالمغرب والمنتخب الوطني وما زاد من ذلك هو المناداة على أخيه الأكبر لتمثيل الأسود في المنافسات الدولية سنة عليه ذلك أن يعمل بجد ُّ بدوره بدعوة للعب مع الكبار وهو الحلم الذي مع قدوم الناخب الوطني هيرفي رونار الذي وج َّ له الدعوة في مارس الماضي في ودية مع المنتخب التونسي وانتهت بفوز المغاربة بهدف نظيف جديد قد َّ أوراق اعتماده بشكل لافت عندما استطاع سد فراغ وسط الميدان وأكد بالملموس على في مجاورة أخيه وبعد ودية النسور اعتبر المتتبعون للشأن الكروي المغربي أن أمرابط الصغير له القدرة على التأقلم مع وضع الفريق الوطني بحكم مؤهلاته التقنية العالية سنه ومن شأن ذلك أن يكون إيجابيا اللاعب والمنتخب إن استطاع الحفاظ على النسق نفسه وتطوير مستواه إلى الأفضل مع خبرة المباريات الدولية وهو ما يستجيب له حاليا إذ وشارك أساسيا في المباراة الأولى عن تصفيات كأس أمم إفريقيا أمام الكاميرون وقد َّ ما عليه وظل ِّ تعليمات المدرب حتى النهاية التي شهدت هزيمة بهدف نظيف من قلب العاصمة تنتظره مهم صعبة بالفعل نظرا مستوى اللاعبين وعدم تمك ُّ من فرض الذات داخل عرين انتظر المغاربة منذ البداية ُّ متحف الجامعة الملكية لكرة القدم على بطولة إفريقية واحدة جيل السبعينات الأجيال منذ ذلك الحين إلى اليوم فيه وفي باقي من يحمل قميص المغرب أن يكون المستقبل سار َّ لمزيد من أخبار الرياضة والرياضي ين زوروا\n[15895, 4762, 17547, 5286, 1, 11532, 8854, 1, 13939, 75, 1, 60, 315, 263, 1496, 127, 2233, 4, 60, 193, 91, 2908, 3, 1155, 16, 4778, 6, 1, 14898, 54, 99, 4040, 16720, 36, 3143, 6232, 10125, 4154, 5, 280, 1270, 1444, 1, 64, 31, 6095, 85, 1050, 143, 6232, 193, 15305, 905, 10841, 18730, 267, 5351, 1, 1664, 5, 102, 15175, 253, 7, 1757, 219, 480, 6, 22, 1938, 11, 2222, 5114, 148, 17548, 7926, 2267, 8, 6825, 844, 4610, 7, 1, 4604, 1496, 6020, 2209, 729, 8855, 1, 455, 199, 44, 5286, 586, 3, 12790, 6, 1, 4047, 435, 1270, 125, 3605, 7909, 59, 5, 724, 17, 52, 17641, 1, 138, 622, 1299, 1, 249, 5124, 1, 2222, 1214, 804, 15, 84, 3674, 2052, 1228, 25, 34, 155, 2118, 211, 42, 25, 4862, 2, 43, 1022, 1, 1766, 92, 98, 1, 1791, 5, 222, 1129, 10428, 5286, 22, 127, 1, 283, 2645, 118, 53, 3, 315, 679, 23, 62, 103, 4002, 1, 2578, 320, 3375, 11, 4207, 2642, 1, 8230, 6, 2559, 2518, 2654, 2, 139, 2, 1160, 194, 493, 456, 5376, 6232, 2, 10, 1, 112, 511, 53, 2446, 326, 2592, 7, 13134, 2518, 6825, 76, 103, 7744, 2914, 66, 37, 5879, 5, 724, 133, 149, 1021, 14635, 586, 621, 3, 158, 2, 457, 1496, 10519, 1, 2559, 5286, 5, 14439, 3700, 3121, 46, 2025, 7, 66, 5352, 18103, 199, 44, 21, 13345, 852, 4, 3180, 2768, 2, 16635, 2129, 1, 1661, 1882, 12503, 888, 1785, 198, 4133, 25, 536, 4240, 2, 37, 53, 8347, 3, 11252, 2784, 10488, 866, 1, 1968, 294, 43, 82, 37, 4, 1542, 12462, 10011, 1307, 7053, 3645, 9, 2645, 54, 8142, 7, 9, 10244, 1228, 25, 1695, 1217, 7, 11842, 2518, 46, 1692, 1, 523, 31, 1, 5403, 9, 42, 887, 4284, 2244, 172, 296, 8419, 295, 94, 2518, 5237, 10211, 103, 7408, 284, 2980, 5944, 7869, 372, 1139, 210, 16913, 3, 1, 12103, 11252, 348, 5403, 16636, 789, 8647, 9339, 1800, 36, 4, 6232, 4751, 46, 2908, 3, 8648, 9, 282, 17, 25, 852, 12790, 607, 4671, 11451, 95, 1866, 37, 4, 186, 6628, 154, 4133, 38, 2980, 1399, 3, 19596, 114, 3306, 4174, 5, 2630, 9, 3003, 255, 294, 54, 10, 7409, 46, 376, 27, 3416, 2740, 1, 45, 70, 8, 1766, 92, 1369, 120, 32, 6559, 311, 2518, 10, 82, 3899, 5286, 3175, 161, 148, 1571, 6, 1521, 3122, 296, 8419, 2, 2937, 749, 11139, 2173, 1197, 2969, 1099, 130, 75, 752, 9483, 10011, 2, 1704, 6185, 89, 18855, 12840, 172, 66, 1277, 10011, 9505, 110, 214, 72, 44, 3, 315, 3452, 437, 4434, 9755, 6826, 66, 37, 5879, 5, 40, 71, 48, 484, 2, 1304, 2209, 14, 4, 186, 1852, 10841, 2518, 1007, 2, 602, 468, 1418, 1178, 1066]\n" ] ], [ [ "### Data overview & Zipf's Law", "_____no_output_____" ] ], [ [ "# Data set records overviwe\nseq_lens = [len(s) for s in sequences]\nprint(\"Average length: %0.1f\" % np.mean(seq_lens))\nprint(\"Max length: %d\" % max(seq_lens))", "Average length: 228.1\nMax length: 4308\n" ], [ "# Zipf's Law\nimport math\n\nwords = []\nwords_freq_log = []\nwords_freq = []\nfor key in tokenizer.word_index:\n words.append(key)\n words_freq.append(tokenizer.word_counts[key])\n words_freq_log.append(math.log10(tokenizer.word_counts[key]))\n\nwords = np.array(words)\nwords_freq = np.array(words_freq)\nwords_freq_log = np.array(words_freq_log)", "_____no_output_____" ], [ "print(words[:10])\nprint(words_freq[:10])", "['في' 'من' 'على' 'أن' 'إلى' 'التي' 'الذي' 'عن' 'مع' 'ما']\n[739946 577150 344521 331489 291319 179210 138086 128015 104664 93524]\n" ], [ "\nplt.plot(words[:20], words_freq_log[:20])\nplt.xlabel('Words')\nplt.ylabel('Freq (log)')\nplt.show()", "_____no_output_____" ] ], [ [ "### Get Data ready for the models", "_____no_output_____" ] ], [ [ "# pad vectors to maximum length\nMAX_SEQUENCE_LENGTH = 300\n\n# pad sequences with 0s\nx_train = pad_sequences(sequences, maxlen=MAX_SEQUENCE_LENGTH)\nx_test = pad_sequences(sequences_test, maxlen=MAX_SEQUENCE_LENGTH)\nprint('Shape of data tensor:', x_train.shape)\nprint('Shape of data test tensor:', x_test.shape)", "Shape of data tensor: (87031, 300)\nShape of data test tensor: (21758, 300)\n" ], [ "# encode y data labels\nencoder = LabelEncoder()\nencoder.fit(train_y)\ny_train = encoder.transform(train_y)\ny_test = encoder.transform(test_y)", "_____no_output_____" ], [ "# Converts the labels to a one-hot representation\nN_CLASSES = np.max(y_train) + 1\ny_train = to_categorical(y_train, N_CLASSES)\ny_test = to_categorical(y_test, N_CLASSES)\nprint('Shape of label tensor:', y_train.shape)", "Shape of label tensor: (87031, 5)\n" ] ], [ [ "## Deep Nenral Network", "_____no_output_____" ] ], [ [ "from keras.layers import Dense, Input, Flatten\nfrom keras.layers import GlobalAveragePooling1D, Embedding\nfrom keras.models import Model\n\nEMBEDDING_DIM = 50\n\n# input: a sequence of MAX_SEQUENCE_LENGTH integers\nsequence_input = Input(shape=(MAX_SEQUENCE_LENGTH,), dtype='int32')\n\nembedding_layer = Embedding(MAX_NB_WORDS, EMBEDDING_DIM,\n input_length=MAX_SEQUENCE_LENGTH,\n trainable=True)\n\nembedded_sequences = embedding_layer(sequence_input)\n\naverage = GlobalAveragePooling1D()(embedded_sequences)\npredictions = Dense(N_CLASSES, activation='softmax')(average)\n\nmodel = Model(sequence_input, predictions)\nmodel.compile(loss='categorical_crossentropy',\n optimizer='adam', metrics=['acc'])", "_____no_output_____" ], [ "model.summary()", "Model: \"model_6\"\n_________________________________________________________________\nLayer (type) Output Shape Param # \n=================================================================\ninput_6 (InputLayer) (None, 300) 0 \n_________________________________________________________________\nembedding_4 (Embedding) (None, 300, 50) 1000000 \n_________________________________________________________________\nglobal_average_pooling1d_4 ( (None, 50) 0 \n_________________________________________________________________\ndense_6 (Dense) (None, 5) 255 \n=================================================================\nTotal params: 1,000,255\nTrainable params: 1,000,255\nNon-trainable params: 0\n_________________________________________________________________\n" ], [ "hist = model.fit(x_train, y_train, validation_split=0.2, epochs=15, batch_size=128)", "/usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/indexed_slices.py:434: UserWarning: Converting sparse IndexedSlices to a dense Tensor of unknown shape. This may consume a large amount of memory.\n \"Converting sparse IndexedSlices to a dense Tensor of unknown shape. \"\n" ], [ "f, ax = plt.subplots()\nax.plot([None] + hist.history['acc'])\nax.plot([None] + hist.history['val_acc'])\nax.legend(['Train accuracy', 'Validation accuracy'], loc=0)\n#ax.set_title('Training accuracy per Epoch')\nax.set_xlabel('Epoch')\nax.set_ylabel('Accuracy (X100%)')\nplt.show()", "_____no_output_____" ], [ "f, ax = plt.subplots()\nax.plot([None] + hist.history['val_acc'])\nax.set_title('Validation accuracy per Epoch')\nax.set_xlabel('Epoch')\nax.set_ylabel('Accuracy (X100%) ')\nplt.show()", "_____no_output_____" ], [ "f, ax = plt.subplots()\nax.plot([None] + hist.history['loss'])\n#ax.plot([None] + hist.history['val_loss'])\n#ax.legend(['Train Loss', 'Validation Loss'], loc=0)\nax.set_title('Training Loss per Epoch')\nax.set_xlabel('Epoch')\nax.set_ylabel('Loss')\nplt.show()", "_____no_output_____" ], [ "output_test = model.predict(x_test)\nprint(\"test auc:\", roc_auc_score(y_test,output_test))", "test auc: 0.9945050063723798\n" ], [ "# Here's how to generate a prediction on individual examples\ntext_labels = encoder.classes_ \n\nfor i in range(50,80):\n prediction = model.predict(np.array([x_test[i]]))\n predicted_label = text_labels[np.argmax(prediction)]\n print(texts_test.iloc[i], \"...\")\n print('Actual label:' + str(test_y.iloc[i]))\n print(\"Predicted label: \" + str(predicted_label) + \"\\n\") ", "_____no_output_____" ] ], [ [ "## LSTM", "_____no_output_____" ], [ "### LSTM 1", "_____no_output_____" ] ], [ [ "# input: a sequence of MAX_SEQUENCE_LENGTH integers\nsequence_input = Input(shape=(MAX_SEQUENCE_LENGTH,), dtype='int32')\nembedded_sequences = embedding_layer(sequence_input)\n\nx = LSTM(128, dropout=0.5, recurrent_dropout=0.2)(embedded_sequences)\npredictions = Dense(N_CLASSES, activation='softmax')(x)\n\n\nmodel = Model(sequence_input, predictions)\nmodel.compile(loss='categorical_crossentropy',\n optimizer='adam',\n metrics=['acc'])", "_____no_output_____" ], [ "model.fit(x_train, y_train, validation_split=0.1,\n nb_epoch=3, batch_size=128)", "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:2: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n \n" ], [ "output_test = model.predict(x_test)\nprint(\"test auc:\", roc_auc_score(y_test,output_test))", "test auc: 0.990678454101022\n" ], [ "# Evaluate the accuracy of our trained model\nscore = model.evaluate(x_test, y_test,\n batch_size=64, verbose=1)\nprint('Test loss:', score[0])\nprint('Test accuracy:', score[1])", "22346/22346 [==============================] - 14s 645us/step\nTest loss: 0.1953141155583434\nTest accuracy: 0.9459858536720276\n" ] ], [ [ "### LSTM 2", "_____no_output_____" ] ], [ [ "# input: a sequence of MAX_SEQUENCE_LENGTH integers\nsequence_input = Input(shape=(MAX_SEQUENCE_LENGTH,), dtype='int32')\nembedded_sequences = embedding_layer(sequence_input)\n\n# 1D convolution with 64 output channels\nx = Conv1D(64, 5)(embedded_sequences)\n# MaxPool divides the length of the sequence by 5\nx = MaxPooling1D(5)(x)\nx = Dropout(0.5)(x)\nx = Conv1D(64, 5)(x)\nx = MaxPooling1D(5)(x)\n# LSTM layer with a hidden size of 64\nx = Dropout(0.3)(x)\nx = LSTM(32)(x)\npredictions = Dense(N_CLASSES, activation='softmax')(x)\n\nmodel = Model(sequence_input, predictions)\nmodel.compile(loss='categorical_crossentropy',\n optimizer='adam',metrics=['acc'])", "_____no_output_____" ], [ "model.summary()", "Model: \"model_5\"\n_________________________________________________________________\nLayer (type) Output Shape Param # \n=================================================================\ninput_5 (InputLayer) (None, 300) 0 \n_________________________________________________________________\nembedding_3 (Embedding) (None, 300, 50) 1000000 \n_________________________________________________________________\nconv1d_3 (Conv1D) (None, 296, 64) 16064 \n_________________________________________________________________\nmax_pooling1d_3 (MaxPooling1 (None, 59, 64) 0 \n_________________________________________________________________\ndropout_3 (Dropout) (None, 59, 64) 0 \n_________________________________________________________________\nconv1d_4 (Conv1D) (None, 55, 64) 20544 \n_________________________________________________________________\nmax_pooling1d_4 (MaxPooling1 (None, 11, 64) 0 \n_________________________________________________________________\ndropout_4 (Dropout) (None, 11, 64) 0 \n_________________________________________________________________\nlstm_2 (LSTM) (None, 32) 12416 \n_________________________________________________________________\ndense_5 (Dense) (None, 5) 165 \n=================================================================\nTotal params: 1,049,189\nTrainable params: 1,049,189\nNon-trainable params: 0\n_________________________________________________________________\n" ], [ "hist = model.fit(x_train, y_train, validation_split=0.2, epochs=15, batch_size=128)", "/usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/indexed_slices.py:434: UserWarning: Converting sparse IndexedSlices to a dense Tensor of unknown shape. This may consume a large amount of memory.\n \"Converting sparse IndexedSlices to a dense Tensor of unknown shape. \"\n" ], [ "f, ax = plt.subplots()\nax.plot([None] + hist.history['acc'])\n#ax.plot([None] + hist.history['val_accuracy'])\n#ax.legend(['Train accuracy', 'Validation accuracy'], loc=0)\nax.set_title('Training accuracy per Epoch')\nax.set_xlabel('Epoch')\nax.set_ylabel('Accuracy (X100%)')\nplt.show()", "_____no_output_____" ], [ "f, ax = plt.subplots()\nax.plot([None] + hist.history['val_acc'])\nax.set_title('Validation accuracy per Epoch')\nax.set_xlabel('Epoch')\nax.set_ylabel('Accuracy (X100%) ')\nplt.show()", "_____no_output_____" ], [ "f, ax = plt.subplots()\nax.plot([None] + hist.history['loss'])\n#ax.plot([None] + hist.history['val_loss'])\n#ax.legend(['Train Loss', 'Validation Loss'], loc=0)\nax.set_title('Training Loss per Epoch')\nax.set_xlabel('Epoch')\nax.set_ylabel('Loss')\nplt.show()", "_____no_output_____" ], [ "output_test = model.predict(x_test)\nprint(\"test auc:\", roc_auc_score(y_test,output_test))", "test auc: 0.9933526571605558\n" ] ], [ [ "# Machine Learning\n", "_____no_output_____" ], [ "## Preprocessing", "_____no_output_____" ] ], [ [ "# Read Data\ndataset = data\ndataset.targe.value_counts()", "_____no_output_____" ], [ "# Extract Coulmns of data (x) and labes (y)\nx=dataset.iloc[:,0]\ny=np.array(dataset.iloc[:,1])", "_____no_output_____" ], [ "# Download Arabic Stop words\nnltk.download('stopwords')\narb_stopwords = set(nltk.corpus.stopwords.words(\"arabic\"))\nprint(arb_stopwords)", "[nltk_data] Downloading package stopwords to /root/nltk_data...\n[nltk_data] Package stopwords is already up-to-date!\n{'إليكم', 'منها', 'قد', 'هكذا', 'هن', 'لا', 'لئن', 'وإذا', 'ذواتي', 'هذان', 'هذين', 'بعض', 'ليستا', 'اللذان', 'إليكما', 'إذ', 'لن', 'بنا', 'فيها', 'فيما', 'ما', 'هي', 'ذا', 'حاشا', 'فمن', 'بها', 'مما', 'إذن', 'ثمة', 'منه', 'آه', 'كما', 'أما', 'عليه', 'أن', 'بكن', 'تين', 'عند', 'كليكما', 'وإذ', 'عليك', 'ذانك', 'اللذين', 'نحو', 'تلكم', 'أوه', 'والذين', 'هم', 'ولكن', 'ليت', 'هاته', 'لها', 'أو', 'هيا', 'ذاك', 'بلى', 'لي', 'نعم', 'هيت', 'لعل', 'إلى', 'أولئك', 'لكيلا', 'عما', 'ريث', 'أينما', 'اللتيا', 'لاسيما', 'اللاتي', 'حبذا', 'بخ', 'بكم', 'هنالك', 'إنما', 'سوف', 'لكي', 'لولا', 'إليك', 'تي', 'لستم', 'لكن', 'أكثر', 'به', 'مذ', 'لهم', 'بهن', 'بين', 'إي', 'هاتين', 'فإن', 'كلما', 'ولا', 'لو', 'فلا', 'حين', 'على', 'ألا', 'بمن', 'لست', 'اللائي', 'بس', 'خلا', 'ذه', 'أولاء', 'ومن', 'آي', 'لستن', 'في', 'أقل', 'له', 'بل', 'ذلكم', 'تلكما', 'أيها', 'لكم', 'إذا', 'لوما', 'لسنا', 'هؤلاء', 'أنى', 'ها', 'ذوا', 'لستما', 'هاتي', 'هلا', 'أين', 'ذلك', 'كذا', 'كل', 'لكما', 'ته', 'ذو', 'ممن', 'إنه', 'دون', 'حتى', 'هناك', 'هذا', 'أم', 'مهما', 'إيه', 'هذي', 'من', 'لنا', 'بكما', 'كأنما', 'أنتما', 'بما', 'إن', 'لدى', 'آها', 'هما', 'إليكن', 'لما', 'كأين', 'ذينك', 'لسن', 'ليسوا', 'لك', 'الذين', 'ذي', 'شتان', 'أنتن', 'فيم', 'ليسا', 'منذ', 'هنا', 'ثم', 'وهو', 'مع', 'حيث', 'كلا', 'كليهما', 'يا', 'كم', 'كيت', 'ذواتا', 'إلا', 'هذه', 'ذات', 'كذلك', 'كأن', 'الذي', 'متى', 'اللتان', 'إما', 'تينك', 'كي', 'هاهنا', 'أف', 'إذما', 'بماذا', 'أنت', 'أي', 'لكنما', 'هيهات', 'ليست', 'ذلكما', 'بهم', 'كيفما', 'هل', 'ذين', 'فإذا', 'كيف', 'ذلكن', 'بعد', 'بهما', 'اللواتي', 'أنتم', 'تلك', 'عل', 'وإن', 'لم', 'بي', 'مه', 'عسى', 'ماذا', 'هاتان', 'هاك', 'سوى', 'أنا', 'وما', 'التي', 'حيثما', 'ذان', 'عدا', 'عن', 'لهن', 'بك', 'والذي', 'كلتا', 'بيد', 'فيه', 'نحن', 'اللتين', 'غير', 'ولو', 'إنا', 'كلاهما', 'هو', 'كأي', 'لهما', 'ليس'}\n" ], [ "from sklearn.feature_extraction.text import CountVectorizer, TfidfTransformer\n\n# Remove stop words and stem words\nstemmer=ISRIStemmer()\nprint('before: ', x.iloc[0])\nx = x.apply(lambda cell : ' '. join([stemmer.stem(word) for word in cell.split() if word not in arb_stopwords]))\nprint('after: ', x.iloc[0])", "before: بين أستوديوهات ورزازات وصحراء مرزوكة وآثار وليلي ثم الرباط والبيضاء انتهى المخرج المغربي سهيل بن بركة من تصوير مشاهد عمله السينمائي الجديد الذي خصصه لتسليط الضوء عن حياة الجاسوس الإسباني دومينغو باديا الذي عاش فترة من القرن التاسع عشر بالمغرب باسم علي باي هذا الفيلم الذي اختار له مخرجه عنوان حلم خليفة يصور حياة علي باي العباسي الذي ما زال أحد أحياء طنجة يحمل اسمه عاش حياة فريدة متنكرا بشخصية تاجر عربي من سلالة الرسول صلى الله عليه وسلم فيما كان يعمل جاسوسا لحساب إسبانيا وكشف مخرج الفيلم سهيل بن بركة في تصريح لهسبريس أن الفيلم السينمائي دخل مرحلة التوضيب التي تتم خارج المغرب مبرزا أن الفيلم الذي يروي حياة الجاسوس الإسباني دومينغو باديا منذ أن قرر من طنجة بدء رحلاته نحو عدد من المناطق في العالم الإسلامي بداية القرن العشرين سيكون جاهزا بعد شهرين ويجمع الفيلم السينمائي عددا من الممثلين من مختلف الجنسيات واختار لدور البطولة الممثلة السينمائية الإيطالية كارولينا كريشنتيني للقيام بدور الإنجليزية الليدي هستر ستانهوب التي اشتهرت في الكتب الغربية بـ زنوبيا والتي عاشت بدورها بالدول العربية وارتبطت بعلي باي بعلاقة عاطفية إضافة إلى وجوه سينمائية معروفة وعن اختيار المخرج المغربي لحياة علي باي العباسي يوضح في تصريح لوكالة الأنباء الفرنسية هذه الشخصية عاشت أحداثا مشوقة كثيرة تستحق أن تسلط عليها الأضواء مشيرا إلى أن الفيلم سيحمل الكثير من المفاجآت لا سيما أن البطل قتل على يد امرأة دست له السم خلال رحلة الحج وأضاف شخصية طموحة وشجاعة ومثقفة ومذهلة في آن واحد كان يرى نفسه مستكشفا في أول الأمر نال علي باي إعجاب السلطان بعلمه فجعله من المقربين منه في ظرف وجيز ودعاه إلى اللحاق به إلى فاس وبرحيله إلى فاس تنتهي قصته مع طنجة وعاش علي باي العباسي بمدينة طنجة على أنه رجل مسلم أصله من الشام ونال ثقة الجميع في هذه المدينة حيث تم تشييد تمثال له في عروسة الشمال نظرا لتمكنه من بعض العلوم خاصة علم الفلك الذي مكنه من رصد كسوف الشمس الذي تزامن مع وجوده في طنجة فكان لعلمه دور كبير ساعده في إخفاء هويته كما أبان هذا الأمر أيضا عن تراجع كبير في ميدان العلم والمعرفة لدى المغاربة والمسلمين بصفة عامة\nafter: استوديوه رزز صحراء رزك آثر ولل ربط يضء نهى خرج غرب سهل بن برك صور شهد عمل نمئ جدد خصص سلط ضوء حية جسس اسب دومينغو باد عاش فتر قرن تسع عشر غرب بسم علي باي يلم خار خرج عنو حلم خلف يصر حية علي باي عبس زال احد حيء طنج حمل اسم عاش حية فرد تنكر شخص تجر عرب سلل رسل صلى الل سلم كان عمل جسس حسب سبن كشف خرج يلم سهل بن برك صرح هسبريس يلم نمئ دخل رحل وضب تتم خرج غرب برز يلم يري حية جسس اسب دومينغو باد قرر طنج بدء رحل عدد نطق علم سلم بدي قرن عشر سيك جهز شهر جمع يلم نمئ عدد مثل خلف جنس خار لدر بطل مثل سينمائية يطل كرل كرش قيم بدر انجليزية ليد هستر هوب شهر كتب غرب بـ زنب والتي عشت بدر دول عرب ربط بعل باي علق عطف ضفة وجه سينمائية عرف وعن خير خرج غرب لحا علي باي عبس وضح صرح وكل باء رنس شخص عشت حدث شوق كثر سحق سلط عليها ضوء شير يلم حمل كثر فجآ سيم بطل قتل يد مرأ دست لسم خلل رحل لحج أضف شخص طمح شجع ثقف ذهل ان وحد كان يرى نفس كشف اول امر نال علي باي عجب سلط علم جعل قرب ظرف وجز ودع لحق فاس رحل فاس نهي قصت طنج وعش علي باي عبس بمد طنج انه رجل سلم اصل شام ونل ثقة جمع دين تم شيد ثال عرس شمل نظر لتم علم خصة علم فلك كنه رصد كسف شمس زمن وجد طنج فكان علم دور كبر سعد خفء هوت ابن امر ايض رجع كبر ميد علم عرف غرب سلم بصف عمة\n" ], [ "# Calculate TF\ncount_vect=CountVectorizer(analyzer='word', max_features=5000)\nx_counts=count_vect.fit_transform(x)", "_____no_output_____" ], [ "print('TF: ', x_counts[0])\nprint('Total Features = ', len(count_vect.get_feature_names()), ' - x (TF matrix) shpae: ', x_counts.shape)", "TF: (0, 413)\t1\n (0, 1977)\t1\n (0, 2607)\t1\n (0, 0)\t1\n (0, 4705)\t1\n (0, 1924)\t2\n (0, 4889)\t1\n (0, 4276)\t1\n (0, 1599)\t5\n (0, 3071)\t6\n (0, 2393)\t2\n (0, 847)\t2\n (0, 726)\t2\n (0, 2677)\t1\n (0, 2556)\t1\n (0, 3003)\t2\n (0, 4256)\t3\n (0, 1253)\t1\n (0, 1616)\t1\n (0, 2353)\t3\n (0, 2738)\t2\n (0, 1557)\t4\n (0, 1292)\t3\n (0, 378)\t2\n (0, 2852)\t2\n :\t:\n (0, 2575)\t1\n (0, 1186)\t1\n (0, 2908)\t1\n (0, 2548)\t1\n (0, 4207)\t1\n (0, 3683)\t1\n (0, 1614)\t1\n (0, 3262)\t1\n (0, 3591)\t1\n (0, 1996)\t1\n (0, 2545)\t1\n (0, 2183)\t1\n (0, 4501)\t1\n (0, 3250)\t1\n (0, 1807)\t1\n (0, 3478)\t2\n (0, 2320)\t1\n (0, 1638)\t1\n (0, 4382)\t1\n (0, 279)\t1\n (0, 588)\t1\n (0, 1950)\t1\n (0, 4072)\t1\n (0, 768)\t1\n (0, 2997)\t1\nTotal Features = 5000 - x (TF matrix) shpae: (50000, 5000)\n" ], [ "# Calculate TF-IDF\ntfidf_transformer = TfidfTransformer()\nx_tfidf = tfidf_transformer.fit_transform(x_counts)\nx_tfidf = np.array(x_tfidf.toarray())", "_____no_output_____" ], [ "x_tfidf[0]", "_____no_output_____" ], [ "x_tfidf.shape", "_____no_output_____" ], [ "x_train, x_test, y_train, y_test = train_test_split(x_tfidf, y, test_size = 0.2, random_state = 5)", "_____no_output_____" ] ], [ [ "## SVM", "_____no_output_____" ] ], [ [ "from sklearn.svm import SVC\n\nclf= SVC(kernel = 'rbf', random_state = 0)\n# Traning\nclf.fit(x_train, y_train)\nclf.score(x_train, y_train)", "_____no_output_____" ], [ "# Testing\ny_pred=clf.predict(x_test)", "_____no_output_____" ], [ "from sklearn.metrics import confusion_matrix,accuracy_score\ncm = confusion_matrix(y_test, y_pred)\nAccuracy_Score = accuracy_score(y_test, y_pred)", "_____no_output_____" ], [ "import seaborn as sn\nprint('Accuracy = %', Accuracy_Score * 100)\n\nprint('Confustion Matirx: ')\nsn.heatmap(cm,annot=True,cmap='Blues', fmt='g')", "Accuracy = % 95.33\nConfustion Matirx: \n" ] ], [ [ "# Download DataSet\n\nData Set link: [https://data.mendeley.com/datasets/v524p5dhpj/2]", "_____no_output_____" ] ], [ [ "# Download data from the link\n!wget https://data.mendeley.com/datasets/v524p5dhpj/2/files/72c2e306-9538-4c74-a28f-558fbe87c382/arabic_dataset_classifiction.csv.zip", "--2020-05-18 22:57:50-- https://data.mendeley.com/datasets/v524p5dhpj/2/files/72c2e306-9538-4c74-a28f-558fbe87c382/arabic_dataset_classifiction.csv.zip\nResolving data.mendeley.com (data.mendeley.com)... 162.159.130.86, 162.159.133.86, 2606:4700:7::a29f:8556, ...\nConnecting to data.mendeley.com (data.mendeley.com)|162.159.130.86|:443... connected.\nHTTP request sent, awaiting response... 302 Found\nLocation: https://md-datasets-public-files-prod.s3.eu-west-1.amazonaws.com/b6de1cfa-56f4-4b51-b8b9-65923207b36d [following]\n--2020-05-18 22:57:51-- https://md-datasets-public-files-prod.s3.eu-west-1.amazonaws.com/b6de1cfa-56f4-4b51-b8b9-65923207b36d\nResolving md-datasets-public-files-prod.s3.eu-west-1.amazonaws.com (md-datasets-public-files-prod.s3.eu-west-1.amazonaws.com)... 52.218.56.120\nConnecting to md-datasets-public-files-prod.s3.eu-west-1.amazonaws.com (md-datasets-public-files-prod.s3.eu-west-1.amazonaws.com)|52.218.56.120|:443... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 82058965 (78M) [application/zip]\nSaving to: ‘arabic_dataset_classifiction.csv.zip’\n\narabic_dataset_clas 100%[===================>] 78.26M 17.2MB/s in 4.6s \n\n2020-05-18 22:57:56 (17.2 MB/s) - ‘arabic_dataset_classifiction.csv.zip’ saved [82058965/82058965]\n\n" ], [ "# Unzip the data file\n# (Recommanded) move it to your Driver after zipping\n!unzip /content/arabic_dataset_classifiction.csv.zip", "Archive: /content/arabic_dataset_classifiction.csv.zip\n inflating: arabic_dataset_classifiction.csv \n creating: __MACOSX/\n inflating: __MACOSX/._arabic_dataset_classifiction.csv \n" ], [ "", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown", "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code" ], [ "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ] ]

ec7b485159f9bf72eec9634785e3d5de961e7a00

23,240

ipynb

Jupyter Notebook

doc/notebooks/CP0.ipynb

romarro/CoolProp-1

c06d2df794d75766b769f5271b98ca3fc36b7f93

[ "MIT" ]

520

2015-01-14T16:49:41.000Z

2022-03-29T07:48:50.000Z

doc/notebooks/CP0.ipynb

romarro/CoolProp-1

c06d2df794d75766b769f5271b98ca3fc36b7f93

[ "MIT" ]

1,647

2015-01-01T07:42:45.000Z

2022-03-31T23:48:56.000Z

doc/notebooks/CP0.ipynb

romarro/CoolProp-1

c06d2df794d75766b769f5271b98ca3fc36b7f93

[ "MIT" ]

320

2015-01-02T17:24:27.000Z

2022-03-19T07:01:00.000Z

79.317406

4,147

0.716136

[ [ [ "empty" ] ] ]

[ "empty" ]

[ [ "empty" ] ]

ec7b4d2b182a38ebae1826ee298b875b622571ee

16,978

ipynb

Jupyter Notebook

lab2/lab2_part1.ipynb

gigaroby/data-intensive

eab1ff55a79cabda0cf2b1c5ee8fa5360b19488b

[ "CNRI-Python", "Apache-1.1" ]

null

lab2/lab2_part1.ipynb

gigaroby/data-intensive

eab1ff55a79cabda0cf2b1c5ee8fa5360b19488b

[ "CNRI-Python", "Apache-1.1" ]

null

lab2/lab2_part1.ipynb

gigaroby/data-intensive

eab1ff55a79cabda0cf2b1c5ee8fa5360b19488b

[ "CNRI-Python", "Apache-1.1" ]

null

30.869091

2,471

0.560372

[ [ [ "<p align=\"center\"><img src=\"logo/spark.png\" alt=\"Hadoop Logo\" width=\"250\"/></p>\n# **Lab 2 - Part 1 - Spark**\n#### The following steps demonstrate how to create a simple Spark application in Scala. In this notebook you will see how to make a base RDD and appy functions to it.\n\n", "_____no_output_____" ], [ "### ** Part 1: Warm Up **", "_____no_output_____" ], [ "Create a collection of integers in the range of 1 ... 10000.", "_____no_output_____" ] ], [ [ "// TODO: Replace <FILL IN> with appropriate code\nval data = 1 to 10000\nprint(data)", "Range(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254, 255, 256, 257, 258, 259, 260, 261, 262, 263, 264, 265, 266, 267, 268, 269, 270, 271, 272, 273, 274, 275, 276, 277, 278, 279, 280, 281, 282, 283, 284, 285, 286, 287, 288, 289, 290, 291, 292, 293, 294, 295, 296, 297, 298, 299, 300, 301, 302, 303, 304, 305, 306, 307, 308, 309, 310, 311, 312, 313, 314, 315, 316, 317, 318, 319, 320, 321, 322, 323, 324, 325, 326, 327, 328, 329, 330, 331, 332, 333, 334, 335, 336, 337, 338, 339, 340, 341, 342, 343, 344, 345, 346, 347, 348, 349, 350, 351, 352, 353, 354, 355, 356, 357, 358, 359, 360, 361, 362, 363, 364, 365, 366, 367, 368, 369, 370, 371, 372, 373, 374, 375, 376, 377, 378, 379, 380, 381, 382, 383, 384, 385, 386, 387, 388, 389, 390, 391, 392, 393, 394, 395, 396, 397, 398, 399, 400, 401, 402, 403, 404, 405, 406, 407, 408, 409, 410, 411, 412, 413, 414, 415, 416, 417, 418, 419, 420, 421, 422, 423, 424, 425, 426, 427, 428, 429, 430, 431, 432, 433, 434, 435, 436, 437, 438, 439, 440, 441, 442, 443, 444, 445, 446, 447, 448, 449, 450, 451, 452, 453, 454, 455, 456, 457, 458, 459, 460, 461, 462, 463, 464, 465, 466, 467, 468, 469, 470, 471, 472, 473, 474, 475, 476, 477, 478, 479, 480, 481, 482, 483, 484, 485, 486, 487, 488, 489, 490, 491, 492, 493, 494, 495, 496, 497, 498, 499, 500, 501, 502, 503, 504, 505, 506, 507, 508, 509, 510, 511, 512, ... )" ] ], [ [ "Use that collection to create a base RDD.", "_____no_output_____" ] ], [ [ "// TODO: Replace <FILL IN> with appropriate code\nval distData = sc.parallelize(data)\nprint(distData)", "ParallelCollectionRDD[6] at parallelize at <console>:29" ] ], [ [ "Namely a `filter()` transformation to keep the values less than 10, then a `collect()` action to collect the results.", "_____no_output_____" ] ], [ [ "val x = distData.filter(x => x < 10)\nx.collect()", "" ] ], [ [ "### ** Part 2: Create an RDD From a File **\n\nThe following steps demonstrate how to create an RDD from a file and apply transofrmations on it in Scala. Creat an RDD, named `pagecounts`, from the input files at `data/pagecounts`. The files entries will look something like this:\n```\n20090507-040000 zh favicon.ico 67 62955\n```", "_____no_output_____" ] ], [ [ "// TODO: Replace <FILL IN> with appropriate code\nval pagecounts = sc.textFile(\"data/pagecounts/\")\nprint(pagecounts)", "data/pagecounts/ MapPartitionsRDD[20] at textFile at <console>:24" ] ], [ [ "Use the `take()` operation of an RDD to get the first 10 records.", "_____no_output_____" ] ], [ [ "// TODO: Replace <FILL IN> with appropriate code\npagecounts.take(10)", "" ] ], [ [ "An alternative way to print the fields is to travers the array and print each record on its own line.", "_____no_output_____" ] ], [ [ "// TODO: Replace <FILL IN> with appropriate code\nfor (x <- pagecounts.take(10)) {\n println(x)\n}", "20090507-040000 aa Main_Page 7 51309\n20090507-040000 ab %D0%90%D0%B8%D0%BD%D1%82%D0%B5%D1%80%D0%BD%D0%B5%D1%82 1 34069\n20090507-040000 ab %D0%98%D1%85%D0%B0%D0%B4%D0%BE%D1%83_%D0%B0%D0%B4%D0%B0%D2%9F%D1%8C%D0%B0 3 65763\n20090507-040000 af.b Tuisblad 1 36231\n20090507-040000 af.d Tuisblad 1 58960\n20090507-040000 af.q Tuisblad 1 44265\n20090507-040000 af Afrikaans 3 80838\n20090507-040000 af Australi%C3%AB 1 132433\n20090507-040000 af Ensiklopedie 2 60584\n20090507-040000 af Internet 1 48816\n" ] ], [ [ "Use the `count()` function to see how many records in total are in this data set. The `pagecounts` folder consists of two files, each with around 700K lines, so in total we have around 1400K lines.", "_____no_output_____" ] ], [ [ "// TODO: Replace <FILL IN> with appropriate code\npagecounts.count()", "" ] ], [ [ "The second field of each record in the data set is the \"project code\" and contains information about the language of the pages. For example, the project code \"en\" indicates an English page. Let's derive an RDD, named `enPages`, containing only English pages from pagecounts. This can be done by applying a `filter()` function to pagecounts. For each record, we can split it by the field delimiter (i.e., a space) and get the second field, and then compare it with the string \"en\". To avoid reading from disks each time we perform any operations on the RDD, we can use `cache()` to cache the RDD into memory. ", "_____no_output_____" ] ], [ [ "// TODO: Replace <FILL IN> with appropriate code\nval enPages = pagecounts.filter(x => {\n var fields = x.split(\" \")\n fields(1) == \"en\"\n}).cache()", "" ] ], [ [ "The above command defines the RDD, but because of lazy evaluation, no computation is done yet. Next time any action is invoked on `enPages`, Spark will cache the data set in memory across the workers in your cluster. So, let's count the number of records, which are there for English pages. The first time this command is run, it will take 2-3 minutes while Spark scans through the entire data set on disk. But since `enPages` was marked as \"cached\" in the previous step, if you run count on the same RDD again, it should return an order of magnitude faster.", "_____no_output_____" ] ], [ [ "// TODO: Replace <FILL IN> with appropriate code\nenPages.count()", "" ] ], [ [ "Let's generate a histogram of total page views on Wikipedia English pages for the date range represented in our dataset (May 5 to May 7, 2009). The high level idea of what we'll be doing is as follows. First, we generate a key value pair for each line; the key is the date (the first eight characters of the first field), and the value is the number of pageviews for that date (the fourth field).", "_____no_output_____" ] ], [ [ "// TODO: Replace <FILL IN> with appropriate code\nval enKeyValuePairs = enPages.map(x => {\n var fields = x.split(\" \")\n var date = (fields(0).split(\"-\"))(0)\n (date, fields(3).toInt)\n})", "" ] ], [ [ "Next, we shuffle the data and group all values of the same key together. Finally we sum up the values for each key. There is a convenient method called `reduceByKey` in Spark for exactly this pattern. Note that the second argument to `reduceByKey` determines the number of reducers to use. By default, Spark assumes that the reduce function is commutative and associative and applies combiners on the mapper side. Since we know there is a very limited number of keys in this case (because there are only 3 unique dates in our data set), let’s use only one reducer.", "_____no_output_____" ] ], [ [ "// TODO: Replace <FILL IN> with appropriate code\nenKeyValuePairs.reduceByKey(_ + _, 1).collect()", "" ] ], [ [ "The `collect()` method at the end converts the result from an RDD to an array. We can combine the previous three commands into one:", "_____no_output_____" ] ], [ [ "// TODO: Replace <FILL IN> with appropriate code\nenPages.map(x => x.split(\" \")).map(x => (x(0).split(\"-\")(0),x(3).toInt)).reduceByKey(_+_, 1).collect()", "" ] ], [ [ "Suppose we want to find pages that were viewed more than 200,000 times during the three days covered by our dataset. Conceptually, this task is similar to the previous query. But, given the large number of pages (23 million distinct page names), the new task is very expensive. We are doing an expensive group-by with a lot of network shuffling of data. To recap, first we split each line of data into its respective fields. Next, we extract the fields for page name and number of page views. We reduce by key again, this time with 40 reducers. Then we filter out pages with less than 200,000 total views over our time window represented by our dataset.", "_____no_output_____" ] ], [ [ "// TODO: Replace <FILL IN> with appropriate code\nenPages.map(x => x.split(\" \")).map(x => (x(2), x(3).toInt))\n.reduceByKey(_+_, 40).filter(x => x._2 >= 200000).map(x => (x._2, x._1)).collect()", "" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

ec7b7619da0a81327ef8bfce798a3ab5be8ce19d

33,353

ipynb

Jupyter Notebook

kMeans.ipynb

andresvillamayor/Ejemplos_ML

4f1cbb2f775e7723d29b5b878ab6798d7dce3b3f

[ "Apache-2.0" ]

null

kMeans.ipynb

andresvillamayor/Ejemplos_ML

4f1cbb2f775e7723d29b5b878ab6798d7dce3b3f

[ "Apache-2.0" ]

null

kMeans.ipynb

andresvillamayor/Ejemplos_ML

4f1cbb2f775e7723d29b5b878ab6798d7dce3b3f

[ "Apache-2.0" ]

null

54.766831

12,628

0.586094

[ [ [ "# Cargar funciones de la librería de python data analysis\nimport pandas as pd \nimport numpy as np\n\n# Leer csv con datos y cargar en el dataframe data\ndata = pd.read_csv(\"creditos.csv\") \n\n# calcular variable edad a partir de fecha de solicitud - fecha de nacimiento\ndata['fechaHora'] = pd.to_datetime(data['fechaHora'])\ndata['nacimiento'] = pd.to_datetime(data['nacimiento'])\ndata['edad'] = ((data['fechaHora']-data['nacimiento'])/np.timedelta64(1,'Y')).astype(int)\n\ndata.head()", "_____no_output_____" ], [ "from sklearn.cluster import KMeans\nimport matplotlib.pyplot as plt \nfrom sklearn import preprocessing\n#feature selection \ndf = data[['edad','cliente_nuevo_o_recurrente','monto_solicitado','tiene_ips',\n 'plazo_solicitado','ingreso_neto_mensual','resultadoFinal']]\n\n# One-hot encoding para variables categoricas\nx = pd.get_dummies(df)\n\n# Normalizacion a [0-1]\nmin_max_scaler = preprocessing.MinMaxScaler()\nxNorm = pd.DataFrame(min_max_scaler.fit_transform(x.values))\nxNorm.head()", "_____no_output_____" ], [ "# Elbow Curve\nnc = range(1, 25)\nkmeans = [KMeans(n_clusters=i) for i in nc]\nscore = [kmeans[i].fit(xNorm).score(xNorm) for i in range(len(kmeans))]\n\nplt.plot(nc, score, color='green') \nplt.xlabel('Clusters')\nplt.ylabel('Score')\nplt.title('Elbow Curve')\nplt.show()", "_____no_output_____" ], [ "kmeans = KMeans(n_clusters=7, max_iter=300, random_state=1)\nkmeans.fit(xNorm)\nprint(\"Score: \" + str(kmeans.score(xNorm)))\nprint(kmeans.labels_)\n#print(kmeans.cluster_centers_)\n\nclusters = pd.DataFrame(min_max_scaler.inverse_transform(kmeans.cluster_centers_), columns=x.columns)\nclusters['tiene_ips'] = clusters['tiene_ips'].round()\nclusters['cliente_nuevo_o_recurrente_N'] = clusters['cliente_nuevo_o_recurrente_N'].round()\nclusters['cliente_nuevo_o_recurrente_R'] = clusters['cliente_nuevo_o_recurrente_R'].round()\nclusters['resultadoFinal_BIEN'] = clusters['resultadoFinal_BIEN'].round()\nclusters['resultadoFinal_MAL'] = clusters['resultadoFinal_MAL'].round()\nclusters", "Score: -387.72895396855506\n[0 0 0 ... 5 3 0]\n" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code" ] ]

ec7b78bbcc90c68aa58d33258c2ca2d864982fb3

288,623

ipynb

Jupyter Notebook

.ipynb_checkpoints/ThermistorNB-checkpoint.ipynb

drtaiga/Thermistor_Bridge

09ec7fcc8f05b230547af5ceea188da7a62448eb

[ "MIT" ]

null

.ipynb_checkpoints/ThermistorNB-checkpoint.ipynb

drtaiga/Thermistor_Bridge

09ec7fcc8f05b230547af5ceea188da7a62448eb

[ "MIT" ]

null

.ipynb_checkpoints/ThermistorNB-checkpoint.ipynb

drtaiga/Thermistor_Bridge

09ec7fcc8f05b230547af5ceea188da7a62448eb

[ "MIT" ]

null

58.213594

23,176

0.619902

[ [ [ "%config InlineBackend.figure_format = 'svg'\nfrom NB_Setup import *\n%matplotlib inline", "_____no_output_____" ] ], [ [ "# Thermistors\n\n$$ R(T) = R_* \\exp \\left[ B \\left( \\frac{1}{T} - \\frac{1}{T_*} \\right) \\right] \\$$", "_____no_output_____" ] ], [ [ "thermres_plot()", "_____no_output_____" ] ], [ [ "# Voltage Divider Circuits", "_____no_output_____" ], [ "## Generic Divider Circuits", "_____no_output_____" ] ], [ [ "draw_dividers()", "_____no_output_____" ] ], [ [ "$$V^{(1A)} = V_o \\left[ \\frac{R_1}{R_0 + R_1} \\right] \\qquad \\qquad V^{(1B)} = V_o \\left[ \\frac{R_0}{R_0 + R_1} \\right] \\tag{1}$$", "_____no_output_____" ] ], [ [ "draw_dividers2()", "_____no_output_____" ] ], [ [ "$$V^{(2A)} = V_o \\left[ \\frac{R(T)}{R_0 + R(T)} \\right] \\qquad \\qquad V^{(2B)} = V_o \\left[ \\frac{R_0}{R_0 + R(T)} \\right] \\tag{1}$$", "_____no_output_____" ] ], [ [ "divider_plot()", "_____no_output_____" ], [ "draw_bridge()", "_____no_output_____" ] ], [ [ "$$\\Delta V^{(3A)} = V^{(3A)}_T - V^{(3A)}_B = V_0 \\left[ \\frac{1}{1 + \\rho} - \\frac{R(T)}{R_0 + R(T)} \\right] $$\n\nand \n\n$$\\Delta V^{(3B)} = V^{(3B)}_T - V^{(3B)}_B = V_0 \\left[ \\frac{1}{1 + \\rho} - \\frac{R_0}{R_0 + R(T)} \\right].$$\n\n... the effects of two resistors in the upper bridge are encapsulated by the resistor ratio $\\rho = R_1/R_3$.", "_____no_output_____" ] ], [ [ "bridge_plot2()", "_____no_output_____" ] ], [ [ "## Thermistor Divider Circuits", "_____no_output_____" ] ], [ [ "draw_divamp()", "_____no_output_____" ] ], [ [ "$$V = V_o \\left[ \\frac{R_0}{R_0 + R(T)} \\right] \\times A_G + V_{\\it{ref}}\n\\tag{1}$$", "_____no_output_____" ], [ "with the $A_G$ is the amplifier gain and $V_{\\it ref}$ is a reference voltage that provides an offset for the amplified signal. The amplifier in figure -- is an instrumentation aplifier rather than an ordinary op-amp. ...", "_____no_output_____" ] ], [ [ "draw_bridgeamp()", "_____no_output_____" ] ], [ [ "# Bridge Circuits", "_____no_output_____" ], [ "## Generic Wheatstone Bridge Circuits", "_____no_output_____" ] ], [ [ "ampcircuits_plot()", "_____no_output_____" ] ], [ [ "# Appendix:", "_____no_output_____" ], [ "## Inflection point", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ] ]

ec7b8717af718d779089331cc46e8b283bb863e8

180,538

ipynb

Jupyter Notebook

scripts/molpal_analysis.ipynb

ashuein/molpal

1e17a0c406516ceaeaf273a6983d06206bcfe76f

[ "MIT" ]

1

2021-01-12T11:51:26.000Z

scripts/molpal_analysis.ipynb

ashuein/molpal

1e17a0c406516ceaeaf273a6983d06206bcfe76f

[ "MIT" ]

null

scripts/molpal_analysis.ipynb

ashuein/molpal

1e17a0c406516ceaeaf273a6983d06206bcfe76f

[ "MIT" ]

null

61.449285

526

0.397866

[ [ [ "# MolPAL Figure Notebook\n\nthe first step in recreating most figures from the main text will be to import everything and load all of the data. To do that, go down to [this cell](#RUN-ALL-CELLS-ABOVE-ME), and run all cells above", "_____no_output_____" ] ], [ [ "import plotly.graph_objects as go\nimport plotly.io as pio\nimport plotly.express as px\nfrom plotly.subplots import make_subplots\n\nimport pandas as pd\n\npio.templates.default = 'plotly_white'\n\nMODELS = ['rf', 'nn', 'mpn']\nMETRICS = ['greedy', 'ucb', 'thompson', 'ei', 'pi']\nMETRIC_NAMES = {'greedy': 'greedy', 'ucb': 'UCB', 'thompson': 'TS',\n 'ei': 'EI', 'pi': 'PI'}\nSPLITS = [0.4, 0.2, 0.1]\n\nDASHES = ['dash', 'dot', 'dashdot']\nMARKERS = ['circle', 'square', 'diamond']\nMETRIC_COLORS = px.colors.qualitative.Plotly\nMODEL_COLORS = px.colors.qualitative.D3", "_____no_output_____" ], [ "E10k_random = {\n 'avg': ([80.023, 85.934, 88.358, 89.827, 90.707, 91.408],\n [1.149, 0.459, 0.431, 0.508, 0.270, 0.337]),\n 'scores': ([1.600, 2.800, 3.200, 4.200, 4.800, 5.600],\n [1.356, 1.470, 1.600, 1.327, 0.748, 0.800]),\n 'smis': ([1.600, 2.800, 3.200, 3.800, 4.400, 5.000],\n [1.356, 1.470, 1.600, 1.600, 1.020, 0.894]),\n}", "_____no_output_____" ], [ "#top-100\nE10k_online = {\n 'mpn': {'avg': {'ei': ([80.166, 92.926, 94.476, 96.067, 97.065, 97.775],\n [0.961, 0.076, 0.291, 0.587, 0.266, 0.265]),\n 'greedy': ([79.909, 92.176, 94.513, 96.101, 96.941, 97.682],\n [0.532, 1.091, 0.242, 0.196, 0.395, 0.501]),\n 'pi': ([79.961, 91.992, 94.141, 95.639, 96.529, 97.398],\n [0.565, 0.900, 0.460, 0.346, 0.370, 0.286]),\n 'thompson': ([79.865, 91.644, 93.073, 95.726, 96.185, 97.442],\n [0.702, 0.865, 1.170, 0.917, 0.886, 0.334]),\n 'ucb': ([79.799, 91.754, 94.093, 95.993, 97.055, 97.825],\n [0.585, 0.809, 0.600, 0.372, 0.451, 0.471])},\n 'scores': {'ei': ([1.800, 15.800, 20.400, 29.000, 36.400, 45.400],\n [0.748, 0.748, 1.356, 3.286, 2.417, 5.004]),\n 'greedy': ([1.600, 14.800, 20.400, 28.600, 36.400, 44.600],\n [1.625, 2.638, 1.020, 2.154, 5.276, 6.946]),\n 'pi': ([1.200, 13.000, 18.200, 24.000, 30.200, 39.400],\n [0.748, 1.414, 2.400, 2.280, 3.600, 3.262]),\n 'thompson': ([0.800, 10.800, 13.400, 25.800, 28.800, 41.400],\n [0.748, 3.868, 5.783, 7.250, 7.859, 4.673]),\n 'ucb': ([1.000, 12.200, 18.400, 27.400, 36.400, 45.800],\n [0.632, 2.400, 1.625, 3.323, 5.886, 7.111])},\n 'smis': {'ei': ([1.800, 13.400, 17.200, 24.800, 31.000, 38.800],\n [0.748, 1.020, 0.980, 3.059, 2.449, 3.429]),\n 'greedy': ([1.200, 12.800, 17.000, 23.600, 30.000, 37.200],\n [1.166, 2.135, 1.414, 2.332, 4.050, 6.013]),\n 'pi': ([1.000, 10.800, 14.200, 19.400, 24.600, 33.200],\n [0.632, 0.748, 2.040, 2.245, 3.200, 2.926]),\n 'thompson': ([0.800, 10.200, 12.600, 23.200, 25.800, 37.000],\n [0.748, 4.354, 6.280, 7.222, 7.935, 5.215]),\n 'ucb': ([0.800, 10.800, 15.400, 22.600, 30.400, 38.800],\n [0.748, 2.040, 2.059, 3.137, 5.122, 6.274])}},\n 'nn': {'avg': {'ei': ([79.708, 92.806, 93.385, 94.906, 95.683, 96.094],\n [0.518, 0.432, 1.239, 1.007, 1.026, 1.119]),\n 'greedy': ([79.683, 92.938, 96.221, 97.473, 97.949, 98.487],\n [0.471, 1.073, 0.634, 0.505, 0.479, 0.305]),\n 'pi': ([79.865, 92.822, 94.919, 95.898, 96.368, 96.931],\n [0.552, 1.015, 0.771, 0.693, 0.698, 0.920]),\n 'thompson': ([79.690, 92.336, 95.169, 96.796, 97.634, 98.162],\n [0.686, 0.635, 0.434, 0.208, 0.157, 0.154]),\n 'ucb': ([79.760, 92.450, 94.948, 96.099, 97.022, 97.587],\n [0.818, 0.992, 0.977, 1.145, 1.097, 0.821])},\n 'scores': {'ei': ([1.600, 12.800, 15.200, 20.000, 24.000, 28.000],\n [1.020, 3.250, 6.431, 5.727, 5.292, 6.693]),\n 'greedy': ([0.800, 15.600, 28.600, 40.200, 46.400, 55.400],\n [0.980, 3.878, 6.312, 7.705, 7.761, 6.184]),\n 'pi': ([1.600, 15.400, 23.200, 29.000, 32.000, 37.600],\n [0.490, 4.224, 3.124, 4.427, 4.195, 6.711]),\n 'thompson': ([1.800, 13.400, 22.600, 34.200, 42.800, 51.200],\n [0.400, 1.497, 3.262, 2.040, 1.600, 3.868]),\n 'ucb': ([0.600, 14.200, 22.600, 29.000, 37.000, 43.400],\n [0.800, 3.250, 6.406, 10.450, 11.610, 11.586])},\n 'smis': {'ei': ([1.600, 11.600, 14.000, 18.000, 21.400, 24.600],\n [1.020, 2.417, 6.164, 5.550, 5.535, 6.216]),\n 'greedy': ([0.800, 14.600, 26.000, 36.000, 41.200, 49.800],\n [0.980, 3.262, 6.573, 8.025, 8.518, 6.369]),\n 'pi': ([1.400, 14.000, 20.600, 25.000, 27.600, 33.000],\n [0.800, 3.950, 2.577, 3.578, 3.007, 5.404]),\n 'thompson': ([1.400, 12.200, 20.600, 30.400, 38.400, 45.800],\n [0.490, 2.227, 4.030, 3.072, 2.800, 3.544]),\n 'ucb': ([0.400, 13.200, 20.600, 26.200, 33.000, 38.400],\n [0.800, 3.487, 6.530, 9.826, 10.526, 10.111])}},\n 'rf': {'avg': {'ei': ([79.725, 89.860, 93.306, 95.026, 95.960, 96.533],\n [0.912, 0.849, 0.676, 0.506, 0.717, 0.491]),\n 'greedy': ([79.731, 92.816, 95.461, 96.767, 97.446, 97.864],\n [0.582, 0.569, 0.566, 0.514, 0.318, 0.192]),\n 'pi': ([79.865, 91.435, 94.335, 95.585, 96.235, 97.057],\n [1.177, 0.483, 0.543, 0.426, 0.489, 0.585]),\n 'thompson': ([79.035, 87.929, 90.754, 92.526, 93.712, 94.575],\n [0.479, 0.571, 0.673, 0.381, 0.360, 0.313]),\n 'ucb': ([79.166, 89.515, 92.760, 94.679, 95.759, 96.457],\n [0.734, 1.759, 0.769, 0.858, 0.556, 0.506])},\n 'scores': {'ei': ([1.200, 7.400, 16.000, 22.000, 27.600, 32.200],\n [0.748, 2.154, 5.550, 6.542, 9.178, 7.250]),\n 'greedy': ([2.200, 15.600, 25.400, 34.000, 41.400, 46.200],\n [1.166, 2.417, 4.964, 5.899, 4.363, 2.135]),\n 'pi': ([1.200, 10.600, 19.600, 25.200, 29.400, 36.800],\n [0.748, 2.871, 5.004, 4.707, 4.673, 6.274]),\n 'thompson': ([1.400, 4.400, 7.400, 10.800, 14.400, 17.400],\n [0.800, 2.728, 3.929, 3.655, 3.007, 3.200]),\n 'ucb': ([0.600, 7.400, 13.400, 19.600, 25.200, 30.200],\n [0.800, 3.555, 4.716, 5.122, 5.776, 5.776])},\n 'smis': {'ei': ([1.200, 6.200, 13.200, 18.200, 22.800, 27.000],\n [0.748, 2.482, 4.707, 5.600, 7.547, 5.762]),\n 'greedy': ([2.200, 13.800, 22.800, 30.400, 37.000, 40.800],\n [1.166, 2.482, 4.833, 6.829, 5.215, 3.655]),\n 'pi': ([0.800, 8.800, 16.800, 21.800, 25.200, 31.400],\n [0.748, 2.227, 4.354, 4.167, 4.792, 5.389]),\n 'thompson': ([1.200, 3.600, 6.400, 9.000, 12.600, 15.200],\n [0.748, 2.577, 3.555, 3.162, 2.245, 3.250]),\n 'ucb': ([0.600, 6.600, 11.200, 16.800, 22.000, 26.800],\n [0.800, 3.200, 4.261, 5.671, 5.727, 5.307])}}}", "_____no_output_____" ], [ "E10k_retrain = {\n 'mpn': {'avg': {'ei': ([79.427, 91.615, 95.612, 97.373, 98.203, 98.795],\n [0.597, 0.933, 0.983, 0.450, 0.361, 0.190]),\n 'greedy': ([79.901, 92.609, 95.734, 97.523, 98.477, 98.942],\n [0.509, 0.377, 0.642, 0.351, 0.163, 0.118]),\n 'pi': ([79.950, 92.379, 95.689, 97.448, 98.381, 98.758],\n [0.472, 0.832, 0.387, 0.164, 0.204, 0.161]),\n 'thompson': ([79.232, 91.442, 95.066, 96.817, 97.858, 98.584],\n [0.770, 0.843, 0.427, 0.377, 0.346, 0.199]),\n 'ucb': ([78.905, 91.669, 95.471, 97.332, 98.168, 98.800],\n [0.394, 1.047, 0.334, 0.379, 0.261, 0.166])},\n 'scores': {'ei': ([1.800, 13.400, 28.000, 42.400, 54.000, 64.400],\n [0.748, 4.224, 6.261, 5.389, 5.865, 3.826]),\n 'greedy': ([1.600, 15.200, 28.400, 44.400, 58.400, 67.000],\n [0.490, 2.638, 4.630, 5.238, 3.611, 3.033]),\n 'pi': ([1.200, 14.600, 25.800, 41.600, 56.400, 63.000],\n [0.748, 2.498, 2.638, 1.625, 3.878, 3.162]),\n 'thompson': ([2.000, 13.200, 23.200, 35.400, 47.800, 60.000],\n [0.632, 2.482, 2.315, 4.454, 5.636, 4.099]),\n 'ucb': ([0.200, 12.400, 26.000, 42.600, 53.200, 64.400],\n [0.400, 3.774, 2.280, 5.004, 4.750, 4.079])},\n 'smis': {'ei': ([1.400, 12.200, 24.600, 37.200, 47.000, 57.000],\n [0.490, 3.544, 5.463, 4.020, 5.020, 3.578]),\n 'greedy': ([1.200, 13.000, 24.000, 38.200, 51.000, 59.600],\n [0.400, 2.191, 3.347, 4.020, 2.280, 2.332]),\n 'pi': ([1.000, 12.800, 21.600, 35.600, 48.400, 55.200],\n [0.894, 2.482, 3.611, 2.154, 3.007, 3.311]),\n 'thompson': ([1.800, 12.000, 20.200, 30.800, 41.600, 52.200],\n [0.748, 2.280, 1.166, 3.311, 5.426, 4.622]),\n 'ucb': ([0.000, 9.800, 21.400, 36.600, 45.800, 57.000],\n [0.000, 3.970, 2.577, 4.454, 4.214, 3.162])}},\n 'nn': {'avg': {'ei': ([79.700, 92.692, 96.330, 97.667, 98.071, 98.423],\n [0.930, 0.798, 0.382, 0.248, 0.367, 0.416]),\n 'greedy': ([79.093, 93.636, 96.678, 97.796, 98.419, 98.965],\n [0.429, 1.159, 0.579, 0.397, 0.312, 0.202]),\n 'pi': ([79.865, 92.806, 96.053, 97.434, 97.965, 98.553],\n [0.857, 0.777, 0.372, 0.163, 0.299, 0.150]),\n 'thompson': ([79.822, 92.437, 95.821, 97.351, 98.222, 98.729],\n [0.481, 0.826, 0.443, 0.149, 0.213, 0.192]),\n 'ucb': ([79.988, 92.452, 95.892, 97.411, 98.160, 98.588],\n [0.944, 0.696, 0.182, 0.172, 0.166, 0.164])},\n 'scores': {'ei': ([0.800, 14.400, 30.800, 44.200, 50.400, 56.000],\n [0.748, 3.720, 4.578, 4.354, 6.621, 7.457]),\n 'greedy': ([1.400, 20.600, 34.400, 46.800, 55.200, 66.800],\n [1.744, 4.176, 5.678, 5.564, 5.946, 5.418]),\n 'pi': ([1.200, 13.800, 28.600, 41.600, 47.800, 57.800],\n [1.166, 1.720, 3.262, 2.871, 4.956, 2.400]),\n 'thompson': ([1.200, 12.800, 25.400, 39.000, 52.800, 61.400],\n [0.748, 1.600, 1.356, 1.095, 3.868, 3.878]),\n 'ucb': ([1.600, 12.200, 27.400, 41.200, 50.600, 58.000],\n [0.800, 2.561, 1.020, 3.059, 3.666, 3.521])},\n 'smis': {'ei': ([0.600, 13.200, 27.600, 39.600, 45.000, 49.800],\n [0.490, 4.445, 4.317, 3.878, 6.325, 6.853]),\n 'greedy': ([1.000, 18.600, 30.600, 40.800, 48.600, 59.200],\n [1.265, 3.878, 5.571, 5.810, 6.468, 6.145]),\n 'pi': ([1.200, 13.000, 27.000, 38.200, 43.200, 51.600],\n [1.166, 1.789, 2.530, 2.135, 3.919, 2.332]),\n 'thompson': ([1.200, 11.400, 22.400, 34.600, 46.200, 54.600],\n [0.748, 2.059, 1.744, 1.625, 2.926, 3.441]),\n 'ucb': ([1.600, 11.000, 25.000, 36.600, 44.400, 51.200],\n [0.800, 2.608, 1.414, 2.577, 3.441, 3.370])}},\n 'rf': {'avg': {'ei': ([79.180, 90.721, 94.074, 95.639, 96.664, 97.158],\n [0.514, 1.350, 0.811, 0.610, 0.741, 0.764]),\n 'greedy': ([79.981, 92.636, 95.469, 96.864, 97.558, 98.206],\n [0.781, 1.316, 0.556, 0.397, 0.359, 0.306]),\n 'pi': ([79.884, 90.338, 94.403, 96.457, 97.274, 97.816],\n [1.071, 2.804, 1.136, 0.295, 0.278, 0.245]),\n 'thompson': ([80.180, 88.761, 91.976, 93.493, 95.002, 95.972],\n [0.413, 0.503, 0.370, 0.367, 0.352, 0.339]),\n 'ucb': ([79.762, 90.365, 94.018, 95.751, 96.968, 97.580],\n [0.629, 0.323, 0.340, 0.271, 0.212, 0.248])},\n 'scores': {'ei': ([0.600, 9.400, 18.200, 26.000, 33.600, 39.400],\n [0.800, 2.871, 4.354, 5.404, 8.163, 9.478]),\n 'greedy': ([0.600, 13.600, 24.400, 34.200, 41.400, 51.600],\n [0.800, 4.499, 3.072, 4.069, 5.987, 5.851]),\n 'pi': ([1.000, 10.800, 21.400, 34.000, 40.200, 47.600],\n [1.095, 4.622, 3.826, 2.098, 2.926, 4.176]),\n 'thompson': ([1.600, 4.800, 9.600, 13.000, 20.000, 27.600],\n [0.490, 1.600, 2.154, 1.789, 2.098, 1.855]),\n 'ucb': ([0.400, 6.400, 15.400, 25.600, 36.400, 43.200],\n [0.490, 2.154, 1.960, 2.728, 3.382, 3.370])},\n 'smis': {'ei': ([0.600, 8.000, 15.400, 22.600, 28.800, 33.800],\n [0.800, 2.280, 3.611, 5.352, 7.547, 9.108]),\n 'greedy': ([0.600, 10.800, 21.000, 30.200, 36.200, 44.800],\n [0.800, 4.354, 3.347, 4.707, 5.492, 5.776]),\n 'pi': ([1.000, 8.800, 18.600, 29.600, 35.400, 41.400],\n [1.095, 3.655, 3.499, 3.072, 4.030, 3.323]),\n 'thompson': ([1.400, 4.200, 8.600, 11.200, 16.800, 22.600],\n [0.490, 1.470, 1.855, 1.600, 2.315, 2.653]),\n 'ucb': ([0.400, 5.600, 13.400, 22.000, 31.600, 37.200],\n [0.490, 2.059, 2.059, 2.898, 3.720, 3.059])}}}", "_____no_output_____" ], [ "E50k_random = {\n 'avg': ([78.964, 85.635, 88.049, 89.505, 90.516, 91.362],\n [0.316, 0.266, 0.252, 0.286, 0.146, 0.193]),\n 'scores': ([0.281, 2.040, 3.160, 4.280, 5.400, 6.600],\n [0.273, 0.344, 0.585, 1.136, 1.012, 1.095]),\n 'smis': ([0.720, 1.960, 3.040, 4.000, 5.040, 6.080],\n [0.271, 0.367, 0.625, 1.152, 1.155, 1.170]),\n}", "_____no_output_____" ], [ "#top-500\nE50k_online = {\n 'mpn': {'avg': {'ei': ([78.817, 94.143, 96.742, 98.010, 98.571, 98.953],\n [0.413, 0.394, 0.163, 0.047, 0.133, 0.111]),\n 'greedy': ([78.456, 94.235, 96.030, 97.700, 98.245, 98.698],\n [0.294, 0.341, 0.712, 0.261, 0.403, 0.318]),\n 'pi': ([78.845, 94.161, 95.903, 97.610, 98.083, 98.640],\n [0.327, 0.248, 0.883, 0.442, 0.592, 0.441]),\n 'thompson': ([78.816, 92.628, 96.029, 97.598, 98.485, 98.935],\n [0.293, 0.606, 0.133, 0.225, 0.131, 0.093]),\n 'ucb': ([78.851, 94.464, 96.884, 98.020, 98.596, 99.000],\n [0.160, 0.147, 0.223, 0.071, 0.239, 0.152])},\n 'scores': {'ei': ([0.602, 22.600, 35.480, 47.880, 56.200, 63.320],\n [0.380, 3.642, 1.904, 1.070, 2.245, 2.594]),\n 'greedy': ([0.399, 23.520, 30.960, 43.760, 51.080, 58.400],\n [0.670, 1.787, 3.563, 3.008, 5.344, 4.943]),\n 'pi': ([0.523, 23.560, 30.600, 42.520, 49.000, 57.680],\n [0.241, 1.183, 4.445, 5.326, 8.494, 8.073]),\n 'thompson': ([0.640, 15.040, 28.440, 41.680, 53.680, 62.200],\n [0.665, 2.378, 1.722, 3.597, 2.830, 2.671]),\n 'ucb': ([0.603, 23.760, 36.600, 47.920, 56.360, 64.640],\n [0.524, 1.286, 2.280, 1.737, 4.421, 3.483])},\n 'smis': {'ei': ([0.960, 21.240, 33.080, 44.760, 52.480, 59.160],\n [0.320, 3.171, 1.719, 1.038, 2.215, 2.490]),\n 'greedy': ([0.960, 22.040, 28.800, 40.920, 47.880, 54.840],\n [0.265, 1.839, 3.119, 2.600, 4.999, 4.560]),\n 'pi': ([1.160, 22.440, 28.760, 40.000, 46.080, 54.280],\n [0.150, 1.203, 3.636, 4.912, 8.029, 7.616]),\n 'thompson': ([1.000, 14.200, 26.800, 39.280, 50.640, 58.440],\n [0.379, 2.184, 1.431, 3.300, 2.590, 2.292]),\n 'ucb': ([1.040, 22.520, 34.120, 44.840, 52.800, 60.640],\n [0.463, 1.348, 2.160, 1.856, 4.280, 3.522])}},\n 'nn': {'avg': {'ei': ([79.020, 94.143, 94.995, 96.462, 96.894, 97.103],\n [0.453, 0.454, 0.919, 1.155, 1.002, 1.051]),\n 'greedy': ([78.994, 94.823, 97.227, 98.365, 98.790, 99.088],\n [0.211, 0.256, 0.201, 0.145, 0.139, 0.104]),\n 'pi': ([79.079, 94.154, 95.026, 96.327, 97.063, 97.254],\n [0.503, 0.407, 1.123, 1.298, 0.874, 0.699]),\n 'thompson': ([78.598, 93.866, 96.408, 97.508, 98.188, 98.560],\n [0.423, 0.437, 0.395, 0.361, 0.347, 0.167]),\n 'ucb': ([78.524, 94.303, 95.975, 97.316, 97.856, 98.291],\n [0.354, 0.673, 0.718, 0.935, 0.731, 0.541])},\n 'scores': {'ei': ([0.724, 22.920, 25.840, 34.800, 37.800, 39.480],\n [0.414, 3.154, 5.107, 8.663, 7.474, 9.225]),\n 'greedy': ([0.801, 26.640, 40.320, 53.360, 60.320, 66.520],\n [0.632, 2.061, 2.081, 2.010, 2.415, 2.211]),\n 'pi': ([0.721, 22.200, 26.320, 33.320, 37.800, 38.840],\n [0.808, 1.802, 5.792, 10.131, 7.915, 7.244]),\n 'thompson': ([0.682, 21.760, 33.120, 42.240, 50.520, 55.920],\n [0.517, 1.903, 3.428, 4.240, 5.683, 3.491]),\n 'ucb': ([0.520, 23.880, 31.080, 41.920, 47.200, 52.720],\n [0.924, 4.287, 4.960, 9.946, 10.347, 10.090])},\n 'smis': {'ei': ([1.320, 21.480, 24.280, 32.720, 35.520, 37.120],\n [0.240, 3.040, 4.857, 8.176, 6.946, 8.646]),\n 'greedy': ([1.040, 25.160, 38.080, 50.160, 56.920, 62.840],\n [0.367, 2.122, 1.862, 2.141, 2.506, 2.041]),\n 'pi': ([1.320, 21.240, 25.040, 31.520, 35.760, 36.680],\n [0.449, 1.525, 5.432, 9.441, 7.587, 6.818]),\n 'random': ([0.720, 1.960, 3.040, 4.000, 5.040, 6.080],\n [0.271, 0.367, 0.625, 1.152, 1.155, 1.170]),\n 'thompson': ([0.960, 20.480, 31.120, 39.560, 47.440, 52.600],\n [0.463, 1.809, 3.367, 4.005, 5.431, 3.272]),\n 'ucb': ([1.080, 22.760, 29.520, 39.720, 44.760, 49.920],\n [0.652, 4.007, 4.475, 9.394, 9.687, 9.301])}},\n 'rf': {'avg': {'ei': ([78.786, 91.711, 94.464, 95.952, 96.705, 97.329],\n [0.208, 0.925, 0.368, 0.417, 0.550, 0.249]),\n 'greedy': ([78.628, 93.700, 96.734, 97.921, 98.378, 98.746],\n [0.490, 0.382, 0.173, 0.104, 0.080, 0.045]),\n 'pi': ([79.068, 92.409, 95.560, 96.727, 97.501, 97.997],\n [0.308, 0.663, 0.267, 0.290, 0.236, 0.132]),\n 'thompson': ([78.810, 89.425, 92.663, 94.463, 95.571, 96.296],\n [0.419, 0.779, 0.385, 0.297, 0.146, 0.111]),\n 'ucb': ([78.854, 92.510, 94.801, 96.293, 97.159, 97.724],\n [0.243, 0.240, 0.684, 0.372, 0.318, 0.147])},\n 'scores': {'ei': ([0.401, 11.880, 20.080, 27.560, 33.080, 38.560],\n [0.553, 2.827, 1.107, 3.563, 5.367, 2.684]),\n 'greedy': ([0.201, 20.600, 36.040, 47.560, 53.800, 60.360],\n [0.285, 1.757, 1.176, 1.587, 1.431, 1.155]),\n 'pi': ([0.601, 14.800, 26.440, 33.360, 40.680, 46.440],\n [0.612, 2.985, 1.582, 2.877, 2.933, 1.830]),\n 'thompson': ([0.202, 6.760, 12.320, 18.360, 23.560, 27.960],\n [0.285, 2.037, 1.532, 1.556, 1.359, 1.235]),\n 'ucb': ([0.640, 14.640, 21.680, 30.120, 37.800, 43.480],\n [0.690, 1.106, 3.465, 3.329, 3.277, 1.792])},\n 'smis': {'ei': ([0.920, 11.520, 19.440, 26.720, 31.920, 37.120],\n [0.449, 2.982, 1.428, 3.704, 5.280, 2.667]),\n 'greedy': ([0.760, 19.240, 33.440, 44.400, 50.200, 56.400],\n [0.388, 1.675, 0.991, 1.486, 1.246, 0.938]),\n 'pi': ([1.200, 14.320, 25.640, 32.040, 39.160, 44.560],\n [0.219, 2.468, 1.405, 2.658, 2.658, 1.394]),\n 'thompson': ([1.000, 6.440, 11.800, 17.600, 22.440, 26.680],\n [0.358, 1.899, 1.585, 1.507, 1.353, 1.306]),\n 'ucb': ([0.920, 13.960, 20.680, 28.800, 36.040, 41.360],\n [0.271, 0.933, 3.461, 3.277, 3.340, 1.932])}}}", "_____no_output_____" ], [ "E50k_retrain = {\n 'mpn': {'avg': {'ei': ([78.787, 94.248, 97.409, 98.469, 98.924, 99.122],\n [0.298, 0.187, 0.076, 0.072, 0.067, 0.032]),\n 'greedy': ([78.867, 94.544, 97.597, 98.590, 99.088, 99.340],\n [0.202, 0.299, 0.124, 0.061, 0.048, 0.030]),\n 'pi': ([78.864, 94.373, 97.496, 98.527, 98.972, 99.170],\n [0.331, 0.311, 0.200, 0.091, 0.067, 0.074]),\n 'thompson': ([78.847, 92.568, 96.386, 98.134, 98.850, 99.250],\n [0.325, 0.637, 0.354, 0.136, 0.099, 0.047]),\n 'ucb': ([78.492, 94.490, 97.507, 98.585, 99.068, 99.329],\n [0.411, 0.124, 0.129, 0.051, 0.022, 0.024])},\n 'scores': {'ei': ([0.399, 23.240, 41.280, 54.040, 62.160, 67.200],\n [0.964, 1.311, 1.230, 1.666, 1.359, 0.938]),\n 'greedy': ([0.921, 24.920, 43.360, 56.800, 66.440, 72.920],\n [0.413, 1.657, 1.091, 1.706, 1.255, 1.269]),\n 'pi': ([0.724, 23.480, 41.720, 55.240, 63.720, 69.160],\n [0.349, 3.090, 2.197, 1.617, 1.163, 1.689]),\n 'thompson': ([0.120, 14.000, 30.600, 48.760, 60.520, 70.360],\n [0.519, 3.067, 2.896, 1.541, 1.378, 0.686]),\n 'ucb': ([0.240, 25.000, 41.920, 56.080, 65.520, 72.720],\n [0.625, 0.522, 1.690, 0.952, 0.816, 0.500])},\n 'smis': {'ei': ([0.920, 22.080, 38.560, 50.400, 58.080, 62.880],\n [0.640, 1.348, 1.335, 1.565, 1.078, 0.900]),\n 'greedy': ([0.960, 23.640, 40.640, 53.360, 62.640, 68.840],\n [0.480, 1.439, 1.091, 1.428, 1.155, 1.031]),\n 'pi': ([1.400, 22.360, 39.320, 52.000, 59.920, 64.920],\n [0.219, 2.888, 1.874, 1.649, 1.269, 1.505]),\n 'thompson': ([0.760, 13.160, 28.440, 45.720, 56.840, 66.040],\n [0.320, 2.951, 2.877, 1.613, 1.222, 0.686]),\n 'ucb': ([0.920, 23.680, 39.240, 52.560, 61.560, 68.400],\n [0.574, 0.412, 1.347, 0.916, 0.674, 0.438])}},\n 'nn': {'avg': {'ei': ([78.627, 94.222, 97.266, 98.122, 98.761, 99.076],\n [0.344, 0.329, 0.160, 0.385, 0.081, 0.118]),\n 'greedy': ([78.708, 94.617, 97.765, 98.687, 99.129, 99.388],\n [0.156, 0.297, 0.097, 0.074, 0.066, 0.048]),\n 'pi': ([78.537, 94.298, 97.360, 98.191, 98.762, 99.077],\n [0.168, 0.460, 0.255, 0.531, 0.287, 0.160]),\n 'thompson': ([79.186, 93.843, 97.374, 98.559, 99.075, 99.350],\n [0.345, 0.349, 0.213, 0.103, 0.048, 0.072]),\n 'ucb': ([78.919, 94.173, 97.583, 98.633, 99.114, 99.385],\n [0.236, 0.248, 0.186, 0.045, 0.071, 0.040])},\n 'scores': {'ei': ([0.321, 22.200, 40.200, 50.080, 59.920, 66.080],\n [0.350, 1.939, 1.152, 4.045, 1.478, 2.963]),\n 'greedy': ([0.520, 25.720, 45.600, 58.640, 67.920, 74.760],\n [0.450, 1.746, 1.437, 1.546, 1.760, 1.127]),\n 'pi': ([0.200, 23.920, 41.960, 52.160, 60.640, 67.200],\n [0.509, 2.664, 2.330, 7.040, 5.278, 4.002]),\n 'thompson': ([0.961, 21.560, 41.400, 56.320, 66.200, 73.360],\n [0.599, 2.289, 2.527, 2.088, 1.233, 2.257]),\n 'ucb': ([0.682, 23.120, 44.480, 58.080, 67.120, 74.440],\n [0.515, 2.042, 1.714, 0.917, 1.929, 1.371])},\n 'smis': {'ei': ([0.760, 21.040, 38.120, 47.240, 56.560, 62.200],\n [0.196, 2.156, 1.237, 4.035, 1.641, 2.912]),\n 'greedy': ([0.800, 24.560, 43.240, 55.360, 63.960, 70.120],\n [0.253, 1.699, 1.335, 1.439, 1.718, 1.107]),\n 'pi': ([0.840, 22.640, 39.800, 49.360, 57.160, 63.080],\n [0.265, 2.381, 1.850, 6.306, 4.739, 3.530]),\n 'thompson': ([1.200, 20.560, 39.560, 53.000, 62.200, 68.920],\n [0.456, 2.214, 2.313, 1.720, 1.403, 2.286]),\n 'ucb': ([0.920, 21.640, 41.520, 54.280, 62.920, 70.040],\n [0.483, 2.221, 1.866, 1.204, 1.709, 1.189])}},\n 'rf': {'avg': {'ei': ([78.750, 91.025, 94.630, 96.294, 97.066, 97.625],\n [0.194, 1.128, 0.984, 0.699, 0.419, 0.193]),\n 'greedy': ([78.896, 93.759, 96.954, 97.977, 98.428, 98.744],\n [0.356, 0.460, 0.248, 0.173, 0.168, 0.154]),\n 'pi': ([78.797, 92.304, 95.889, 96.980, 97.623, 97.924],\n [0.245, 0.950, 0.263, 0.108, 0.136, 0.151]),\n 'thompson': ([78.623, 89.782, 93.667, 95.524, 96.717, 97.487],\n [0.529, 0.448, 0.268, 0.245, 0.202, 0.226]),\n 'ucb': ([78.800, 92.206, 95.625, 96.862, 97.713, 98.160],\n [0.498, 1.071, 0.428, 0.155, 0.121, 0.106])},\n 'scores': {'ei': ([0.439, 11.320, 21.000, 30.320, 36.160, 41.880],\n [0.837, 3.900, 5.547, 5.814, 4.869, 2.685]),\n 'greedy': ([0.762, 21.200, 37.640, 47.600, 53.600, 59.120],\n [0.389, 2.605, 2.949, 2.245, 2.577, 2.892]),\n 'pi': ([0.400, 14.720, 28.120, 34.960, 41.760, 45.480],\n [0.782, 3.971, 1.870, 1.127, 1.704, 2.355]),\n 'thompson': ([0.841, 7.240, 15.960, 24.320, 32.440, 39.840],\n [0.571, 1.317, 1.405, 2.328, 2.330, 2.949]),\n 'ucb': ([0.760, 14.680, 27.240, 35.160, 43.360, 49.040],\n [1.340, 4.135, 2.772, 1.359, 1.235, 1.394])},\n 'smis': {'ei': ([0.840, 11.040, 20.400, 29.040, 34.600, 40.120],\n [0.496, 3.908, 5.684, 5.946, 5.055, 2.700]),\n 'greedy': ([1.080, 20.080, 35.440, 44.520, 50.160, 55.080],\n [0.392, 2.551, 3.018, 2.560, 2.877, 2.968]),\n 'pi': ([0.880, 14.080, 26.960, 33.560, 40.040, 43.360],\n [0.574, 3.715, 1.822, 1.286, 1.592, 2.189]),\n 'thompson': ([1.160, 6.840, 15.040, 23.000, 30.760, 37.640],\n [0.196, 1.305, 1.607, 2.356, 2.368, 2.932]),\n 'ucb': ([1.120, 14.080, 26.200, 33.840, 41.760, 46.880],\n [0.826, 4.015, 2.647, 1.286, 0.814, 1.287])}}}", "_____no_output_____" ], [ "HTS_004_random = {\n 'avg': ([83.817, 86.459, 87.815, 88.779, 89.488, 90.094],\n [0.137, 0.061, 0.116, 0.112, 0.145, 0.147]),\n 'scores': ([0.440, 0.880, 1.300, 1.800, 2.220, 2.620],\n [0.215, 0.172, 0.219, 0.261, 0.232, 0.147]),\n 'smis': ([0.420, 0.860, 1.260, 1.700, 2.080, 2.420],\n [0.214, 0.174, 0.265, 0.290, 0.172, 0.075]),\n}", "_____no_output_____" ], [ "#top-1000\nHTS_004_online = {\n 'mpn': {'avg': {'ei': ([83.684, 99.075, 99.483, 99.628, 99.679, 99.712],\n [0.098, 0.046, 0.029, 0.024, 0.039, 0.049]),\n 'greedy': ([83.934, 99.126, 99.582, 99.758, 99.830, 99.871],\n [0.101, 0.073, 0.030, 0.028, 0.024, 0.016]),\n 'pi': ([83.815, 99.116, 99.525, 99.620, 99.669, 99.701],\n [0.247, 0.151, 0.052, 0.054, 0.041, 0.037]),\n 'thompson': ([83.902, 97.755, 99.391, 99.715, 99.822, 99.884],\n [0.106, 0.210, 0.066, 0.024, 0.007, 0.017]),\n 'ucb': ([83.781, 99.205, 99.678, 99.819, 99.874, 99.913],\n [0.060, 0.098, 0.037, 0.034, 0.018, 0.013])},\n 'scores': {'ei': ([0.500, 68.580, 81.040, 86.680, 88.880, 90.200],\n [0.276, 1.269, 1.454, 1.026, 1.452, 1.716]),\n 'greedy': ([0.480, 70.540, 86.540, 92.200, 94.140, 95.240],\n [0.172, 1.752, 0.983, 0.603, 0.595, 0.413]),\n 'pi': ([0.480, 70.440, 83.440, 87.620, 89.540, 90.540],\n [0.183, 3.158, 1.860, 2.040, 1.389, 1.106]),\n 'thompson': ([0.500, 42.840, 79.520, 91.260, 94.020, 95.780],\n [0.179, 3.456, 2.577, 0.768, 0.117, 0.306]),\n 'ucb': ([0.540, 72.380, 89.720, 93.960, 95.460, 96.680],\n [0.162, 2.153, 1.081, 0.831, 0.535, 0.387])},\n 'smis': {'ei': ([0.440, 65.120, 77.060, 82.380, 84.420, 85.720],\n [0.206, 1.315, 1.204, 0.924, 1.388, 1.683]),\n 'greedy': ([0.420, 67.000, 82.040, 88.560, 91.480, 93.200],\n [0.133, 1.664, 1.063, 1.080, 0.861, 0.540]),\n 'pi': ([0.420, 66.800, 79.340, 83.380, 85.300, 86.600],\n [0.160, 3.082, 1.640, 2.004, 1.463, 1.337]),\n 'thompson': ([0.500, 40.700, 75.560, 87.400, 91.480, 93.720],\n [0.179, 3.130, 2.613, 1.099, 0.515, 0.546]),\n 'ucb': ([0.480, 68.780, 85.220, 90.840, 93.000, 94.500],\n [0.194, 2.245, 1.237, 1.191, 0.687, 0.352])}},\n 'nn': {'avg': {'ei': ([83.784, 97.804, 97.898, 98.614, 98.821, 98.874],\n [0.276, 0.058, 0.067, 0.193, 0.267, 0.237]),\n 'greedy': ([83.707, 97.845, 99.122, 99.599, 99.701, 99.786],\n [0.062, 0.256, 0.172, 0.067, 0.046, 0.033]),\n 'pi': ([83.759, 97.862, 97.896, 98.340, 98.519, 98.583],\n [0.134, 0.068, 0.080, 0.233, 0.510, 0.552]),\n 'thompson': ([83.789, 96.998, 98.501, 98.907, 99.180, 99.275],\n [0.145, 0.110, 0.619, 0.388, 0.398, 0.399]),\n 'ucb': ([83.789, 97.953, 98.888, 99.354, 99.533, 99.693],\n [0.053, 0.167, 0.415, 0.191, 0.096, 0.032])},\n 'scores': {'ei': ([0.540, 43.420, 44.920, 58.800, 63.500, 64.940],\n [0.196, 0.808, 1.521, 4.155, 6.591, 5.569]),\n 'greedy': ([0.400, 45.760, 72.200, 87.480, 90.860, 93.020],\n [0.063, 4.216, 4.731, 2.044, 1.188, 0.842]),\n 'pi': ([0.320, 44.920, 45.500, 53.880, 58.420, 60.280],\n [0.172, 1.455, 1.784, 4.895, 12.256, 14.546]),\n 'thompson': ([0.440, 32.600, 57.960, 66.640, 74.600, 77.500],\n [0.233, 1.558, 11.872, 10.615, 11.430, 11.338]),\n 'ucb': ([0.380, 46.520, 65.280, 78.360, 84.400, 90.100],\n [0.223, 2.727, 9.609, 6.412, 3.852, 1.056])},\n 'smis': {'ei': ([0.540, 41.320, 42.820, 55.640, 60.220, 61.520],\n [0.196, 0.631, 1.359, 3.723, 6.355, 5.538]),\n 'greedy': ([0.380, 43.340, 68.300, 82.860, 86.860, 89.760],\n [0.075, 3.975, 4.689, 2.258, 1.513, 0.954]),\n 'pi': ([0.300, 42.620, 43.160, 50.800, 55.200, 57.040],\n [0.190, 1.134, 1.604, 4.689, 11.611, 13.848]),\n 'thompson': ([0.400, 31.020, 54.960, 63.260, 70.680, 73.380],\n [0.253, 1.373, 11.216, 9.827, 10.458, 10.548]),\n 'ucb': ([0.360, 44.160, 62.320, 74.260, 80.180, 86.080],\n [0.185, 2.548, 9.264, 6.075, 3.677, 1.607])}},\n 'rf': {'avg': {'ei': ([83.809, 95.565, 96.119, 97.158, 97.618, 97.927],\n [0.109, 0.474, 0.711, 0.613, 0.324, 0.098]),\n 'greedy': ([83.846, 97.655, 98.781, 99.146, 99.346, 99.455],\n [0.152, 0.139, 0.107, 0.141, 0.083, 0.050]),\n 'pi': ([83.904, 96.265, 96.694, 97.153, 97.593, 97.795],\n [0.134, 0.646, 0.646, 0.574, 0.484, 0.410]),\n 'thompson': ([83.830, 95.328, 97.106, 98.056, 98.502, 98.694],\n [0.182, 0.240, 0.131, 0.110, 0.115, 0.071]),\n 'ucb': ([83.776, 96.838, 97.310, 98.029, 98.443, 98.589],\n [0.195, 0.375, 0.540, 0.348, 0.115, 0.137])},\n 'scores': {'ei': ([0.460, 21.540, 26.260, 35.580, 40.640, 44.980],\n [0.102, 3.571, 6.112, 6.739, 4.589, 1.759]),\n 'greedy': ([0.420, 40.640, 61.000, 70.380, 76.240, 80.620],\n [0.160, 1.796, 2.791, 3.816, 3.090, 2.268]),\n 'pi': ([0.260, 26.920, 30.640, 35.260, 40.520, 43.180],\n [0.102, 5.416, 6.246, 6.260, 5.662, 5.568]),\n 'thompson': ([0.380, 19.380, 35.080, 47.660, 55.380, 59.980],\n [0.117, 1.408, 1.174, 2.461, 2.475, 2.025]),\n 'ucb': ([0.380, 31.440, 37.160, 46.460, 53.540, 56.360],\n [0.204, 3.910, 6.637, 6.309, 2.153, 2.579])},\n 'smis': {'ei': ([0.440, 20.780, 25.380, 34.300, 39.300, 43.460],\n [0.102, 3.498, 6.019, 6.710, 4.732, 1.911]),\n 'greedy': ([0.400, 38.700, 58.260, 67.140, 72.520, 76.480],\n [0.179, 1.628, 2.743, 3.781, 2.815, 2.060]),\n 'pi': ([0.240, 26.020, 29.660, 34.100, 39.180, 41.760],\n [0.102, 5.202, 6.021, 6.005, 5.364, 5.324]),\n 'thompson': ([0.340, 18.540, 33.600, 45.480, 52.860, 57.100],\n [0.185, 1.476, 1.170, 2.307, 2.331, 1.833]),\n 'ucb': ([0.360, 30.280, 35.760, 44.580, 51.260, 53.980],\n [0.174, 3.869, 6.344, 5.977, 1.918, 2.374])}}}", "_____no_output_____" ], [ "HTS_004_retrain = {\n 'mpn': {'avg': {'ei': ([83.904, 99.143, 99.633, 99.791, 99.855, 99.891],\n [0.087, 0.113, 0.050, 0.046, 0.025, 0.017]),\n 'greedy': ([83.858, 99.129, 99.688, 99.862, 99.921, 99.941],\n [0.239, 0.107, 0.023, 0.008, 0.008, 0.005]),\n 'pi': ([83.788, 99.074, 99.670, 99.807, 99.864, 99.904],\n [0.164, 0.210, 0.020, 0.018, 0.019, 0.017]),\n 'thompson': ([83.788, 97.784, 99.410, 99.730, 99.869, 99.920],\n [0.070, 0.356, 0.032, 0.018, 0.022, 0.013]),\n 'ucb': ([83.813, 99.217, 99.757, 99.885, 99.926, 99.944],\n [0.065, 0.065, 0.035, 0.012, 0.013, 0.013])},\n 'scores': {'ei': ([0.460, 70.260, 86.960, 92.900, 94.840, 95.820],\n [0.206, 2.702, 1.803, 1.431, 0.768, 0.614]),\n 'greedy': ([0.460, 70.840, 90.480, 94.960, 96.700, 97.680],\n [0.120, 2.351, 0.652, 0.242, 0.290, 0.232]),\n 'pi': ([0.380, 68.900, 88.720, 93.300, 94.920, 96.160],\n [0.172, 4.281, 1.144, 0.569, 0.546, 0.575]),\n 'thompson': ([0.320, 43.020, 80.680, 91.660, 95.160, 96.740],\n [0.147, 5.601, 1.434, 0.516, 0.496, 0.338]),\n 'ucb': ([0.460, 72.160, 92.120, 95.660, 97.220, 98.080],\n [0.174, 1.204, 1.034, 0.436, 0.462, 0.421])},\n 'smis': {'ei': ([0.440, 67.000, 82.920, 89.540, 92.300, 93.660],\n [0.206, 2.452, 1.589, 2.083, 1.105, 0.671]),\n 'greedy': ([0.400, 67.300, 86.140, 92.300, 94.380, 94.760],\n [0.089, 2.348, 0.736, 0.303, 0.098, 0.136]),\n 'pi': ([0.380, 65.440, 84.200, 89.720, 92.220, 93.940],\n [0.172, 4.109, 0.851, 0.995, 0.933, 0.709]),\n 'thompson': ([0.300, 40.760, 76.340, 87.980, 92.780, 94.500],\n [0.126, 5.166, 1.493, 0.760, 0.730, 0.126]),\n 'ucb': ([0.440, 68.560, 88.100, 93.240, 94.620, 94.880],\n [0.185, 1.244, 1.114, 0.233, 0.160, 0.172])}},\n 'nn': {'avg': {'ei': ([83.720, 97.690, 98.626, 98.796, 99.139, 99.327],\n [0.097, 0.089, 0.049, 0.056, 0.028, 0.048]),\n 'greedy': ([83.851, 97.667, 99.174, 99.598, 99.813, 99.894],\n [0.081, 0.207, 0.080, 0.016, 0.002, 0.002]),\n 'pi': ([83.843, 97.654, 98.284, 98.661, 98.961, 99.165],\n [0.115, 0.176, 0.220, 0.196, 0.124, 0.071]),\n 'thompson': ([83.851, 97.349, 99.056, 99.580, 99.763, 99.847],\n [0.035, 0.193, 0.106, 0.047, 0.007, 0.011]),\n 'ucb': ([83.838, 98.043, 99.259, 99.618, 99.741, 99.843],\n [0.117, 0.138, 0.040, 0.029, 0.038, 0.017])},\n 'scores': {'ei': ([0.460, 42.160, 59.000, 62.700, 70.767, 76.700],\n [0.102, 1.428, 0.712, 0.993, 1.112, 2.304]),\n 'greedy': ([0.320, 42.180, 74.175, 88.200, 93.550, 95.700],\n [0.194, 3.384, 1.875, 0.374, 0.050, 0.000]),\n 'pi': ([0.420, 41.600, 52.325, 60.100, 67.100, 71.875],\n [0.160, 2.642, 4.471, 4.566, 3.237, 1.906]),\n 'thompson': ([0.520, 36.840, 68.780, 86.575, 92.467, 94.533],\n [0.172, 2.825, 2.908, 2.200, 0.464, 0.249]),\n 'ucb': ([0.420, 48.100, 75.240, 88.460, 91.780, 94.380],\n [0.075, 2.396, 1.109, 0.905, 0.983, 0.471])},\n 'smis': {'ei': ([0.420, 39.960, 56.067, 59.600, 67.333, 72.900],\n [0.117, 1.363, 0.713, 0.993, 0.988, 2.223]),\n 'greedy': ([0.320, 40.180, 70.225, 83.375, 90.350, 93.150],\n [0.194, 3.020, 1.895, 0.630, 0.150, 0.050]),\n 'pi': ([0.340, 39.500, 49.850, 57.350, 64.000, 68.500],\n [0.120, 2.551, 4.275, 4.542, 3.163, 1.826]),\n 'thompson': ([0.500, 35.140, 65.340, 81.925, 88.500, 91.567],\n [0.179, 2.932, 2.851, 2.013, 0.616, 0.287]),\n 'ucb': ([0.400, 45.700, 71.260, 83.560, 87.920, 91.480],\n [0.089, 2.293, 1.007, 0.950, 1.451, 0.783])}},\n 'rf': {'avg': {'ei': ([83.961, 96.632, 97.269, 97.540, 97.751, 97.905],\n [0.090, 0.565, 0.475, 0.363, 0.272, 0.263]),\n 'greedy': ([83.898, 97.481, 98.662, 99.215, 99.414, 99.532],\n [0.066, 0.332, 0.078, 0.060, 0.047, 0.024]),\n 'pi': ([83.836, 96.361, 97.050, 97.435, 97.645, 97.798],\n [0.143, 0.725, 0.379, 0.182, 0.192, 0.184]),\n 'thompson': ([83.842, 95.147, 97.833, 98.700, 99.055, 99.256],\n [0.145, 0.439, 0.108, 0.034, 0.041, 0.035]),\n 'ucb': ([83.876, 96.993, 98.350, 98.723, 98.884, 99.033],\n [0.197, 0.393, 0.171, 0.158, 0.137, 0.100])},\n 'scores': {'ei': ([0.680, 30.020, 36.180, 39.660, 42.420, 44.840],\n [0.172, 5.605, 5.639, 4.786, 3.899, 3.979]),\n 'greedy': ([0.380, 38.380, 58.060, 72.000, 79.420, 84.300],\n [0.117, 4.407, 2.170, 1.601, 1.607, 1.108]),\n 'pi': ([0.300, 28.920, 34.640, 38.860, 41.440, 43.500],\n [0.141, 5.284, 3.733, 2.217, 2.391, 2.601]),\n 'thompson': ([0.420, 18.860, 43.260, 59.740, 68.300, 74.080],\n [0.117, 3.526, 1.904, 0.796, 0.974, 1.040]),\n 'ucb': ([0.300, 33.220, 52.220, 60.080, 64.060, 68.160],\n [0.167, 4.175, 3.466, 4.154, 3.682, 2.738])},\n 'smis': {'ei': ([0.640, 29.040, 34.940, 38.300, 40.940, 43.220],\n [0.162, 5.434, 5.494, 4.627, 3.815, 3.860]),\n 'greedy': ([0.340, 36.700, 55.400, 68.520, 75.340, 79.800],\n [0.102, 4.283, 1.869, 1.474, 1.553, 0.867]),\n 'pi': ([0.280, 28.000, 33.660, 37.700, 40.200, 42.200],\n [0.147, 5.164, 3.697, 2.136, 2.393, 2.668]),\n 'thompson': ([0.380, 18.160, 41.340, 56.840, 64.920, 70.340],\n [0.160, 3.615, 1.930, 0.907, 1.148, 1.222]),\n 'ucb': ([0.300, 32.120, 50.380, 57.680, 61.380, 65.160],\n [0.167, 4.023, 3.323, 3.999, 3.490, 2.628])}}}", "_____no_output_____" ], [ "HTS_002_random = {\n 'avg': ([80.697, 83.715, 85.348, 86.373, 87.173, 87.751],\n [0.234, 0.153, 0.163, 0.179, 0.175, 0.144]),\n 'scores': ([0.120, 0.340, 0.660, 0.840, 1.140, 1.340],\n [0.075, 0.102, 0.102, 0.185, 0.301, 0.383]),\n 'smis': ([0.100, 0.320, 0.640, 0.800, 1.100, 1.300],\n [0.063, 0.098, 0.120, 0.155, 0.261, 0.341]),\n}", "_____no_output_____" ], [ "HTS_002_online = {\n 'mpn': {'avg': {'ei': ([80.955, 98.023, 99.010, 99.327, 99.417, 99.484],\n [0.165, 0.166, 0.087, 0.103, 0.082, 0.074]),\n 'greedy': ([80.828, 97.961, 98.929, 99.364, 99.522, 99.619],\n [0.091, 0.206, 0.060, 0.052, 0.028, 0.043]),\n 'pi': ([80.777, 98.000, 98.989, 99.277, 99.376, 99.413],\n [0.195, 0.243, 0.094, 0.084, 0.087, 0.084]),\n 'thompson': ([80.899, 95.969, 98.664, 99.335, 99.527, 99.646],\n [0.265, 0.255, 0.138, 0.047, 0.026, 0.013]),\n 'ucb': ([80.732, 98.118, 99.030, 99.444, 99.560, 99.665],\n [0.228, 0.117, 0.106, 0.069, 0.044, 0.016])},\n 'scores': {'ei': ([0.220, 45.800, 65.220, 74.560, 77.780, 80.520],\n [0.133, 2.781, 2.374, 3.102, 3.062, 3.081]),\n 'greedy': ([0.120, 46.280, 65.700, 79.000, 85.040, 88.740],\n [0.075, 3.178, 1.431, 1.519, 0.723, 1.306]),\n 'pi': ([0.180, 46.160, 65.620, 72.700, 76.200, 77.760],\n [0.117, 3.992, 2.035, 2.373, 2.768, 2.826]),\n 'thompson': ([0.320, 25.060, 59.180, 77.300, 84.840, 89.280],\n [0.183, 1.810, 2.112, 1.119, 0.869, 0.299]),\n 'ucb': ([0.180, 48.420, 67.040, 80.720, 85.740, 89.500],\n [0.075, 1.901, 2.284, 2.248, 1.493, 0.341])},\n 'smis': {'ei': ([0.220, 44.060, 62.600, 71.340, 74.280, 76.800],\n [0.133, 2.619, 2.224, 2.895, 2.948, 2.887]),\n 'greedy': ([0.120, 44.340, 62.600, 75.060, 80.640, 84.120],\n [0.075, 2.889, 1.178, 1.311, 0.695, 1.477]),\n 'pi': ([0.180, 44.440, 62.860, 69.620, 72.900, 74.300],\n [0.117, 3.811, 1.946, 2.176, 2.651, 2.653]),\n 'thompson': ([0.300, 23.960, 56.420, 73.500, 80.520, 84.860],\n [0.167, 1.765, 2.007, 1.071, 0.652, 0.393]),\n 'ucb': ([0.160, 46.380, 64.040, 76.760, 81.420, 85.080],\n [0.080, 1.604, 2.095, 2.201, 1.379, 0.564])}},\n 'nn': {'avg': {'ei': ([80.669, 96.461, 96.763, 97.207, 97.426, 97.494],\n [0.070, 0.091, 0.430, 0.890, 0.753, 0.795]),\n 'greedy': ([80.792, 96.592, 98.399, 99.082, 99.370, 99.473],\n [0.190, 0.522, 0.277, 0.109, 0.033, 0.028]),\n 'pi': ([80.799, 96.474, 96.690, 97.178, 97.315, 97.458],\n [0.139, 0.161, 0.302, 0.651, 0.558, 0.484]),\n 'thompson': ([80.914, 95.463, 97.716, 98.516, 98.967, 99.189],\n [0.144, 0.079, 0.758, 0.390, 0.253, 0.145]),\n 'ucb': ([80.666, 96.664, 98.169, 98.655, 99.033, 99.132],\n [0.158, 0.253, 0.628, 0.581, 0.325, 0.380])},\n 'scores': {'ei': ([0.160, 28.560, 31.840, 37.980, 39.820, 40.940],\n [0.049, 1.122, 4.955, 12.077, 10.884, 11.675]),\n 'greedy': ([0.200, 29.720, 54.240, 69.540, 78.200, 82.200],\n [0.110, 5.704, 4.350, 2.277, 0.901, 0.762]),\n 'pi': ([0.180, 28.420, 30.740, 36.460, 37.720, 39.320],\n [0.117, 1.342, 3.409, 8.817, 7.923, 7.133]),\n 'thompson': ([0.280, 20.600, 43.320, 56.540, 66.480, 72.480],\n [0.183, 0.648, 9.902, 7.893, 6.285, 4.160]),\n 'ucb': ([0.220, 30.360, 50.260, 60.000, 67.740, 70.720],\n [0.075, 3.484, 8.661, 9.605, 6.738, 8.550])},\n 'smis': {'ei': ([0.140, 27.100, 30.360, 36.080, 37.840, 38.920],\n [0.049, 0.914, 4.894, 11.509, 10.336, 11.124]),\n 'greedy': ([0.200, 28.140, 51.720, 66.340, 74.380, 78.060],\n [0.110, 5.412, 4.293, 2.104, 0.877, 0.605]),\n 'pi': ([0.180, 27.020, 29.280, 34.700, 35.860, 37.380],\n [0.117, 1.335, 3.323, 8.474, 7.672, 6.940]),\n 'thompson': ([0.240, 19.620, 41.460, 54.040, 63.460, 68.920],\n [0.196, 0.435, 9.593, 7.510, 5.951, 3.854]),\n 'ucb': ([0.220, 29.020, 48.540, 57.800, 64.980, 67.860],\n [0.075, 3.482, 8.569, 9.421, 6.631, 8.332])}},\n 'rf': {'avg': {'ei': ([80.879, 94.098, 94.503, 95.673, 96.228, 96.462],\n [0.133, 0.810, 0.549, 0.501, 0.638, 0.748]),\n 'greedy': ([80.919, 96.169, 97.852, 98.565, 98.848, 99.012],\n [0.170, 0.563, 0.377, 0.019, 0.090, 0.111]),\n 'pi': ([80.847, 94.119, 94.628, 95.738, 96.478, 96.793],\n [0.101, 1.029, 0.985, 0.918, 0.629, 0.600]),\n 'thompson': ([80.707, 92.112, 94.474, 96.130, 97.011, 97.386],\n [0.157, 0.195, 0.783, 0.684, 0.432, 0.312]),\n 'ucb': ([80.739, 93.947, 95.344, 96.841, 97.265, 97.938],\n [0.177, 3.115, 1.934, 0.972, 0.678, 0.092])},\n 'scores': {'ei': ([0.220, 15.940, 17.560, 24.080, 27.960, 30.020],\n [0.147, 4.237, 3.593, 3.756, 5.407, 7.074]),\n 'greedy': ([0.180, 26.780, 45.000, 56.200, 62.780, 66.920],\n [0.117, 4.782, 5.383, 0.438, 2.155, 2.403]),\n 'pi': ([0.120, 15.260, 17.980, 23.900, 29.360, 32.300],\n [0.098, 5.091, 4.810, 5.677, 4.980, 5.495]),\n 'thompson': ([0.200, 8.080, 17.400, 26.960, 34.800, 38.460],\n [0.089, 0.770, 4.023, 5.549, 4.827, 4.257]),\n 'ucb': ([0.280, 18.680, 23.980, 33.360, 37.420, 45.760],\n [0.133, 8.614, 9.668, 9.174, 7.951, 1.551])},\n 'smis': {'ei': ([0.220, 15.420, 16.980, 23.240, 26.960, 28.960],\n [0.147, 4.304, 3.711, 3.821, 5.478, 7.107]),\n 'greedy': ([0.180, 25.540, 43.240, 53.740, 60.080, 64.000],\n [0.117, 4.558, 5.205, 0.546, 2.218, 2.370]),\n 'pi': ([0.100, 14.600, 17.320, 22.940, 28.260, 31.100],\n [0.089, 5.004, 4.667, 5.295, 4.738, 5.187]),\n 'thompson': ([0.180, 7.680, 16.700, 25.820, 33.420, 36.940],\n [0.075, 0.747, 3.888, 5.348, 4.574, 3.999]),\n 'ucb': ([0.260, 18.040, 23.200, 32.240, 36.080, 44.080],\n [0.136, 8.520, 9.573, 9.059, 7.832, 1.427])}}}", "_____no_output_____" ], [ "HTS_002_retrain = {\n 'mpn': {'avg': {'ei': ([80.939, 97.785, 99.102, 99.521, 99.655, 99.744],\n [0.252, 0.273, 0.059, 0.021, 0.026, 0.014]),\n 'greedy': ([80.812, 98.144, 99.252, 99.566, 99.717, 99.814],\n [0.130, 0.032, 0.057, 0.063, 0.043, 0.035]),\n 'pi': ([80.854, 97.914, 99.211, 99.550, 99.675, 99.759],\n [0.180, 0.191, 0.092, 0.028, 0.019, 0.027]),\n 'thompson': ([80.785, 95.529, 98.555, 99.278, 99.550, 99.714],\n [0.166, 0.962, 0.163, 0.033, 0.028, 0.014]),\n 'ucb': ([80.823, 98.160, 99.313, 99.634, 99.761, 99.836],\n [0.134, 0.192, 0.037, 0.017, 0.014, 0.015])},\n 'scores': {'ei': ([0.200, 43.380, 68.420, 82.060, 88.060, 91.680],\n [0.063, 4.464, 1.388, 0.977, 1.076, 0.662]),\n 'greedy': ([0.220, 49.280, 73.940, 86.280, 91.180, 93.700],\n [0.040, 0.571, 1.630, 1.885, 1.105, 0.892]),\n 'pi': ([0.220, 44.900, 70.720, 82.900, 88.280, 92.100],\n [0.075, 2.575, 2.507, 1.495, 1.034, 0.707]),\n 'thompson': ([0.160, 23.300, 56.840, 75.140, 85.600, 90.760],\n [0.102, 5.548, 2.957, 1.125, 0.844, 0.418]),\n 'ucb': ([0.220, 49.160, 74.980, 87.680, 92.200, 94.260],\n [0.075, 3.230, 1.254, 0.556, 0.456, 0.393])},\n 'smis': {'ei': ([0.180, 41.700, 65.360, 78.340, 83.760, 87.380],\n [0.075, 4.346, 1.392, 1.001, 1.177, 0.655]),\n 'greedy': ([0.220, 47.240, 70.320, 81.760, 87.240, 90.720],\n [0.040, 0.615, 1.620, 1.940, 1.439, 0.950]),\n 'pi': ([0.200, 43.300, 67.920, 79.220, 84.160, 87.960],\n [0.089, 2.377, 2.411, 1.399, 1.009, 0.931]),\n 'thompson': ([0.160, 22.220, 54.300, 71.340, 81.060, 86.740],\n [0.102, 5.284, 2.669, 1.237, 0.723, 0.554]),\n 'ucb': ([0.220, 47.120, 71.520, 83.360, 88.520, 91.560],\n [0.075, 3.191, 1.238, 0.320, 0.504, 0.476])}},\n 'nn': {'avg': {'ei': ([80.988, 96.399, 97.699, 98.080, 98.414, 98.537],\n [0.103, 0.202, 0.226, 0.238, 0.238, 0.228]),\n 'greedy': ([80.730, 96.734, 98.570, 99.132, 99.447, 99.627],\n [0.120, 0.258, 0.110, 0.084, 0.025, 0.027]),\n 'pi': ([80.920, 96.366, 97.548, 98.034, 98.411, 98.674],\n [0.124, 0.229, 0.280, 0.160, 0.178, 0.129]),\n 'thompson': ([80.778, 95.606, 98.015, 98.892, 99.359, 99.532],\n [0.194, 0.215, 0.102, 0.066, 0.024, 0.026]),\n 'ucb': ([81.034, 96.737, 98.624, 99.203, 99.463, 99.586],\n [0.134, 0.208, 0.121, 0.076, 0.057, 0.015])},\n 'scores': {'ei': ([0.260, 27.560, 42.180, 48.120, 54.140, 56.640],\n [0.196, 2.001, 3.159, 3.610, 4.268, 4.340]),\n 'greedy': ([0.300, 31.020, 58.380, 72.180, 82.600, 88.800],\n [0.228, 2.805, 2.262, 2.009, 0.607, 0.751]),\n 'pi': ([0.260, 27.920, 40.300, 47.460, 53.560, 59.120],\n [0.162, 2.212, 3.423, 2.451, 2.899, 3.115]),\n 'thompson': ([0.240, 21.580, 47.140, 65.300, 78.420, 85.040],\n [0.150, 1.777, 1.617, 1.962, 1.333, 0.926]),\n 'ucb': ([0.200, 31.200, 58.640, 72.900, 81.940, 86.740],\n [0.126, 2.427, 2.127, 1.874, 1.975, 0.543])},\n 'smis': {'ei': ([0.260, 26.340, 40.380, 46.080, 51.700, 54.020],\n [0.196, 2.150, 3.338, 3.633, 4.026, 4.074]),\n 'greedy': ([0.280, 29.400, 55.460, 68.440, 78.020, 83.920],\n [0.223, 2.630, 2.310, 1.732, 0.511, 0.776]),\n 'pi': ([0.220, 26.560, 38.420, 45.300, 51.240, 56.600],\n [0.172, 2.133, 3.101, 2.195, 2.681, 3.027]),\n 'thompson': ([0.240, 20.500, 44.940, 61.840, 74.200, 80.360],\n [0.150, 1.872, 1.416, 2.154, 1.252, 0.887]),\n 'ucb': ([0.200, 29.720, 55.980, 69.480, 77.880, 82.140],\n [0.126, 2.404, 2.113, 1.712, 1.766, 0.561])}},\n 'rf': {'avg': {'ei': ([80.728, 94.054, 95.671, 96.353, 96.601, 96.865],\n [0.150, 0.688, 0.552, 0.349, 0.376, 0.281]),\n 'greedy': ([80.845, 95.389, 97.640, 98.471, 98.971, 99.231],\n [0.146, 0.679, 0.287, 0.184, 0.104, 0.075]),\n 'pi': ([80.847, 94.176, 95.481, 96.188, 96.347, 96.536],\n [0.097, 0.527, 0.776, 0.634, 0.584, 0.522]),\n 'thompson': ([80.946, 92.597, 96.079, 97.651, 98.230, 98.599],\n [0.177, 0.375, 0.300, 0.059, 0.063, 0.048]),\n 'ucb': ([80.848, 94.674, 97.064, 97.754, 98.019, 98.253],\n [0.110, 0.521, 0.137, 0.092, 0.143, 0.151])},\n 'scores': {'ei': ([0.160, 15.440, 23.760, 28.480, 30.620, 32.560],\n [0.120, 3.469, 5.234, 3.417, 3.804, 3.064]),\n 'greedy': ([0.180, 20.940, 40.440, 54.280, 65.620, 72.280],\n [0.075, 4.852, 3.657, 3.396, 2.386, 1.912]),\n 'pi': ([0.220, 15.740, 21.960, 26.760, 27.580, 29.280],\n [0.075, 2.024, 4.126, 5.073, 5.028, 5.150]),\n 'thompson': ([0.340, 9.360, 26.020, 41.640, 50.580, 57.480],\n [0.102, 1.286, 2.322, 0.731, 1.237, 1.353]),\n 'ucb': ([0.240, 18.800, 34.760, 43.080, 46.980, 50.980],\n [0.102, 3.536, 1.499, 1.248, 2.698, 2.870])},\n 'smis': {'ei': ([0.140, 14.740, 22.800, 27.360, 29.380, 31.280],\n [0.102, 3.495, 5.246, 3.402, 3.771, 3.024]),\n 'greedy': ([0.180, 20.180, 38.800, 51.920, 62.720, 69.000],\n [0.075, 4.710, 3.538, 3.276, 2.415, 1.889]),\n 'pi': ([0.220, 15.160, 21.240, 25.920, 26.740, 28.340],\n [0.075, 1.712, 3.859, 4.882, 4.843, 4.987]),\n 'thompson': ([0.340, 8.940, 25.080, 40.000, 48.440, 54.840],\n [0.102, 1.157, 2.203, 0.994, 1.375, 1.488]),\n 'ucb': ([0.220, 18.260, 33.560, 41.420, 45.140, 48.860],\n [0.098, 3.729, 1.661, 1.280, 2.818, 2.892])}}}", "_____no_output_____" ], [ "HTS_001_random = {\n 'avg': ([77.065, 80.761, 82.523, 83.792, 84.702, 85.363],\n [0.203, 0.176, 0.119, 0.111, 0.111, 0.108]),\n 'scores': ([0.040, 0.120, 0.200, 0.300, 0.440, 0.600],\n [0.080, 0.147, 0.110, 0.200, 0.206, 0.219]),\n 'smis': ([0.040, 0.120, 0.180, 0.240, 0.380, 0.520],\n [0.080, 0.147, 0.117, 0.136, 0.172, 0.172]),\n}", "_____no_output_____" ], [ "HTS_001_online = {'mpn': {'avg': {'ei': ([77.011, 95.719, 97.870, 98.487, 98.804, 98.951],\n [0.322, 0.385, 0.107, 0.105, 0.049, 0.061]),\n 'greedy': ([77.119, 95.799, 97.538, 98.648, 98.920, 99.168],\n [0.170, 0.325, 0.241, 0.134, 0.181, 0.114]),\n 'pi': ([76.971, 96.051, 97.963, 98.569, 98.829, 98.972],\n [0.173, 0.363, 0.151, 0.075, 0.086, 0.092]),\n 'thompson': ([76.989, 93.091, 97.319, 98.531, 98.990, 99.241],\n [0.127, 0.342, 0.215, 0.149, 0.094, 0.034]),\n 'ucb': ([77.228, 96.095, 98.084, 98.855, 99.093, 99.327],\n [0.071, 0.250, 0.192, 0.058, 0.073, 0.053])},\n 'scores': {'ei': ([0.140, 26.500, 44.200, 53.640, 59.940, 63.800],\n [0.049, 2.754, 1.472, 1.879, 1.237, 1.374]),\n 'greedy': ([0.080, 26.580, 40.760, 58.380, 65.260, 71.900],\n [0.075, 2.431, 2.600, 2.795, 4.731, 3.341]),\n 'pi': ([0.060, 29.180, 45.760, 55.240, 60.740, 64.340],\n [0.120, 2.161, 2.019, 1.120, 1.909, 1.768]),\n 'thompson': ([0.020, 12.600, 38.020, 56.140, 66.640, 73.340],\n [0.040, 1.568, 2.121, 2.667, 2.043, 0.960]),\n 'ucb': ([0.080, 28.940, 47.700, 62.240, 68.740, 75.780],\n [0.075, 1.714, 2.711, 1.479, 1.995, 2.118])},\n 'smis': {'ei': ([0.140, 25.800, 42.580, 51.540, 57.620, 61.320],\n [0.049, 2.725, 1.422, 2.058, 1.359, 1.541]),\n 'greedy': ([0.080, 25.700, 39.260, 55.860, 62.260, 68.400],\n [0.075, 2.347, 2.418, 2.617, 4.408, 3.455]),\n 'pi': ([0.060, 28.320, 44.080, 53.240, 58.500, 61.920],\n [0.120, 2.130, 2.060, 1.098, 1.806, 1.775]),\n 'thompson': ([0.020, 12.280, 36.500, 53.720, 63.700, 70.040],\n [0.040, 1.607, 2.013, 2.552, 2.131, 1.227]),\n 'ucb': ([0.080, 28.260, 45.980, 59.680, 65.900, 72.360],\n [0.075, 1.664, 2.627, 1.393, 1.925, 1.804])}},\n 'nn': {'avg': {'ei': ([77.223, 94.523, 94.856, 95.303, 95.696, 96.225],\n [0.242, 0.313, 0.811, 1.265, 1.045, 0.978]),\n 'greedy': ([77.295, 94.600, 97.088, 98.146, 98.702, 98.956],\n [0.293, 0.494, 0.283, 0.147, 0.079, 0.087]),\n 'pi': ([77.224, 94.173, 95.544, 97.069, 97.277, 97.387],\n [0.142, 0.196, 0.636, 0.340, 0.179, 0.299]),\n 'thompson': ([77.048, 93.544, 95.970, 96.949, 97.639, 97.898],\n [0.180, 0.305, 1.197, 1.009, 0.531, 0.619]),\n 'ucb': ([76.843, 94.479, 96.181, 96.965, 97.413, 97.903],\n [0.188, 0.436, 1.259, 1.119, 0.994, 0.624])},\n 'scores': {'ei': ([0.060, 17.320, 19.520, 22.480, 23.940, 27.400],\n [0.049, 2.359, 5.685, 9.806, 9.066, 9.594]),\n 'greedy': ([0.160, 17.740, 35.400, 49.220, 59.540, 65.840],\n [0.080, 3.028, 3.051, 2.116, 1.714, 2.226]),\n 'pi': ([0.180, 16.100, 23.960, 35.240, 37.340, 38.800],\n [0.133, 0.853, 3.759, 3.433, 2.030, 3.569]),\n 'thompson': ([0.180, 13.580, 27.520, 35.360, 42.140, 46.100],\n [0.117, 1.250, 6.867, 8.682, 7.335, 9.103]),\n 'ucb': ([0.040, 17.700, 29.480, 35.940, 40.740, 45.780],\n [0.049, 2.784, 8.671, 10.293, 11.121, 8.874])},\n 'smis': {'ei': ([0.060, 16.520, 18.680, 21.540, 22.880, 26.180],\n [0.049, 2.301, 5.470, 9.528, 8.871, 9.425]),\n 'greedy': ([0.140, 16.780, 33.860, 47.220, 57.020, 63.180],\n [0.080, 2.806, 2.958, 2.167, 1.577, 1.951]),\n 'pi': ([0.160, 15.300, 22.980, 33.860, 35.880, 37.320],\n [0.136, 0.927, 3.645, 3.262, 1.790, 3.337]),\n 'thompson': ([0.160, 12.780, 26.400, 33.860, 40.500, 44.300],\n [0.102, 1.329, 6.647, 8.350, 7.025, 8.794]),\n 'ucb': ([0.040, 16.940, 28.360, 34.480, 39.140, 43.860],\n [0.049, 2.750, 8.468, 10.064, 10.847, 8.647])}},\n 'rf': {'avg': {'ei': ([77.188, 90.755, 91.591, 93.206, 94.276, 94.899],\n [0.121, 1.514, 1.641, 2.137, 1.553, 1.293]),\n 'greedy': ([77.047, 93.235, 95.601, 96.695, 97.141, 97.598],\n [0.115, 0.876, 0.358, 0.323, 0.308, 0.488]),\n 'pi': ([77.085, 91.782, 93.112, 94.547, 95.315, 96.054],\n [0.181, 0.882, 0.868, 0.658, 1.139, 0.852]),\n 'thompson': ([77.190, 88.720, 91.614, 93.629, 94.869, 95.634],\n [0.179, 0.460, 0.959, 0.822, 0.369, 0.329]),\n 'ucb': ([77.003, 91.568, 93.413, 94.253, 95.221, 95.837],\n [0.185, 1.298, 1.391, 1.428, 1.016, 0.906])},\n 'scores': {'ei': ([0.140, 7.760, 9.700, 14.340, 17.880, 20.200],\n [0.102, 3.388, 4.380, 6.545, 5.657, 5.511]),\n 'greedy': ([0.020, 13.200, 24.240, 31.340, 35.380, 40.980],\n [0.040, 3.514, 3.463, 3.787, 3.727, 6.469]),\n 'pi': ([0.140, 10.020, 13.680, 18.880, 22.940, 27.120],\n [0.136, 2.641, 2.846, 3.387, 6.511, 5.721]),\n 'thompson': ([0.160, 3.860, 9.680, 15.380, 20.380, 24.120],\n [0.120, 0.786, 2.691, 3.413, 1.957, 1.629]),\n 'ucb': ([0.080, 10.000, 15.740, 18.500, 22.580, 26.240],\n [0.040, 4.964, 5.188, 5.964, 6.033, 6.776])},\n 'smis': {'ei': ([0.140, 7.440, 9.340, 13.920, 17.380, 19.580],\n [0.102, 3.335, 4.309, 6.387, 5.450, 5.247]),\n 'greedy': ([0.020, 12.500, 23.360, 30.120, 34.100, 39.480],\n [0.040, 3.416, 3.539, 3.868, 3.597, 6.155]),\n 'pi': ([0.140, 9.560, 13.140, 18.300, 22.320, 26.400],\n [0.136, 2.546, 2.773, 3.372, 6.471, 5.628]),\n 'thompson': ([0.120, 3.680, 9.260, 14.800, 19.560, 23.120],\n [0.098, 0.763, 2.515, 3.247, 1.847, 1.522]),\n 'ucb': ([0.080, 9.580, 15.140, 17.820, 21.860, 25.420],\n [0.040, 4.803, 5.095, 5.837, 5.854, 6.569])}}}", "_____no_output_____" ], [ "HTS_001_retrain = {\n 'mpn': {'avg': {'ei': ([77.087, 95.947, 97.954, 98.876, 99.213, 99.423],\n [0.138, 0.431, 0.192, 0.055, 0.064, 0.026]),\n 'greedy': ([77.241, 96.071, 98.266, 98.916, 99.257, 99.470],\n [0.117, 0.270, 0.115, 0.062, 0.022, 0.030]),\n 'pi': ([77.132, 95.993, 98.123, 98.830, 99.187, 99.407],\n [0.158, 0.593, 0.093, 0.051, 0.041, 0.016]),\n 'thompson': ([76.906, 93.051, 97.070, 98.357, 98.923, 99.271],\n [0.194, 0.523, 0.401, 0.177, 0.134, 0.035]),\n 'ucb': ([77.163, 95.855, 98.319, 99.052, 99.375, 99.536],\n [0.179, 0.444, 0.129, 0.049, 0.031, 0.025])},\n 'scores': {'ei': ([0.100, 28.080, 45.440, 61.580, 70.240, 77.000],\n [0.063, 3.226, 2.664, 1.326, 1.360, 1.246]),\n 'greedy': ([0.120, 29.440, 51.540, 65.240, 74.660, 82.220],\n [0.075, 1.623, 1.798, 1.372, 0.852, 0.928]),\n 'pi': ([0.120, 28.540, 47.440, 60.660, 69.780, 76.100],\n [0.075, 4.280, 1.728, 1.469, 1.423, 0.867]),\n 'thompson': ([0.100, 12.580, 35.340, 52.980, 65.440, 74.800],\n [0.000, 2.104, 4.061, 3.383, 3.447, 1.862]),\n 'ucb': ([0.200, 27.260, 51.220, 66.920, 77.160, 83.980],\n [0.167, 3.342, 2.420, 1.382, 1.084, 0.801])},\n 'smis': {'ei': ([0.100, 27.240, 43.800, 59.240, 67.580, 73.880],\n [0.063, 3.013, 2.598, 1.385, 1.367, 1.182]),\n 'greedy': ([0.120, 28.480, 49.380, 62.140, 70.940, 78.100],\n [0.075, 1.476, 1.761, 1.255, 0.905, 0.829]),\n 'pi': ([0.120, 27.640, 45.600, 58.220, 66.940, 73.000],\n [0.075, 4.101, 1.575, 1.595, 1.407, 0.746]),\n 'thompson': ([0.100, 12.200, 34.040, 50.680, 62.460, 71.200],\n [0.000, 2.066, 3.568, 3.149, 3.231, 1.793]),\n 'ucb': ([0.200, 26.720, 49.320, 64.240, 73.680, 79.920],\n [0.167, 3.161, 2.278, 1.322, 0.999, 0.708])}},\n 'nn': {'avg': {'ei': ([77.132, 94.567, 96.391, 96.959, 97.412, 97.737],\n [0.276, 0.295, 0.588, 0.619, 0.420, 0.287]),\n 'greedy': ([76.881, 94.686, 97.377, 98.306, 98.775, 99.075],\n [0.335, 0.513, 0.195, 0.132, 0.078, 0.085]),\n 'pi': ([76.955, 94.528, 96.593, 97.227, 97.551, 97.935],\n [0.157, 0.156, 0.491, 0.153, 0.163, 0.169]),\n 'thompson': ([76.912, 93.766, 97.011, 98.079, 98.719, 99.053],\n [0.189, 0.335, 0.196, 0.082, 0.054, 0.027]),\n 'ucb': ([77.051, 95.030, 97.167, 98.245, 98.781, 99.044],\n [0.209, 0.296, 0.363, 0.099, 0.070, 0.047])},\n 'scores': {'ei': ([0.140, 17.920, 29.200, 34.320, 39.260, 43.300],\n [0.150, 1.560, 4.996, 5.578, 4.445, 3.823]),\n 'greedy': ([0.160, 17.960, 39.060, 53.020, 63.020, 70.500],\n [0.150, 3.215, 2.786, 2.316, 1.766, 1.785]),\n 'pi': ([0.100, 18.040, 31.480, 37.220, 41.060, 46.260],\n [0.089, 1.017, 3.046, 1.653, 1.966, 2.276]),\n 'thompson': ([0.140, 14.440, 34.160, 48.080, 60.020, 67.960],\n [0.102, 1.960, 2.741, 1.105, 1.120, 0.794]),\n 'ucb': ([0.120, 20.640, 36.300, 50.760, 61.460, 68.040],\n [0.075, 2.246, 3.511, 1.652, 1.669, 0.911])},\n 'smis': {'ei': ([0.120, 17.120, 27.960, 32.940, 37.680, 41.480],\n [0.160, 1.548, 4.763, 5.262, 4.086, 3.531]),\n 'greedy': ([0.160, 16.860, 36.880, 50.400, 59.940, 66.880],\n [0.150, 3.140, 2.595, 2.140, 1.638, 1.623]),\n 'pi': ([0.100, 17.060, 30.020, 35.460, 39.220, 44.180],\n [0.089, 1.013, 2.908, 1.661, 1.873, 2.180]),\n 'thompson': ([0.140, 13.760, 32.640, 45.980, 57.320, 64.760],\n [0.102, 1.927, 2.587, 1.053, 1.048, 0.750]),\n 'ucb': ([0.120, 19.660, 34.660, 48.620, 58.780, 64.840],\n [0.075, 2.207, 3.383, 1.787, 1.845, 1.169])}},\n 'rf': {'avg': {'ei': ([76.999, 89.700, 93.547, 94.633, 95.232, 95.616],\n [0.209, 1.906, 0.796, 0.517, 0.566, 0.369]),\n 'greedy': ([77.075, 93.313, 96.596, 97.451, 97.976, 98.540],\n [0.239, 0.623, 0.171, 0.171, 0.068, 0.244]),\n 'pi': ([77.111, 92.042, 94.651, 95.478, 95.927, 96.274],\n [0.183, 1.109, 0.335, 0.548, 0.319, 0.367]),\n 'thompson': ([77.073, 88.705, 93.618, 95.832, 96.847, 97.429],\n [0.249, 0.226, 0.415, 0.222, 0.142, 0.176]),\n 'ucb': ([77.273, 89.118, 94.814, 96.101, 96.935, 97.155],\n [0.189, 4.484, 0.751, 0.431, 0.381, 0.419])},\n 'scores': {'ei': ([0.060, 4.540, 14.020, 18.120, 20.760, 22.600],\n [0.080, 2.620, 3.558, 3.170, 3.693, 2.744]),\n 'greedy': ([0.160, 14.100, 30.740, 39.000, 45.860, 55.800],\n [0.102, 2.609, 1.515, 2.352, 1.701, 4.894]),\n 'pi': ([0.120, 10.600, 18.620, 22.720, 25.340, 27.920],\n [0.098, 4.363, 2.107, 3.421, 2.072, 2.697]),\n 'thompson': ([0.100, 3.760, 15.140, 25.200, 32.660, 38.760],\n [0.063, 0.882, 2.148, 1.858, 1.674, 2.483]),\n 'ucb': ([0.100, 6.900, 20.820, 27.400, 34.000, 36.160],\n [0.089, 3.401, 3.022, 3.666, 3.795, 4.247])},\n 'smis': {'ei': ([0.060, 4.380, 13.540, 17.480, 20.020, 21.860],\n [0.080, 2.529, 3.391, 3.112, 3.570, 2.603]),\n 'greedy': ([0.120, 13.560, 29.400, 37.300, 43.800, 53.420],\n [0.117, 2.372, 1.432, 2.249, 1.390, 4.640]),\n 'pi': ([0.120, 10.260, 18.000, 22.000, 24.540, 27.100],\n [0.098, 4.414, 2.183, 3.362, 2.088, 2.739]),\n 'thompson': ([0.080, 3.700, 14.680, 24.440, 31.440, 37.100],\n [0.075, 0.888, 2.134, 1.868, 1.765, 2.552]),\n 'ucb': ([0.100, 6.660, 20.060, 26.360, 32.820, 34.940],\n [0.089, 3.313, 2.904, 3.588, 3.798, 4.206])}}}", "_____no_output_____" ], [ "# top-1000, replicate of 3\nHTS_02_004 = {\n 'mpn': {'avg': {'greedy': ([89.441, 99.326],\n [0.082, 0.078])},\n 'scores': {'greedy': ([2.067, 77.367],\n [0.685, 2.145])},\n 'smis': {'greedy': ([2.000, 73.500],\n [0.638, 2.040])}},\n 'nn': {'avg': {'greedy': ([89.414, 98.477],\n [0.050, 0.023])},\n 'scores': {'greedy': ([2.200, 56.400],\n [0.356, 0.245])},\n 'smis': {'greedy': ([2.100, 53.500],\n [0.356, 0.245])}},\n 'rf': {'avg': {'greedy': ([89.403, 97.857],\n [0.139, 0.230])},\n 'scores': {'greedy': ([1.633, 42.867],\n [0.492, 3.229])},\n 'smis': {'greedy': ([1.467, 40.833],\n [0.419, 3.065])}}}", "_____no_output_____" ], [ "# top-1000, replicate of 3\nHTS_004_02 = {\n 'mpn': {'avg': {'greedy': ([83.876, 99.854],\n [0.113, 0.032])},\n 'scores': {'greedy': ([0.367, 94.933],\n [0.047, 0.818])},\n 'smis': {'greedy': ([0.367, 92.400],\n [0.047, 0.726])}},\n 'nn': {'avg': {'greedy': ([83.905, 99.506],\n [0.179, 0.100])},\n 'scores': {'greedy': ([0.567, 87.300],\n [0.262, 1.980])},\n 'smis': {'greedy': ([0.533, 82.700],\n [0.249, 2.276])}},\n 'rf': {'avg': {'greedy': ([83.805, 99.344], \n [0.116, 0.081])},\n 'scores': {'greedy': ([0.467, 76.567],\n [0.170, 2.788])},\n 'smis': {'greedy': ([0.433, 72.833],\n [0.189, 2.370])}}}", "_____no_output_____" ], [ "AmpC_001_random = {\n 'avg': ([48.793, 54.653, 58.036, 60.576, 62.639, 64.436],\n [0.052, 0.040, 0.046, 0.030, 0.049, 0.053]),\n 'scores': ([0.113, 0.221, 0.315, 0.422, 0.528, 0.632],\n [0.010, 0.007, 0.010, 0.021, 0.013, 0.006]),\n 'smis': ([0.113, 0.221, 0.314, 0.421, 0.527, 0.631],\n [0.009, 0.006, 0.010, 0.020, 0.012, 0.005])}", "_____no_output_____" ], [ "AmpC_001_online = {\n 'mpn': {'avg': {'ei': ([48.719, 87.471, 88.509, 91.554, 92.119, 94.148],\n [0.002, 0.643, 0.330, 0.249, 0.382, 0.318]),\n 'greedy': ([48.792, 89.567, 91.003, 94.531, 95.176, 96.269],\n [0.006, 0.571, 0.746, 0.300, 0.143, 0.332]),\n 'pi': ([48.727, 87.548, 89.254, 92.218, 93.118, 94.500],\n [0.018, 0.123, 0.148, 0.118, 0.173, 0.130]),\n 'thompson': ([48.754, 87.180, 92.068, 95.072, 96.336, 97.212],\n [0.020, 0.233, 0.640, 0.239, 0.294, 0.264]),\n 'ucb': ([48.725, 87.942, 90.297, 93.999, 94.498, 95.017],\n [0.025, 0.092, 0.689, 0.265, 0.265, 0.765])},\n 'scores': {'ei': ([0.092, 6.909, 8.430, 13.688, 15.743, 23.308],\n [0.004, 0.955, 0.830, 0.942, 1.629, 1.730]),\n 'greedy': ([0.086, 9.327, 13.062, 24.671, 28.444, 36.536],\n [0.002, 1.317, 1.752, 1.603, 0.864, 2.844]),\n 'pi': ([0.109, 6.942, 9.574, 15.919, 19.329, 25.243],\n [0.003, 0.846, 0.922, 0.961, 0.897, 0.809]),\n 'thompson': ([0.099, 6.130, 15.966, 28.510, 37.591, 45.871],\n [0.003, 0.032, 1.956, 1.280, 2.453, 2.891]),\n 'ucb': ([0.103, 9.069, 13.482, 23.994, 26.445, 29.092],\n [0.003, 0.273, 1.798, 1.620, 1.731, 4.300])},\n 'smis': {'ei': ([0.091, 6.904, 8.424, 13.680, 15.734, 23.291],\n [0.005, 0.952, 0.828, 0.942, 1.628, 1.727]),\n 'greedy': ([0.086, 9.322, 13.056, 24.656, 28.429, 36.514],\n [0.002, 1.314, 1.750, 1.604, 0.863, 2.840]),\n 'pi': ([0.109, 6.937, 9.568, 15.902, 19.309, 25.219],\n [0.003, 0.847, 0.922, 0.962, 0.899, 0.811]),\n 'thompson': ([0.099, 6.125, 15.956, 28.488, 37.563, 45.836],\n [0.003, 0.033, 1.954, 1.278, 2.453, 2.892]),\n 'ucb': ([0.102, 9.064, 13.476, 23.978, 26.427, 29.071],\n [0.002, 0.272, 1.796, 1.614, 1.725, 4.291])}},\n 'nn': {'avg': {'ei': ([48.770, 87.837, 87.842, 88.916, 88.999, 89.010],\n [0.010, 0.282, 0.278, 1.292, 1.393, 1.392]),\n 'greedy': ([48.711, 88.540, 90.187, 91.955, 92.945, 93.277],\n [0.044, 0.238, 0.740, 0.126, 0.220, 0.666]),\n 'pi': ([48.777, 87.822, 88.418, 89.057, 89.068, 89.245],\n [0.018, 0.279, 0.999, 1.865, 1.862, 1.873]),\n 'thompson': ([48.770, 88.125, 88.815, 91.440, 92.197, 93.158],\n [0.025, 0.318, 0.160, 0.337, 0.347, 0.238]),\n 'ucb': ([48.725, 88.218, 88.356, 89.263, 89.616, 90.164],\n [0.015, 0.055, 0.102, 0.940, 1.246, 1.783])},\n 'scores': {'ei': ([0.102, 6.172, 6.177, 7.699, 7.841, 7.852],\n [0.012, 0.202, 0.198, 1.974, 2.167, 2.173]),\n 'greedy': ([0.097, 7.398, 10.118, 13.457, 16.550, 17.945],\n [0.004, 0.143, 1.560, 0.504, 0.795, 2.718]),\n 'pi': ([0.097, 6.469, 7.346, 8.454, 8.461, 8.625],\n [0.003, 0.367, 1.454, 2.993, 2.994, 3.159]),\n 'thompson': ([0.103, 7.253, 8.081, 12.188, 13.921, 16.940],\n [0.020, 0.225, 0.507, 0.818, 0.737, 0.894]),\n 'ucb': ([0.095, 7.289, 7.387, 8.283, 8.860, 10.020],\n [0.009, 0.169, 0.203, 1.057, 1.594, 2.744])},\n 'smis': {'ei': ([0.102, 6.161, 6.165, 7.687, 7.829, 7.839],\n [0.012, 0.205, 0.200, 1.971, 2.163, 2.169]),\n 'greedy': ([0.097, 7.389, 10.107, 13.443, 16.531, 17.924],\n [0.004, 0.141, 1.557, 0.503, 0.797, 2.718]),\n 'pi': ([0.097, 6.462, 7.338, 8.446, 8.453, 8.617],\n [0.003, 0.366, 1.455, 2.995, 2.996, 3.160]),\n 'thompson': ([0.103, 7.247, 8.074, 12.175, 13.907, 16.925],\n [0.020, 0.224, 0.507, 0.816, 0.736, 0.893]),\n 'ucb': ([0.095, 7.280, 7.378, 8.274, 8.851, 10.011],\n [0.009, 0.168, 0.202, 1.056, 1.593, 2.743])}},\n 'rf': {'avg': {'ei': ([48.786, 85.084, 85.670, 88.656, 88.782, 89.748],\n [0.011, 0.538, 0.456, 0.673, 0.601, 1.216]),\n 'greedy': ([48.744, 88.730, 89.659, 91.901, 92.037, 93.013],\n [0.029, 0.066, 0.132, 0.099, 0.084, 0.090]),\n 'pi': ([48.722, 84.991, 85.261, 88.422, 88.644, 89.244],\n [0.080, 1.046, 1.083, 0.462, 0.651, 0.588]),\n 'thompson': ([48.796, 84.519, 84.777, 89.450, 89.491, 91.249],\n [0.036, 0.371, 0.297, 0.198, 0.187, 0.103]),\n 'ucb': ([48.721, 85.789, 86.115, 89.014, 89.457, 90.494],\n [0.050, 0.519, 0.559, 0.449, 0.326, 0.957])},\n 'scores': {'ei': ([0.087, 5.282, 5.557, 7.711, 7.851, 9.399],\n [0.023, 0.556, 0.540, 0.558, 0.599, 1.775]),\n 'greedy': ([0.095, 6.779, 8.435, 12.539, 12.965, 15.893],\n [0.003, 0.386, 0.425, 0.305, 0.123, 0.443]),\n 'pi': ([0.104, 4.763, 4.971, 7.375, 7.655, 8.226],\n [0.009, 0.461, 0.541, 1.141, 1.342, 1.090]),\n 'thompson': ([0.100, 3.874, 4.013, 7.945, 8.029, 11.051],\n [0.006, 0.044, 0.111, 0.114, 0.114, 0.118]),\n 'ucb': ([0.099, 4.741, 4.980, 7.754, 8.429, 10.671],\n [0.014, 0.809, 0.793, 0.495, 0.273, 1.594])},\n 'smis': {'ei': ([0.087, 5.281, 5.556, 7.709, 7.849, 9.395],\n [0.023, 0.556, 0.539, 0.558, 0.599, 1.772]),\n 'greedy': ([0.095, 6.770, 8.425, 12.526, 12.951, 15.879],\n [0.003, 0.385, 0.426, 0.304, 0.123, 0.443]),\n 'pi': ([0.104, 4.761, 4.970, 7.373, 7.653, 8.223],\n [0.009, 0.460, 0.540, 1.140, 1.340, 1.088]),\n 'thompson': ([0.100, 3.870, 4.009, 7.939, 8.023, 11.044],\n [0.006, 0.048, 0.114, 0.111, 0.112, 0.121]),\n 'ucb': ([0.099, 4.740, 4.978, 7.751, 8.425, 10.667],\n [0.014, 0.810, 0.795, 0.495, 0.272, 1.593])}}}", "_____no_output_____" ], [ "AmpC_001_retrain = {\n 'mpn': {'avg': {'ei': ([48.753, 87.369, 90.782, 93.038, 94.460, 95.385],\n [0.012, 1.004, 0.514, 0.435, 0.298, 0.295]),\n 'greedy': ([48.716, 89.005, 94.439, 96.243, 97.120, 97.677],\n [0.027, 0.259, 0.021, 0.098, 0.105, 0.041]),\n 'pi': ([48.784, 87.347, 90.939, 93.136, 94.490, 95.487],\n [0.003, 0.334, 0.207, 0.329, 0.259, 0.175]),\n 'thompson': ([48.726, 87.683, 92.966, 95.363, 96.707, 97.533],\n [0.041, 0.234, 0.089, 0.091, 0.091, 0.032]),\n 'ucb': ([48.757, 87.136, 92.169, 94.834, 96.397, 97.362],\n [0.038, 0.888, 0.365, 0.079, 0.064, 0.050])},\n 'scores': {'ei': ([0.095, 7.989, 13.075, 19.487, 25.430, 30.499],\n [0.002, 1.286, 1.255, 1.548, 1.359, 1.795]),\n 'greedy': ([0.099, 8.230, 23.930, 35.933, 44.867, 51.994],\n [0.009, 0.625, 0.276, 0.934, 1.238, 0.507]),\n 'pi': ([0.095, 6.857, 12.847, 19.564, 25.386, 31.132],\n [0.010, 0.452, 0.949, 1.724, 1.701, 1.333]),\n 'thompson': ([0.101, 7.165, 18.356, 30.160, 40.716, 49.434],\n [0.013, 0.290, 0.332, 0.630, 0.785, 0.314]),\n 'ucb': ([0.085, 7.342, 16.827, 27.923, 38.273, 47.144],\n [0.009, 1.458, 1.426, 0.439, 0.537, 0.490])},\n 'smis': {'ei': ([0.095, 7.983, 13.065, 19.472, 25.411, 30.475],\n [0.002, 1.287, 1.255, 1.548, 1.359, 1.796]),\n 'greedy': ([0.099, 8.227, 23.917, 35.911, 44.836, 51.958],\n [0.009, 0.624, 0.276, 0.937, 1.240, 0.507]),\n 'pi': ([0.095, 6.851, 12.839, 19.549, 25.366, 31.106],\n [0.010, 0.452, 0.950, 1.723, 1.700, 1.333]),\n 'thompson': ([0.101, 7.161, 18.345, 30.141, 40.687, 49.397],\n [0.013, 0.288, 0.330, 0.628, 0.788, 0.315]),\n 'ucb': ([0.085, 7.335, 16.814, 27.901, 38.247, 47.111],\n [0.009, 1.456, 1.427, 0.442, 0.531, 0.488])}},\n 'nn': {'avg': {'ei': ([48.719, 88.003, 90.455, 91.357, 91.942, 92.414],\n [0.008, 0.321, 0.568, 0.628, 0.080, 0.303]),\n 'greedy': ([48.722, 88.826, 92.886, 94.420, 95.372, 95.999],\n [0.022, 0.095, 0.150, 0.064, 0.051, 0.035]),\n 'pi': ([48.753, 88.109, 90.675, 91.120, 91.700, 92.722],\n [0.041, 0.371, 0.337, 0.227, 0.451, 0.428]),\n 'thompson': ([48.799, 87.707, 92.071, 93.928, 95.026, 95.731],\n [0.053, 0.443, 0.219, 0.094, 0.058, 0.100]),\n 'ucb': ([48.766, 87.963, 92.267, 94.083, 95.107, 95.800],\n [0.057, 0.190, 0.030, 0.071, 0.103, 0.070])},\n 'scores': {'ei': ([0.099, 6.889, 10.001, 11.855, 13.142, 14.499],\n [0.006, 0.431, 0.979, 1.409, 0.288, 0.828]),\n 'greedy': ([0.094, 8.027, 16.143, 22.403, 28.315, 33.251],\n [0.018, 0.277, 0.541, 0.337, 0.447, 0.329]),\n 'pi': ([0.111, 6.864, 10.086, 10.975, 12.335, 15.207],\n [0.008, 0.526, 0.777, 0.659, 1.031, 1.421]),\n 'thompson': ([0.116, 6.352, 13.411, 19.985, 25.988, 31.043],\n [0.003, 0.428, 0.697, 0.540, 0.403, 0.783]),\n 'ucb': ([0.101, 6.774, 14.020, 20.736, 26.458, 31.547],\n [0.012, 0.350, 0.081, 0.351, 0.762, 0.608])},\n 'smis': {'ei': ([0.099, 6.881, 9.990, 11.841, 13.127, 14.482],\n [0.006, 0.431, 0.978, 1.408, 0.291, 0.824]),\n 'greedy': ([0.094, 8.017, 16.127, 22.386, 28.293, 33.225],\n [0.018, 0.274, 0.539, 0.334, 0.447, 0.331]),\n 'pi': ([0.111, 6.855, 10.075, 10.963, 12.323, 15.192],\n [0.008, 0.528, 0.779, 0.660, 1.030, 1.419]),\n 'thompson': ([0.115, 6.346, 13.398, 19.966, 25.967, 31.017],\n [0.003, 0.428, 0.694, 0.537, 0.401, 0.781]),\n 'ucb': ([0.101, 6.766, 14.006, 20.714, 26.433, 31.520],\n [0.012, 0.350, 0.080, 0.350, 0.762, 0.609])}},\n 'rf': {'avg': {'ei': ([48.737, 84.648, 87.105, 88.320, 89.255, 90.028],\n [0.028, 0.555, 0.317, 0.715, 0.924, 1.030]),\n 'greedy': ([48.778, 88.889, 91.863, 93.255, 94.062, 94.764],\n [0.042, 0.197, 0.331, 0.252, 0.366, 0.346]),\n 'pi': ([48.765, 85.709, 88.188, 88.990, 89.615, 90.443],\n [0.028, 0.410, 0.435, 0.412, 0.487, 0.539]),\n 'thompson': ([48.759, 84.805, 89.590, 91.976, 93.251, 94.076],\n [0.044, 0.533, 0.682, 0.783, 0.532, 0.381]),\n 'ucb': ([48.767, 86.557, 88.683, 89.745, 91.040, 92.007],\n [0.016, 0.804, 0.541, 0.052, 0.596, 0.421])},\n 'scores': {'ei': ([0.103, 4.907, 6.023, 7.016, 8.475, 9.755],\n [0.005, 0.783, 0.828, 0.830, 1.949, 2.338]),\n 'greedy': ([0.094, 6.791, 12.151, 16.664, 20.207, 24.003],\n [0.016, 0.564, 1.272, 1.320, 1.936, 2.233]),\n 'pi': ([0.101, 5.143, 6.630, 7.634, 8.439, 9.752],\n [0.010, 0.269, 0.691, 0.731, 1.026, 1.264]),\n 'thompson': ([0.106, 4.057, 7.879, 12.683, 16.671, 20.072],\n [0.014, 0.436, 1.566, 2.461, 2.255, 2.048]),\n 'ucb': ([0.096, 5.465, 7.327, 8.596, 11.537, 13.808],\n [0.013, 1.562, 1.404, 1.016, 1.049, 0.959])},\n 'smis': {'ei': ([0.103, 4.906, 6.021, 7.012, 8.471, 9.751],\n [0.005, 0.784, 0.828, 0.830, 1.949, 2.338]),\n 'greedy': ([0.094, 6.783, 12.137, 16.647, 20.187, 23.981],\n [0.016, 0.563, 1.272, 1.320, 1.938, 2.233]),\n 'pi': ([0.101, 5.141, 6.629, 7.631, 8.437, 9.749],\n [0.010, 0.269, 0.691, 0.731, 1.026, 1.263]),\n 'thompson': ([0.106, 4.055, 7.875, 12.678, 16.661, 20.060],\n [0.014, 0.436, 1.565, 2.461, 2.252, 2.048]),\n 'ucb': ([0.095, 5.463, 7.323, 8.591, 11.530, 13.801],\n [0.012, 1.559, 1.402, 1.015, 1.047, 0.957])}}}", "_____no_output_____" ], [ "AmpC_002_random = {\n 'avg': ([54.602, 60.453, 64.333, 67.379, 69.976, 72.228],\n [0.062, 0.051, 0.084, 0.103, 0.101, 0.096]),\n 'scores': ([0.209, 0.398, 0.595, 0.805, 1.010, 1.218],\n [0.008, 0.011, 0.031, 0.028, 0.027, 0.048]),\n 'smis': ([0.209, 0.398, 0.595, 0.805, 1.009, 1.217],\n [0.008, 0.011, 0.031, 0.028, 0.028, 0.048])}", "_____no_output_____" ], [ "AmpC_002_online = {\n 'mpn': {'avg': {'ei': ([54.571, 91.517, 92.589, 95.318, 95.959, 96.977],\n [0.041, 0.372, 0.245, 0.192, 0.237, 0.232]),\n 'greedy': ([54.624, 92.892, 95.324, 97.289, 97.400, 97.414],\n [0.024, 0.276, 0.531, 0.346, 0.247, 0.242]),\n 'pi': ([54.621, 91.464, 92.734, 95.337, 95.989, 96.911],\n [0.019, 0.109, 0.187, 0.029, 0.044, 0.079]),\n 'thompson': ([54.582, 91.630, 95.726, 97.507, 98.157, 98.460],\n [0.042, 0.388, 0.139, 0.065, 0.057, 0.186]),\n 'ucb': ([54.576, 91.625, 92.986, 95.907, 96.452, 96.783],\n [0.071, 0.358, 0.579, 0.282, 0.450, 0.706])},\n 'scores': {'ei': ([0.207, 14.079, 17.473, 29.541, 34.071, 42.694],\n [0.022, 0.995, 0.818, 1.125, 1.759, 2.409]),\n 'greedy': ([0.185, 16.261, 29.485, 46.891, 48.132, 48.314],\n [0.016, 1.357, 3.478, 4.064, 2.999, 2.952]),\n 'pi': ([0.196, 13.624, 17.801, 29.745, 34.260, 41.929],\n [0.012, 0.132, 0.973, 0.335, 0.446, 0.709]),\n 'thompson': ([0.195, 13.073, 32.226, 49.138, 58.151, 63.145],\n [0.011, 1.157, 1.111, 0.905, 0.913, 3.135]),\n 'ucb': ([0.207, 13.601, 18.778, 33.919, 38.360, 41.698],\n [0.015, 1.149, 2.469, 2.298, 3.826, 6.658])},\n 'smis': {'ei': ([0.207, 14.072, 17.464, 29.521, 34.045, 42.663],\n [0.022, 0.991, 0.812, 1.123, 1.756, 2.409]),\n 'greedy': ([0.185, 16.247, 29.464, 46.855, 48.095, 48.277],\n [0.016, 1.357, 3.473, 4.059, 2.994, 2.948]),\n 'pi': ([0.196, 13.612, 17.787, 29.722, 34.232, 41.896],\n [0.012, 0.132, 0.971, 0.334, 0.444, 0.712]),\n 'thompson': ([0.195, 13.063, 32.199, 49.105, 58.113, 63.101],\n [0.011, 1.156, 1.109, 0.903, 0.916, 3.130]),\n 'ucb': ([0.207, 13.593, 18.764, 33.895, 38.335, 41.671],\n [0.015, 1.148, 2.471, 2.299, 3.826, 6.657])}},\n 'nn': {'avg': {'ei': ([54.597, 91.333, 91.343, 91.408, 91.465, 91.483],\n [0.032, 0.236, 0.237, 0.270, 0.235, 0.222]),\n 'greedy': ([54.630, 91.824, 92.551, 93.746, 93.954, 94.689],\n [0.013, 0.144, 0.374, 1.182, 1.125, 1.092]),\n 'pi': ([54.567, 91.526, 91.530, 92.158, 92.220, 92.256],\n [0.070, 0.037, 0.039, 0.911, 0.871, 0.876]),\n 'thompson': ([54.615, 91.061, 91.435, 92.282, 92.527, 92.828],\n [0.035, 0.311, 0.363, 1.476, 1.707, 2.123]),\n 'ucb': ([54.575, 91.296, 91.623, 92.667, 92.889, 93.286],\n [0.030, 0.140, 0.043, 0.815, 0.836, 1.139])},\n 'scores': {'ei': ([0.193, 11.553, 11.569, 11.633, 11.676, 11.685],\n [0.028, 0.392, 0.390, 0.415, 0.388, 0.385]),\n 'greedy': ([0.195, 13.037, 15.312, 20.443, 21.219, 24.866],\n [0.003, 0.184, 1.149, 4.744, 4.866, 6.017]),\n 'pi': ([0.225, 12.030, 12.042, 14.131, 14.196, 14.237],\n [0.017, 0.082, 0.087, 3.018, 2.983, 3.019]),\n 'thompson': ([0.200, 10.845, 11.717, 14.773, 15.933, 18.031],\n [0.013, 1.072, 1.135, 5.271, 6.819, 9.773]),\n 'ucb': ([0.198, 11.369, 12.125, 15.184, 15.862, 17.622],\n [0.008, 0.441, 0.132, 2.578, 2.672, 4.159])},\n 'smis': {'ei': ([0.193, 11.540, 11.556, 11.619, 11.663, 11.672],\n [0.028, 0.394, 0.392, 0.417, 0.390, 0.387]),\n 'greedy': ([0.195, 13.026, 15.300, 20.423, 21.199, 24.844],\n [0.003, 0.185, 1.148, 4.737, 4.859, 6.009]),\n 'pi': ([0.225, 12.018, 12.030, 14.118, 14.183, 14.224],\n [0.017, 0.080, 0.085, 3.014, 2.978, 3.014]),\n 'thompson': ([0.200, 10.833, 11.705, 14.758, 15.918, 18.014],\n [0.013, 1.069, 1.133, 5.267, 6.814, 9.765]),\n 'ucb': ([0.198, 11.359, 12.113, 15.172, 15.849, 17.607],\n [0.008, 0.438, 0.131, 2.577, 2.670, 4.154])}},\n 'rf': {'avg': {'ei': ([54.629, 89.144, 89.227, 91.656, 92.087, 93.031],\n [0.087, 0.188, 0.232, 0.699, 0.685, 0.658]),\n 'greedy': ([54.536, 92.144, 92.650, 94.400, 94.558, 95.669],\n [0.057, 0.098, 0.241, 0.259, 0.301, 0.237]),\n 'pi': ([54.610, 89.187, 89.371, 91.292, 92.151, 92.964],\n [0.031, 0.329, 0.436, 0.998, 0.086, 0.628]),\n 'thompson': ([54.625, 90.027, 90.498, 93.554, 93.597, 94.755],\n [0.013, 0.180, 0.230, 0.066, 0.101, 0.543]),\n 'ucb': ([54.594, 89.654, 89.835, 92.342, 92.389, 93.227],\n [0.025, 0.172, 0.339, 0.288, 0.291, 0.459])},\n 'scores': {'ei': ([0.199, 8.636, 8.759, 13.615, 14.649, 17.320],\n [0.007, 0.196, 0.230, 1.773, 2.046, 2.277]),\n 'greedy': ([0.221, 12.919, 14.668, 21.807, 22.691, 30.355],\n [0.021, 0.395, 0.860, 1.676, 1.968, 1.965]),\n 'pi': ([0.183, 8.001, 8.319, 12.207, 14.028, 16.668],\n [0.008, 0.421, 0.661, 2.141, 0.210, 2.331]),\n 'thompson': ([0.217, 8.603, 9.431, 18.118, 18.305, 24.219],\n [0.021, 0.256, 0.409, 0.211, 0.281, 3.282]),\n 'ucb': ([0.187, 7.604, 8.030, 13.806, 13.936, 16.763],\n [0.016, 0.107, 0.448, 0.763, 0.787, 1.779])},\n 'smis': {'ei': ([0.199, 8.634, 8.757, 13.610, 14.643, 17.314],\n [0.007, 0.197, 0.232, 1.770, 2.043, 2.273]),\n 'greedy': ([0.221, 12.905, 14.651, 21.789, 22.673, 30.333],\n [0.020, 0.396, 0.861, 1.678, 1.969, 1.966]),\n 'pi': ([0.183, 7.999, 8.317, 12.203, 14.024, 16.661],\n [0.008, 0.422, 0.662, 2.142, 0.211, 2.330]),\n 'thompson': ([0.217, 8.597, 9.425, 18.104, 18.291, 24.201],\n [0.021, 0.257, 0.412, 0.211, 0.281, 3.279]),\n 'ucb': ([0.187, 7.601, 8.027, 13.798, 13.928, 16.754],\n [0.016, 0.108, 0.449, 0.759, 0.782, 1.777])}}}", "_____no_output_____" ], [ "AmpC_002_retrain = {\n 'mpn': {'avg': {'ei': ([54.595, 91.112, 94.196, 95.839, 96.951, 97.716],\n [0.015, 0.096, 0.148, 0.097, 0.041, 0.023]),\n 'greedy': ([54.542, 92.580, 96.611, 97.855, 98.455, 98.614],\n [0.022, 0.211, 0.183, 0.147, 0.126, 0.094]),\n 'pi': ([54.575, 91.188, 94.499, 96.142, 97.266, 97.842],\n [0.056, 0.330, 0.192, 0.149, 0.094, 0.044]),\n 'thompson': ([54.585, 91.780, 95.884, 97.551, 98.062, 98.678],\n [0.030, 0.259, 0.116, 0.100, 0.313, 0.171]),\n 'ucb': ([54.634, 91.079, 95.379, 97.595, 98.680, 99.161],\n [0.003, 0.306, 0.054, 0.040, 0.056, 0.029])},\n 'scores': {'ei': ([0.196, 12.596, 23.377, 33.123, 42.395, 50.805],\n [0.003, 0.451, 0.905, 0.935, 0.500, 0.292]),\n 'greedy': ([0.194, 14.833, 39.277, 54.836, 64.313, 67.093],\n [0.022, 1.144, 1.899, 1.859, 1.703, 2.146]),\n 'pi': ([0.191, 12.791, 25.151, 35.516, 45.620, 52.303],\n [0.008, 1.069, 1.172, 1.250, 0.999, 0.619]),\n 'thompson': ([0.193, 13.231, 33.189, 49.799, 57.181, 67.449],\n [0.022, 0.905, 0.913, 1.186, 4.715, 3.324]),\n 'ucb': ([0.203, 12.353, 30.471, 49.814, 65.689, 75.505],\n [0.019, 0.623, 0.232, 0.507, 1.090, 0.711])},\n 'smis': {'ei': ([0.196, 12.587, 23.359, 33.095, 42.361, 50.765],\n [0.003, 0.449, 0.904, 0.933, 0.504, 0.292]),\n 'greedy': ([0.194, 14.821, 39.251, 54.798, 64.266, 67.044],\n [0.022, 1.141, 1.895, 1.859, 1.702, 2.144]),\n 'pi': ([0.191, 12.784, 25.135, 35.494, 45.590, 52.271],\n [0.008, 1.066, 1.170, 1.248, 0.994, 0.620]),\n 'thompson': ([0.193, 13.223, 33.170, 49.768, 57.143, 67.404],\n [0.022, 0.904, 0.913, 1.184, 4.711, 3.322]),\n 'ucb': ([0.203, 12.346, 30.455, 49.784, 65.649, 75.457],\n [0.019, 0.621, 0.229, 0.505, 1.088, 0.708])}},\n 'nn': {'avg': {'ei': ([54.557, 91.470, 92.868, 93.674, 93.902, 94.746],\n [0.029, 0.212, 0.625, 0.595, 0.427, 0.165]),\n 'greedy': ([54.622, 92.035, 95.098, 96.409, 97.247, 97.721],\n [0.037, 0.345, 0.115, 0.022, 0.036, 0.033]),\n 'pi': ([54.648, 91.621, 93.386, 93.751, 94.064, 94.418],\n [0.015, 0.156, 0.155, 0.265, 0.171, 0.194]),\n 'thompson': ([54.606, 91.416, 94.613, 96.145, 96.919, 97.534],\n [0.032, 0.100, 0.030, 0.027, 0.044, 0.068]),\n 'ucb': ([54.567, 91.665, 94.862, 96.216, 96.919, 97.516],\n [0.042, 0.259, 0.124, 0.157, 0.087, 0.040])},\n 'scores': {'ei': ([0.216, 11.909, 16.061, 19.054, 19.995, 24.179],\n [0.009, 0.636, 2.289, 2.487, 1.937, 0.969]),\n 'greedy': ([0.219, 13.545, 26.499, 36.979, 46.306, 52.826],\n [0.016, 0.915, 0.752, 0.183, 0.474, 0.503]),\n 'pi': ([0.195, 12.300, 17.680, 19.214, 20.612, 22.334],\n [0.014, 0.671, 0.699, 1.299, 0.980, 1.148]),\n 'thompson': ([0.189, 11.516, 23.353, 34.457, 42.294, 50.149],\n [0.017, 0.328, 0.232, 0.239, 0.545, 1.021]),\n 'ucb': ([0.203, 12.597, 24.945, 35.081, 42.222, 49.799],\n [0.003, 0.800, 0.771, 1.469, 0.957, 0.546])},\n 'smis': {'ei': ([0.216, 11.897, 16.047, 19.039, 19.980, 24.161],\n [0.009, 0.636, 2.291, 2.487, 1.938, 0.969]),\n 'greedy': ([0.218, 13.531, 26.478, 36.950, 46.269, 52.785],\n [0.016, 0.911, 0.748, 0.182, 0.471, 0.500]),\n 'pi': ([0.195, 12.287, 17.665, 19.197, 20.595, 22.315],\n [0.014, 0.674, 0.699, 1.297, 0.978, 1.144]),\n 'thompson': ([0.189, 11.504, 23.332, 34.425, 42.261, 50.109],\n [0.017, 0.328, 0.228, 0.240, 0.541, 1.017]),\n 'ucb': ([0.203, 12.585, 24.922, 35.051, 42.186, 49.758],\n [0.003, 0.798, 0.771, 1.466, 0.958, 0.547])}},\n 'rf': {'avg': {'ei': ([54.588, 88.782, 89.924, 90.379, 91.597, 92.442],\n [0.036, 0.406, 0.261, 0.159, 0.561, 0.853]),\n 'greedy': ([54.618, 91.876, 94.532, 95.662, 96.569, 97.190],\n [0.012, 0.364, 0.329, 0.316, 0.138, 0.138]),\n 'pi': ([54.620, 89.807, 90.733, 91.349, 92.081, 92.829],\n [0.020, 0.089, 0.547, 0.558, 0.706, 0.442]),\n 'thompson': ([54.596, 89.959, 93.019, 94.819, 95.835, 96.798],\n [0.020, 0.137, 0.343, 0.137, 0.172, 0.166]),\n 'ucb': ([54.620, 89.696, 91.205, 91.982, 93.484, 94.807],\n [0.053, 0.386, 0.455, 0.333, 0.220, 0.401])},\n 'scores': {'ei': ([0.172, 6.993, 8.669, 9.440, 12.229, 14.629],\n [0.011, 0.448, 0.659, 0.424, 1.458, 2.664]),\n 'greedy': ([0.195, 12.132, 22.873, 30.537, 38.577, 45.543],\n [0.019, 1.504, 2.149, 2.490, 1.419, 1.806]),\n 'pi': ([0.201, 9.471, 11.030, 12.143, 13.815, 15.959],\n [0.028, 0.737, 1.457, 1.455, 2.215, 1.649]),\n 'thompson': ([0.214, 8.539, 15.739, 23.995, 31.353, 40.843],\n [0.017, 0.296, 0.983, 0.805, 1.430, 1.907]),\n 'ucb': ([0.207, 8.055, 10.563, 12.635, 17.811, 24.393],\n [0.010, 0.935, 1.243, 0.885, 0.712, 1.955])},\n 'smis': {'ei': ([0.172, 6.992, 8.665, 9.437, 12.224, 14.623],\n [0.011, 0.449, 0.660, 0.424, 1.455, 2.661]),\n 'greedy': ([0.195, 12.122, 22.853, 30.512, 38.545, 45.505],\n [0.019, 1.502, 2.147, 2.488, 1.422, 1.806]),\n 'pi': ([0.201, 9.468, 11.027, 12.139, 13.809, 15.952],\n [0.027, 0.739, 1.456, 1.455, 2.213, 1.644]),\n 'thompson': ([0.214, 8.535, 15.729, 23.977, 31.329, 40.811],\n [0.017, 0.293, 0.980, 0.804, 1.429, 1.902]),\n 'ucb': ([0.206, 8.052, 10.559, 12.631, 17.800, 24.379],\n [0.009, 0.935, 1.242, 0.885, 0.710, 1.950])}}}", "_____no_output_____" ], [ "AmpC_004_random = {\n 'avg': ([60.527, 67.424, 72.200, 75.909, 78.821, 81.033],\n [0.061, 0.060, 0.040, 0.074, 0.052, 0.040]),\n 'scores': ([0.411, 0.787, 1.189, 1.585, 1.977, 2.381],\n [0.040, 0.046, 0.050, 0.057, 0.056, 0.076]),\n 'smis': ([0.410, 0.785, 1.187, 1.583, 1.975, 2.379],\n [0.040, 0.046, 0.052, 0.059, 0.058, 0.079])}", "_____no_output_____" ], [ "AmpC_004_online = {\n 'mpn': {'avg': {'ei': ([60.466, 94.494, 95.478, 97.411, 97.540, 98.413],\n [0.058, 0.611, 0.476, 0.293, 0.231, 0.122]),\n 'greedy': ([60.481, 95.846, 97.267, 98.591, 98.695, 98.950],\n [0.056, 0.194, 0.060, 0.050, 0.113, 0.286]),\n 'pi': ([60.478, 94.360, 95.272, 97.608, 97.815, 98.320],\n [0.037, 0.090, 0.034, 0.024, 0.172, 0.509]),\n 'thompson': ([60.486, 95.013, 97.539, 98.664, 98.871, 98.989],\n [0.032, 0.126, 0.199, 0.101, 0.193, 0.254]),\n 'ucb': ([60.448, 94.918, 96.968, 98.743, 99.017, 99.256],\n [0.082, 0.060, 0.347, 0.133, 0.105, 0.246])},\n 'scores': {'ei': ([0.391, 24.933, 30.864, 47.396, 48.764, 60.919],\n [0.040, 3.515, 3.297, 3.405, 2.794, 2.176]),\n 'greedy': ([0.393, 32.399, 46.498, 66.301, 68.261, 73.729],\n [0.007, 1.697, 0.804, 0.962, 1.985, 5.727]),\n 'pi': ([0.389, 23.393, 28.959, 49.654, 52.258, 60.230],\n [0.012, 0.470, 0.172, 0.282, 2.309, 7.773]),\n 'thompson': ([0.390, 26.394, 49.765, 67.285, 71.391, 73.989],\n [0.025, 0.841, 2.535, 1.924, 3.871, 5.190]),\n 'ucb': ([0.405, 26.710, 43.082, 66.695, 71.859, 77.763],\n [0.018, 0.439, 3.640, 2.436, 2.228, 5.817])},\n 'smis': {'ei': ([0.391, 24.915, 30.841, 47.360, 48.727, 60.874],\n [0.040, 3.509, 3.290, 3.399, 2.789, 2.172]),\n 'greedy': ([0.393, 32.385, 46.469, 66.251, 68.211, 73.673],\n [0.007, 1.696, 0.800, 0.960, 1.982, 5.721]),\n 'pi': ([0.389, 23.381, 28.941, 49.618, 52.218, 60.187],\n [0.012, 0.472, 0.173, 0.277, 2.307, 7.765]),\n 'thompson': ([0.390, 26.376, 49.730, 67.237, 71.338, 73.935],\n [0.025, 0.836, 2.523, 1.917, 3.860, 5.180]),\n 'ucb': ([0.405, 26.693, 43.055, 66.657, 71.814, 77.713],\n [0.018, 0.435, 3.636, 2.436, 2.227, 5.811])}},\n 'nn': {'avg': {'ei': ([60.459, 94.279, 94.486, 95.705, 95.774, 96.059],\n [0.031, 0.020, 0.308, 1.028, 1.088, 1.378]),\n 'greedy': ([60.455, 94.584, 95.124, 96.749, 96.939, 97.589],\n [0.054, 0.132, 0.127, 0.173, 0.215, 0.380]),\n 'pi': ([60.572, 94.205, 94.217, 94.245, 94.270, 94.337],\n [0.051, 0.282, 0.292, 0.280, 0.294, 0.304]),\n 'thompson': ([60.470, 94.164, 94.781, 95.515, 95.593, 95.640],\n [0.053, 0.167, 0.693, 1.114, 1.153, 1.115]),\n 'ucb': ([60.532, 94.416, 94.513, 95.161, 95.212, 95.409],\n [0.017, 0.013, 0.107, 1.006, 1.005, 1.164])},\n 'scores': {'ei': ([0.394, 21.761, 22.985, 32.114, 32.817, 36.260],\n [0.038, 0.064, 1.733, 7.701, 8.459, 12.559]),\n 'greedy': ([0.376, 23.454, 26.665, 40.370, 42.419, 50.717],\n [0.009, 0.919, 0.873, 1.761, 2.376, 5.224]),\n 'pi': ([0.395, 21.414, 21.457, 21.534, 21.627, 21.930],\n [0.025, 1.495, 1.523, 1.486, 1.580, 1.642]),\n 'thompson': ([0.417, 21.323, 24.937, 30.672, 31.360, 31.609],\n [0.045, 0.965, 4.338, 8.148, 8.770, 8.641]),\n 'ucb': ([0.376, 22.499, 22.993, 27.906, 28.235, 30.047],\n [0.013, 0.144, 0.693, 7.609, 7.716, 9.718])},\n 'smis': {'ei': ([0.393, 21.740, 22.963, 32.085, 32.789, 36.228],\n [0.038, 0.064, 1.733, 7.697, 8.456, 12.551]),\n 'greedy': ([0.376, 23.437, 26.645, 40.333, 42.380, 50.675],\n [0.009, 0.921, 0.877, 1.759, 2.371, 5.221]),\n 'pi': ([0.395, 21.392, 21.435, 21.512, 21.605, 21.908],\n [0.025, 1.495, 1.522, 1.486, 1.579, 1.641]),\n 'thompson': ([0.417, 21.299, 24.911, 30.643, 31.331, 31.580],\n [0.045, 0.963, 4.336, 8.144, 8.766, 8.637]),\n 'ucb': ([0.375, 22.476, 22.970, 27.879, 28.207, 30.019],\n [0.012, 0.144, 0.692, 7.603, 7.708, 9.710])}},\n 'rf': {'avg': {'ei': ([60.522, 91.419, 91.608, 94.121, 94.694, 95.045],\n [0.013, 0.353, 0.430, 1.535, 1.196, 1.123]),\n 'greedy': ([60.495, 94.490, 94.722, 96.652, 96.762, 97.702],\n [0.028, 0.309, 0.301, 0.134, 0.183, 0.255]),\n 'pi': ([60.461, 91.558, 91.650, 93.211, 94.143, 94.814],\n [0.079, 0.397, 0.365, 1.029, 0.194, 0.506]),\n 'thompson': ([60.517, 93.728, 94.099, 96.371, 96.453, 97.595],\n [0.103, 0.062, 0.119, 0.085, 0.088, 0.109]),\n 'ucb': ([60.528, 92.941, 93.029, 95.785, 95.928, 96.291],\n [0.043, 0.289, 0.316, 0.254, 0.188, 0.323])},\n 'scores': {'ei': ([0.389, 11.376, 11.890, 22.675, 25.065, 27.063],\n [0.029, 1.654, 1.852, 7.726, 7.844, 7.968]),\n 'greedy': ([0.365, 22.169, 23.573, 39.183, 40.344, 52.307],\n [0.012, 1.888, 1.904, 1.502, 2.045, 3.572]),\n 'pi': ([0.390, 12.247, 12.473, 17.640, 20.956, 24.781],\n [0.014, 1.124, 1.062, 3.321, 0.913, 2.875]),\n 'thompson': ([0.411, 18.621, 20.454, 36.401, 37.205, 50.696],\n [0.005, 0.339, 0.633, 0.725, 0.858, 1.473]),\n 'ucb': ([0.418, 15.949, 16.311, 31.344, 32.459, 35.755],\n [0.007, 1.212, 1.337, 2.073, 1.601, 3.087])},\n 'smis': {'ei': ([0.387, 11.371, 11.885, 22.662, 25.050, 27.047],\n [0.030, 1.651, 1.848, 7.717, 7.835, 7.958]),\n 'greedy': ([0.365, 22.149, 23.554, 39.155, 40.315, 52.269],\n [0.012, 1.888, 1.904, 1.503, 2.048, 3.573]),\n 'pi': ([0.390, 12.241, 12.467, 17.631, 20.945, 24.765],\n [0.014, 1.124, 1.062, 3.320, 0.911, 2.872]),\n 'thompson': ([0.411, 18.605, 20.437, 36.369, 37.173, 50.657],\n [0.005, 0.340, 0.634, 0.725, 0.858, 1.470]),\n 'ucb': ([0.417, 15.946, 16.307, 31.324, 32.438, 35.732],\n [0.008, 1.210, 1.335, 2.066, 1.594, 3.083])}}}", "_____no_output_____" ], [ "AmpC_004_retrain = {\n 'mpn': {'avg': {'ei': ([60.526, 94.018, 96.738, 98.250, 98.729, 99.173],\n [0.058, 0.329, 0.228, 0.095, 0.435, 0.239]),\n 'greedy': ([60.508, 95.474, 98.223, 99.013, 99.293, 99.556],\n [0.007, 0.140, 0.138, 0.095, 0.134, 0.089]),\n 'pi': ([60.443, 94.290, 97.140, 98.334, 98.953, 99.307],\n [0.054, 0.174, 0.100, 0.060, 0.064, 0.035]),\n 'thompson': ([60.500, 94.875, 97.803, 98.777, 99.285, 99.565],\n [0.055, 0.145, 0.034, 0.048, 0.012, 0.009]),\n 'ucb': ([60.556, 95.143, 98.512, 99.461, 99.721, 99.828],\n [0.029, 0.015, 0.052, 0.026, 0.006, 0.004])},\n 'scores': {'ei': ([0.394, 22.079, 40.393, 58.185, 66.873, 75.440],\n [0.033, 1.931, 2.213, 1.545, 7.860, 5.359]),\n 'greedy': ([0.385, 29.185, 60.558, 75.195, 81.395, 87.895],\n [0.021, 1.212, 2.087, 1.613, 3.184, 2.317]),\n 'pi': ([0.378, 23.500, 44.471, 59.473, 70.519, 78.268],\n [0.017, 1.006, 0.973, 1.059, 1.501, 0.863]),\n 'thompson': ([0.386, 25.513, 53.571, 69.967, 81.050, 87.947],\n [0.030, 1.179, 0.514, 0.938, 0.238, 0.201]),\n 'ucb': ([0.392, 28.151, 62.949, 83.239, 91.105, 94.764],\n [0.030, 0.321, 0.929, 0.701, 0.132, 0.118])},\n 'smis': {'ei': ([0.394, 22.063, 40.365, 58.145, 66.830, 75.390],\n [0.033, 1.925, 2.207, 1.545, 7.858, 5.355]),\n 'greedy': ([0.384, 29.165, 60.514, 75.141, 81.341, 87.851],\n [0.022, 1.213, 2.087, 1.613, 3.184, 2.327]),\n 'pi': ([0.378, 23.490, 44.446, 59.433, 70.473, 78.219],\n [0.017, 1.010, 0.972, 1.057, 1.499, 0.861]),\n 'thompson': ([0.386, 25.493, 53.534, 69.916, 80.993, 87.906],\n [0.030, 1.178, 0.515, 0.936, 0.234, 0.199]),\n 'ucb': ([0.391, 28.129, 62.905, 83.183, 91.061, 94.739],\n [0.029, 0.323, 0.926, 0.700, 0.129, 0.117])}},\n 'nn': {'avg': {'ei': ([60.506, 94.133, 95.036, 95.889, 96.540, 96.903],\n [0.079, 0.237, 0.376, 0.061, 0.315, 0.162]),\n 'greedy': ([60.496, 94.842, 97.097, 98.104, 98.633, 98.942],\n [0.024, 0.109, 0.037, 0.025, 0.058, 0.076]),\n 'pi': ([60.496, 94.195, 95.179, 96.362, 96.669, 97.075],\n [0.029, 0.103, 0.020, 0.208, 0.164, 0.167]),\n 'thompson': ([60.475, 94.145, 96.768, 97.878, 98.488, 98.923],\n [0.083, 0.106, 0.039, 0.025, 0.068, 0.055]),\n 'ucb': ([60.437, 94.247, 96.720, 97.910, 98.566, 98.646],\n [0.038, 0.404, 0.146, 0.030, 0.015, 0.074])},\n 'scores': {'ei': ([0.399, 20.869, 25.993, 32.073, 38.081, 41.799],\n [0.011, 1.220, 2.474, 0.531, 3.171, 1.754]),\n 'greedy': ([0.395, 25.126, 44.356, 58.696, 68.289, 74.691],\n [0.021, 0.780, 0.432, 0.337, 0.905, 1.441]),\n 'pi': ([0.415, 21.327, 26.951, 36.413, 39.425, 43.917],\n [0.027, 0.470, 0.118, 2.067, 1.648, 2.065]),\n 'thompson': ([0.416, 20.933, 40.499, 54.969, 65.277, 73.772],\n [0.021, 0.554, 0.404, 0.510, 1.314, 1.230]),\n 'ucb': ([0.395, 21.653, 40.113, 55.651, 66.853, 68.367],\n [0.017, 2.239, 1.661, 0.520, 0.254, 1.449])},\n 'smis': {'ei': ([0.399, 20.844, 25.964, 32.041, 38.043, 41.758],\n [0.011, 1.220, 2.470, 0.530, 3.167, 1.751]),\n 'greedy': ([0.395, 25.100, 44.319, 58.648, 68.234, 74.630],\n [0.021, 0.779, 0.434, 0.333, 0.901, 1.440]),\n 'pi': ([0.415, 21.309, 26.930, 36.387, 39.396, 43.883],\n [0.027, 0.470, 0.119, 2.067, 1.647, 2.067]),\n 'thompson': ([0.416, 20.913, 40.463, 54.925, 65.225, 73.713],\n [0.021, 0.559, 0.405, 0.507, 1.313, 1.229]),\n 'ucb': ([0.395, 21.632, 40.079, 55.607, 66.797, 68.311],\n [0.017, 2.235, 1.662, 0.520, 0.253, 1.448])}},\n 'rf': {'avg': {'ei': ([60.494, 91.279, 92.061, 92.995, 94.211, 95.468],\n [0.056, 0.185, 0.220, 0.097, 0.165, 0.592]),\n 'greedy': ([60.527, 94.457, 96.494, 97.626, 98.329, 98.789],\n [0.039, 0.119, 0.239, 0.156, 0.179, 0.130]),\n 'pi': ([60.491, 91.360, 91.996, 93.007, 93.592, 95.032],\n [0.051, 0.082, 0.313, 0.594, 0.698, 0.710]),\n 'thompson': ([60.466, 93.764, 96.031, 97.454, 98.158, 98.779],\n [0.031, 0.105, 0.090, 0.068, 0.155, 0.100]),\n 'ucb': ([60.457, 93.274, 94.043, 95.781, 96.885, 97.465],\n [0.029, 0.374, 0.430, 0.500, 0.651, 0.582])},\n 'scores': {'ei': ([0.384, 11.603, 13.287, 16.153, 21.417, 29.141],\n [0.029, 0.825, 0.730, 0.341, 0.636, 4.427]),\n 'greedy': ([0.375, 21.977, 37.721, 51.615, 62.747, 71.376],\n [0.015, 0.850, 2.496, 2.075, 2.715, 2.077]),\n 'pi': ([0.373, 11.414, 13.066, 16.087, 18.525, 26.371],\n [0.021, 0.338, 1.070, 2.637, 3.363, 4.693]),\n 'thompson': ([0.437, 19.010, 33.323, 49.008, 59.905, 71.694],\n [0.031, 0.531, 0.814, 0.941, 2.611, 1.943]),\n 'ucb': ([0.373, 17.559, 20.793, 31.487, 42.086, 49.165],\n [0.027, 1.650, 2.073, 4.080, 7.028, 7.737])},\n 'smis': {'ei': ([0.384, 11.599, 13.283, 16.147, 21.407, 29.125],\n [0.029, 0.824, 0.730, 0.339, 0.635, 4.422]),\n 'greedy': ([0.374, 21.957, 37.689, 51.573, 62.699, 71.321],\n [0.014, 0.850, 2.494, 2.076, 2.715, 2.075]),\n 'pi': ([0.372, 11.407, 13.059, 16.077, 18.514, 26.357],\n [0.022, 0.339, 1.071, 2.637, 3.364, 4.693]),\n 'thompson': ([0.437, 18.996, 33.295, 48.973, 59.863, 71.644],\n [0.031, 0.530, 0.812, 0.939, 2.609, 1.941]),\n 'ucb': ([0.373, 17.549, 20.780, 31.467, 42.059, 49.133],\n [0.027, 1.649, 2.071, 4.074, 7.022, 7.728])}}}", "_____no_output_____" ], [ "all_results = {\n '10k': {\n 1.0: {'online': E10k_online,\n 'retrain': E10k_retrain,\n 'random': E10k_random},\n 'size': 10560,\n 'topk': 100,\n 'y_min': 75\n },\n '50k': {\n 1.0: {'online': E50k_online,\n 'retrain': E50k_retrain,\n 'random': E50k_random},\n 'size': 50240,\n 'topk': 500,\n 'y_min': 75\n },\n 'HTS': {\n 0.1: {'online': HTS_001_online,\n 'retrain': HTS_001_retrain,\n 'random': HTS_001_random},\n 0.2: {'online': HTS_002_online,\n 'retrain': HTS_002_retrain,\n 'random': HTS_002_random},\n 0.4: {'online': HTS_004_online,\n 'retrain': HTS_004_retrain,\n 'random': HTS_004_random},\n 'size': 2141514,\n 'topk': 1000,\n 'y_min': 75\n },\n 'AmpC': {\n 0.1: {'online': AmpC_001_online,\n 'retrain': AmpC_001_retrain,\n 'random': AmpC_001_random},\n 0.2: {'online': AmpC_002_online,\n 'retrain': AmpC_002_retrain,\n 'random': AmpC_002_random},\n 0.4: {'online': AmpC_004_online,\n 'retrain': AmpC_004_retrain,\n 'random': AmpC_004_random},\n 'size': 99459561,\n 'topk': 50000,\n 'y_min': 45\n }\n}", "_____no_output_____" ], [ "# 0.1/0.1 split, top-1k\nHTS_convergence = {\n 'mpn': {'average': [76.890, 95.170, 98.299, 98.927, 99.275, 99.415, 99.539, 99.629],\n 'smis': [0.0, 22.4, 50.0, 62.9, 72.4, 77.1, 81.4, 84.5],\n 'scores': [0.0, 22.8, 51.8, 65.4, 76.0, 81.1, 85.9, 88.7]},\n 'nn': {'average': [76.911, 94.357, 97.226, 98.204, 98.810, 99.018, 99.329, 99.445, 99.530],\n 'smis': [0.0, 14.9, 34.6, 48.3, 60.4, 65.8, 73.9, 77.8, 80.6],\n 'scores': [0.0, 15.7, 36.3, 50.7, 63.2, 69.3, 78.0, 82.2, 85.4]},\n 'rf': {'average': [77.008, 92.383, 96.318, 97.508, 98.300, 98.766, 98.946, 99.154, 99.264, 99.320],\n 'smis': [0.0, 9.0, 27.0, 37.6, 48.6, 57.9, 62.4, 67.2, 70.0, 71.9],\n 'scores': [0.0, 9.2, 27.5, 38.8, 50.4, 60.4, 65.1, 70.7, 73.8, 75.8]}}", "_____no_output_____" ], [ "HTS_common_smis_by_iter_online = {\n 0.1: {'mpn': {'ei': [0, 287, 410, 475, 542, 557],\n 'greedy': [0, 314, 359, 567, 591, 658],\n 'pi': [0, 291, 458, 500, 533, 542],\n 'thompson': [0, 8, 121, 365, 492, 616],\n 'ucb': [0, 320, 448, 572, 589, 690]},\n 'nn': {'ei': [0, 76, 69, 54, 57, 77],\n 'greedy': [0, 202, 314, 464, 569, 597],\n 'pi': [0, 58, 105, 289, 309, 312],\n 'thompson': [0, 7, 68, 99, 167, 182],\n 'ucb': [0, 76, 98, 127, 127, 180]},\n 'rf': {'ei': [0, 0, 0, 0, 0, 3],\n 'greedy': [0, 20, 41, 109, 160, 187],\n 'pi': [0, 0, 3, 21, 85, 98],\n 'thompson': [0, 0, 0, 7, 57, 109],\n 'ucb': [0, 10, 87, 82, 82, 85]}},\n 0.2: {'mpn': {'ei': [0, 404, 551, 636, 640, 651],\n 'greedy': [0, 362, 577, 710, 727, 748],\n 'pi': [0, 393, 569, 646, 651, 644],\n 'thompson': [0, 15, 290, 551, 643, 722],\n 'ucb': [0, 385, 583, 701, 708, 746]},\n 'nn': {'ei': [0, 145, 146, 142, 154, 162],\n 'greedy': [0, 320, 490, 647, 733, 767],\n 'pi': [0, 151, 154, 147, 159, 171],\n 'thompson': [0, 17, 123, 291, 456, 552],\n 'ucb': [0, 220, 384, 376, 495, 502]},\n 'rf': {'ei': [0, 1, 2, 9, 12, 13],\n 'greedy': [0, 137, 219, 485, 529, 566],\n 'pi': [0, 1, 6, 20, 48, 58],\n 'thompson': [0, 0, 5, 40, 186, 241],\n 'ucb': [0, 0, 25, 68, 100, 300]}},\n 0.4: {'mpn': {'ei': [0, 567, 594, 654, 689, 713],\n 'greedy': [0, 649, 709, 831, 888, 918],\n 'pi': [0, 583, 622, 680, 715, 730],\n 'thompson': [0, 74, 505, 745, 854, 912],\n 'ucb': [0, 646, 749, 845, 900, 930]},\n 'nn': {'ei': [0, 223, 251, 364, 417, 446],\n 'greedy': [0, 519, 719, 759, 817, 866],\n 'pi': [0, 245, 251, 338, 291, 287],\n 'thompson': [0, 66, 206, 320, 401, 429],\n 'ucb': [0, 385, 460, 595, 671, 772]},\n 'rf': {'ei': [0, 3, 9, 65, 118, 210],\n 'greedy': [0, 432, 538, 628, 666, 658],\n 'pi': [0, 75, 103, 135, 189, 210],\n 'thompson': [0, 2, 119, 273, 375, 384],\n 'ucb': [0, 23, 50, 187, 381, 420]}}\n}\n\nHTS_common_smis_by_iter_retrain = {\n 0.1: {'mpn': {'ei': [0, 253, 446, 574, 626, 689],\n 'greedy': [0, 303, 576, 676, 758, 791],\n 'pi': [0, 298, 512, 616, 656, 715],\n 'thompson': [0, 6, 115, 283, 440, 601],\n 'ucb': [0, 296, 493, 660, 743, 773]},\n 'nn': {'ei': [0, 74, 150, 204, 313, 366],\n 'greedy': [0, 201, 478, 668, 734, 779],\n 'pi': [0, 80, 237, 331, 367, 394],\n 'thompson': [0, 10, 180, 383, 562, 649],\n 'ucb': [0, 138, 368, 539, 678, 679]},\n 'rf': {'ei': [0, 0, 0, 4, 8, 23],\n 'greedy': [0, 5, 51, 184, 269, 452],\n 'pi': [0, 0, 6, 74, 103, 116],\n 'thompson': [0, 0, 34, 209, 267, 348],\n 'ucb': [0, 0, 119, 193, 293, 309]}},\n 0.2: {'mpn': {'ei': [0, 263, 570, 675, 704, 776],\n 'greedy': [0, 498, 708, 755, 820, 894],\n 'pi': [0, 327, 622, 688, 737, 797],\n 'thompson': [0, 10, 279, 549, 643, 757],\n 'ucb': [0, 475, 711, 762, 843, 894]},\n 'nn': {'ei': [0, 140, 301, 361, 435, 452],\n 'greedy': [0, 434, 704, 815, 828, 871],\n 'pi': [0, 132, 275, 353, 428, 453],\n 'thompson': [0, 26, 276, 516, 682, 748],\n 'ucb': [0, 243, 567, 733, 753, 800]},\n 'rf': {'ei': [0, 1, 23, 97, 106, 137],\n 'greedy': [0, 42, 269, 387, 604, 724],\n 'pi': [0, 7, 25, 77, 86, 106],\n 'thompson': [0, 0, 34, 303, 362, 478],\n 'ucb': [0, 8, 252, 429, 478, 510]}},\n 0.4: {'mpn': {'ei': [0, 587, 667, 795, 867, 902],\n 'greedy': [0, 618, 804, 911, 953, 963],\n 'pi': [0, 534, 709, 816, 871, 915],\n 'thompson': [0, 74, 531, 752, 877, 919],\n 'ucb': [0, 632, 816, 925, 943, 935]},\n 'nn': {'ei': [0, 238, 569, 575, 675, 715],\n 'greedy': [0, 570, 827, 869, 951, 978],\n 'pi': [0, 268, 425, 473, 538, 622],\n 'thompson': [0, 84, 474, 722, 882, 939],\n 'ucb': [0, 417, 709, 795, 855, 926]},\n 'rf': {'ei': [0, 27, 46, 95, 186, 226],\n 'greedy': [0, 189, 441, 733, 729, 795],\n 'pi': [0, 46, 101, 165, 264, 331],\n 'thompson': [0, 4, 179, 399, 556, 634],\n 'ucb': [0, 81, 312, 438, 478, 554]}}\n}", "_____no_output_____" ], [ "E10k_union_smis_retrain = {\n 'mpn': {'ei': [519, 861, 1140, 1311, 1471, 1600],\n 'greedy': [514, 772, 1042, 1249, 1403, 1578],\n 'pi': [508, 800, 1069, 1236, 1393, 1536],\n 'thompson': [512, 895, 1195, 1435, 1648, 1822],\n 'ucb': [506, 828, 1100, 1305, 1478, 1640]},\n 'nn': {'ei': [519, 923, 1186, 1452, 1765, 2070],\n 'greedy': [515, 910, 1172, 1395, 1570, 1732],\n 'pi': [519, 926, 1234, 1492, 1788, 2001],\n 'thompson': [513, 918, 1218, 1456, 1627, 1775],\n 'ucb': [515, 926, 1179, 1412, 1583, 1744]},\n 'rf': {'ei': [510, 985, 1360, 1685, 1984, 2241],\n 'greedy': [516, 941, 1281, 1520, 1742, 1919],\n 'pi': [510, 968, 1319, 1568, 1820, 2075],\n 'thompson': [514, 999, 1427, 1835, 2172, 2507],\n 'ucb': [519, 996, 1370, 1657, 1868, 2095]}\n}\n\nE50k_union_smis_retrain = {\n 'mpn': {'ei': [2442, 3777, 4680, 5500, 6329, 7119],\n 'greedy': [2438, 3686, 4585, 5461, 6223, 7032],\n 'pi': [2437, 3729, 4654, 5465, 6267, 7161],\n 'thompson': [2434, 4344, 5622, 6411, 7247, 7870],\n 'ucb': [2440, 3668, 4637, 5411, 6258, 7023]},\n 'nn': {'ei': [2443, 4015, 5104, 6414, 7571, 8679],\n 'greedy': [2437, 3961, 4769, 5431, 6022, 6544],\n 'pi': [2436, 3959, 5150, 6560, 7648, 8761],\n 'thompson': [2462, 4227, 5170, 5818, 6437, 6959],\n 'ucb': [2434, 4053, 4925, 5580, 6164, 6696]},\n 'rf': {'ei': [2451, 4602, 6181, 7481, 8732, 9893],\n 'greedy': [2445, 4131, 4934, 5642, 6428, 7106],\n 'pi': [2438, 4386, 5628, 6950, 8240, 9483],\n 'thompson': [2445, 4687, 6584, 8239, 9587, 10836],\n 'ucb': [2430, 4345, 5578, 6950, 8034, 9116]}\n}", "_____no_output_____" ], [ "HTS_intersection_smis_online = {\n 0.1: {'mpn': {'ei': [10686, 15707, 20790, 26448, 32863, 38821],\n 'greedy': [10682, 15360, 21985, 26185, 32611, 37807],\n 'pi': [10691, 15775, 21199, 26991, 32398, 38013],\n 'thompson': [10689, 19789, 25916, 30591, 35279, 39820],\n 'ucb': [10682, 15718, 21350, 25570, 32154, 36446]},\n 'nn': {'ei': [10686, 17636, 27806, 35279, 44661, 52823],\n 'greedy': [10693, 15984, 21439, 26856, 31478, 36426],\n 'pi': [10686, 17622, 26732, 32948, 42245, 48641],\n 'thompson': [10691, 19028, 26577, 33758, 40633, 48483],\n 'ucb': [10683, 17515, 25787, 33519, 41550, 48461]},\n 'rf': {'ei': [10684, 20463, 30709, 40230, 49149, 57727],\n 'greedy': [10687, 18885, 26634, 32087, 40330, 46737],\n 'pi': [10692, 20342, 30102, 38600, 47056, 55368],\n 'thompson': [10681, 20919, 30827, 40026, 49158, 57738],\n 'ucb': [10689, 19597, 28477, 37797, 46996, 55803]}},\n 0.2: {'mpn': {'ei': [21330, 30905, 40827, 51900, 65390, 76970],\n 'greedy': [21331, 31115, 43322, 51150, 62549, 70862],\n 'pi': [21347, 31009, 40982, 52101, 63631, 75153],\n 'thompson': [21332, 37797, 47820, 55846, 64280, 72898],\n 'ucb': [21345, 30789, 42967, 51034, 62016, 71688]},\n 'nn': {'ei': [21339, 33855, 54021, 71242, 85249, 97256],\n 'greedy': [21350, 30195, 39989, 49518, 57577, 66290],\n 'pi': [21325, 33492, 53654, 70341, 88444, 97081],\n 'thompson': [21336, 36534, 48142, 59066, 68633, 78866],\n 'ucb': [21348, 32184, 44552, 56964, 68075, 82882]},\n 'rf': {'ei': [21334, 39097, 59278, 76728, 93197, 112255],\n 'greedy': [21342, 33917, 48247, 59901, 74441, 87288],\n 'pi': [21343, 39641, 59651, 76862, 93724, 110352],\n 'thompson': [21322, 41135, 60442, 76552, 93343, 109832],\n 'ucb': [21330, 38829, 56877, 71586, 89522, 104610]}},\n 0.4: {'mpn': {'ei': [42472, 60886, 85271, 107016, 128315, 150345],\n 'greedy': [42497, 59653, 81235, 97256, 115706, 130478],\n 'pi': [42483, 60472, 84103, 108413, 130376, 150972],\n 'thompson': [42508, 71672, 90374, 105287, 120655, 133771],\n 'ucb': [42490, 60021, 79660, 96576, 116595, 132428]},\n 'nn': {'ei': [42522, 65769, 105401, 125242, 159336, 171939],\n 'greedy': [42489, 57089, 72717, 89684, 106228, 123014],\n 'pi': [42496, 65302, 105648, 131060, 158766, 172462],\n 'thompson': [42519, 68998, 92833, 119993, 144478, 171245],\n 'ucb': [42494, 61335, 89349, 107954, 136428, 150832]},\n 'rf': {'ei': [42508, 75895, 114197, 142269, 174207, 205535],\n 'greedy': [42505, 63177, 89032, 109855, 137817, 162966],\n 'pi': [42512, 72402, 105881, 136261, 170318, 200344],\n 'thompson': [42473, 77295, 113569, 141028, 170529, 199158],\n 'ucb': [42535, 70928, 107662, 134741, 169029, 200558]}}\n}\n\nHTS_intersection_smis_retrain = {\n 0.1: {'mpn': {'ei': [10688, 15720, 20010, 23635, 28248, 32577],\n 'greedy': [10691, 15694, 19327, 23105, 26720, 29849],\n 'pi': [10691, 15754, 19792, 23738, 27741, 31631],\n 'thompson': [10697, 19679, 25714, 30499, 34974, 38645],\n 'ucb': [10686, 15626, 19579, 23216, 26540, 29878]},\n 'nn': {'ei': [10695, 17595, 24872, 33185, 41013, 49236],\n 'greedy': [10690, 16108, 19271, 21993, 24378, 27369],\n 'pi': [10686, 17580, 24365, 32492, 40787, 48178],\n 'thompson': [10675, 18890, 23187, 26574, 29027, 31806],\n 'ucb': [10686, 17165, 21434, 24970, 27540, 31772]},\n 'rf': {'ei': [10681, 21005, 30020, 38782, 47135, 55790],\n 'greedy': [10686, 19151, 24234, 28787, 31887, 35929],\n 'pi': [10690, 20325, 27737, 35464, 44363, 53039],\n 'thompson': [10688, 20988, 29592, 37123, 44302, 50846],\n 'ucb': [10692, 20641, 27850, 35624, 41939, 50688]}},\n 0.2: {'mpn': {'ei': [21341, 32616, 40030, 48217, 57254, 65806],\n 'greedy': [21353, 30415, 37947, 44542, 50942, 57041],\n 'pi': [21338, 32203, 39701, 47931, 56763, 64843],\n 'thompson': [21343, 38634, 48668, 56850, 64222, 71360],\n 'ucb': [21344, 30433, 37861, 44686, 51380, 57881]},\n 'nn': {'ei': [21330, 33790, 48847, 66135, 81600, 97863],\n 'greedy': [21352, 29598, 34802, 40432, 45217, 50382],\n 'pi': [21336, 33700, 49855, 66010, 81079, 95444],\n 'thompson': [21325, 36307, 43935, 49119, 53282, 58230],\n 'ucb': [21348, 32032, 38783, 45160, 52772, 59558]},\n 'rf': {'ei': [21336, 39771, 57288, 73902, 91191, 107190],\n 'greedy': [21334, 35859, 45486, 52240, 56539, 62690],\n 'pi': [21352, 38455, 55276, 71233, 88634, 106287],\n 'thompson': [21338, 40888, 56793, 69270, 80600, 90675],\n 'ucb': [21325, 37870, 51109, 64817, 78571, 91339]}},\n 0.4: {'mpn': {'ei': [42510, 60711, 79150, 96571, 112905, 129702],\n 'greedy': [42490, 60267, 73209, 85490, 97573, 108971],\n 'pi': [42495, 61570, 78156, 95570, 113005, 127987],\n 'thompson': [42496, 72017, 87698, 102635, 117048, 130359],\n 'ucb': [42499, 59736, 73507, 86779, 99818, 111995]},\n 'nn': {'ei': [42500, 65546, 98218, 128223, 152412, 178024],\n 'greedy': [42500, 57106, 66708, 75909, 85261, 95448],\n 'pi': [42462, 64925, 97658, 128572, 153130, 181438],\n 'thompson': [42490, 67841, 80125, 88336, 97228, 105883],\n 'ucb': [42477, 61066, 74259, 84028, 101318, 111424]},\n 'rf': {'ei': [42480, 71782, 102182, 135475, 166549, 197109],\n 'greedy': [42500, 64980, 82142, 94578, 106366, 116845],\n 'pi': [42510, 74537, 106604, 136851, 166823, 195612],\n 'thompson': [42520, 77848, 101661, 121834, 140229, 158613],\n 'ucb': [42491, 71171, 95932, 120656, 145471, 166648]}}\n}", "_____no_output_____" ], [ "smis_results ={\n '10k': {\n 'union': E10k_union_smis_retrain,\n },\n '50k': {\n 'union': E50k_union_smis_retrain,\n },\n 'HTS' : {\n 'intersection': {\n 'online': HTS_intersection_smis_online,\n 'retrain': HTS_intersection_smis_retrain\n },\n 'union': {\n 'online': HTS_union_smis_online,\n 'retrain': HTS_union_smis_retrain\n }\n }\n}", "_____no_output_____" ], [ "one_shot_results = {\n 2: HTS_02_004,\n 0.4: HTS_004_02\n}", "_____no_output_____" ] ], [ [ "#### RUN ALL CELLS ABOVE ME\n\nNow we're ready to make some figures and nicer data presentation", "_____no_output_____" ], [ "## Writing CSVs of the full data\n\nthese two functions allow you to write a formatted CSV of the full data that you just loaded in above.", "_____no_output_____" ], [ "The first function will write the final results for a given library and batch size (including both online and full retraining results as well as data for the top-k SMILES and top-k average metrics.", "_____no_output_____" ] ], [ [ "def write_final_results_csv(library, batch_size):\n lib_results = all_results[library][batch_size]\n results_df = []\n for training in ['retrain', 'online']:\n results = lib_results[training]\n for model in MODELS:\n for metric in METRICS:\n if metric == 'greedy':\n metric_ = metric.capitalize()\n elif metric == 'thompson':\n metric_ = 'TS'\n else:\n metric_ = metric.upper()\n\n scores = results[model]['scores'][metric]\n smis = results[model]['smis'][metric]\n avg = results[model]['avg'][metric]\n results_df.append({\n 'Training': training,\n 'Model': model.upper(),\n 'Metric': metric_,\n 'Scores ($\\pm$ s.d.)': f'{scores[0][-1]:0.1f} ({scores[1][-1]:0.1f})',\n 'SMILES ($\\pm$ s.d.)': f'{smis[0][-1]:0.1f} ({smis[1][-1]:0.1f})',\n 'Average ($\\pm$ s.d.)': f'{avg[0][-1]:0.2f} ({avg[1][-1]:0.2f})'\n })\n df = pd.DataFrame(results_df).set_index(['Training', 'Model', 'Metric'])\n df.to_csv(f'{library}_{batch_size}_final_results.csv')\n return df", "_____no_output_____" ] ], [ [ "The second function will write the results of each iteration for a given library, batch size, and scoring metric (including both online and full retraining results).", "_____no_output_____" ] ], [ [ "def write_results_csv_by_iter(library, batch_size, score_mode):\n lib_results = all_results[library][batch_size]\n results_df = []\n for training in ['retrain', 'online']:\n results = lib_results[training]\n for model in MODELS:\n for metric in METRICS:\n if metric == 'greedy':\n metric_ = metric.capitalize()\n elif metric == 'thompson':\n metric_ = 'TS'\n else:\n metric_ = metric.upper()\n\n means = results[model][score_mode][metric][0]\n sds = results[model][score_mode][metric][1]\n row = {\n 'Training': training,\n 'Model': model.upper(),\n 'Metric': metric_,\n }\n for i, (mean, sd) in enumerate(zip(means, sds)):\n row[f'iter_{i} ($\\pm$ s.d.)'] = f'{mean:0.1f} ({sd:0.1f})'\n results_df.append(row)\n df = pd.DataFrame(results_df).set_index(['Training', 'Model', 'Metric'])\n df.to_csv(f'{library}_{batch_size}_{score_mode}_results_full.csv', index=False)\n return df", "_____no_output_____" ] ], [ [ "running the cell below will generate the CSVs for both the full dataset of each experiment for each score metric and the abbreviated (final) results of each experiment. If you only want one or the other, comment out the appropriate lines", "_____no_output_____" ] ], [ [ "for library in ['10k', '50k']:\n write_final_results_csv(library, 1.0)\n for score_mode in ['scores', 'smis', 'avg']:\n write_results_csv_by_iter(library, 1.0, score_mode)\n\nfor library in ['HTS', 'AmpC']:\n for batch_size in [0.4, 0.2, 0.1]:\n write_final_results_csv(library, batch_size)\n for score_mode in ['scores', 'smis', 'avg']:\n write_results_csv_by_iter(library, batch_size, score_mode) ", "_____no_output_____" ] ], [ [ "## Generating the main text figures\n\n### Recreating the 10k and 50k figures\nthe below function will, for a given library and split generate a three-paneled figure, where each panel will correspond to a given surrogate model architecture and contain the traces corresponding to each acquisition metric. This function was used to generate the 10k and 50k figures in the main text. By default, this function only produces results of full model retraining. If si_fig is set to true, it will show the results on online model training at full opacity and a faded trace of full model retraining as well.", "_____no_output_____" ] ], [ [ "def gen_metrics_figure(library, split, si_fig=False):\n score = 'scores'\n\n retrain_results = all_results[library][split]['retrain']\n online_results = all_results[library][split]['online']\n random_results = all_results[library][split]['random']\n size = all_results[library]['size']\n topk = all_results[library]['topk']\n y_min = all_results[library]['y_min']\n\n results_series = make_subplots(\n rows=1, cols=len(retrain_results),\n shared_xaxes=True, shared_yaxes=True,\n x_title='Number of Ligands Explored', y_title=f'Percentage of Top-{topk} Scores Found',\n subplot_titles=[model.upper() for model in MODELS]\n )\n\n xs = [int(size * split/100 * i) for i in range(1, 7)]\n\n for i, model in enumerate(MODELS):\n for j, metric in enumerate(METRICS):\n # full retrain trace\n ys, y_sds = retrain_results[model][score][metric]\n results_series.add_trace(go.Scatter(\n x=xs, y=ys, opacity=0.33 if si_fig else 1.,\n error_y=dict(type='data', array=y_sds, visible=not si_fig),\n marker=dict(color=METRIC_COLORS[j]),\n mode='lines+markers', name=METRIC_NAMES[metric],\n legendgroup=metric, showlegend=(model=='rf' and not si_fig)\n ), row=1, col=i+1)\n \n # online train trace\n if si_fig:\n ys, y_sds = online_results[model][score][metric]\n results_series.add_trace(go.Scatter(\n x=xs, y=ys,\n error_y=dict(type='data', array=y_sds, visible=True),\n marker=dict(color=METRIC_COLORS[j]),\n mode='lines+markers', name=METRIC_NAMES[metric],\n legendgroup=metric, showlegend=(model=='rf')\n ), row=1, col=i+1)\n \n ys, y_sds = random_results[score]\n metric_trace = go.Scatter(\n x=xs, y=ys, \n error_y=dict(type='data', array=y_sds, visible=True),\n marker=dict(color='slategray'),\n mode='lines+markers',\n name='random', legendgroup='random',\n showlegend=model=='rf',\n )\n results_series.add_trace(metric_trace, row=1, col=i+1)\n \n results_series.update_xaxes(row=1, col=i+1, \n rangemode='tozero', nticks=10)\n \n border_params = dict(\n showgrid=True, zeroline=False, showticklabels=True, \n visible=True, mirror=True, linewidth=2,\n linecolor='black'\n )\n results_series.update_xaxes(**border_params, tickangle=-35)\n results_series.update_yaxes(**border_params)\n \n results_series['layout']['legend']['title']['text'] = 'Metric'\n results_series.update_traces(\n marker=dict(symbol='circle', line_width=2, size=7.5),\n #line=dict(dash='solid')\n )\n \n for i in results_series['layout']['annotations']:\n i['font'] = dict(color='black', size=20, family='sans-serif')\n results_series['layout']['annotations'][-2]['yanchor']='bottom'\n results_series['layout']['annotations'][-2]['y'] = -0.15\n HEIGHT = 500\n results_series.update_layout(\n legend_title_text='Metric', #legend_traceorder='reversed',\n height=HEIGHT, width=HEIGHT/1.5*3,\n font=dict(color='black', size=18, family='sans-serif'),\n )\n return results_series\n\ndef gen_10k50k_figure(library, si_fig):\n return gen_metrics_figure(library, 1., si_fig)", "_____no_output_____" ] ], [ [ "The below cell will generate the 10k and 50k figures from the main text. Setting the second argument to `True` in the function calls will produce the 10k and 50k SI figures.", "_____no_output_____" ] ], [ [ "fig = gen_10k50k_figure('10k', False)\nfig.show()\nfig = gen_10k50k_figure('50k', False)\nfig.show()", "_____no_output_____" ] ], [ [ "### Recreating the HTS and AmpC figures\n\nThe HTS and AmpC figures used a different figure design, where each panel now corresponds to a given initialization/exploration batch size and each panel contains the results of all three models' results with a greedy acquisition metric. The function below will generate those figures. To run these functions, go to the cell below.", "_____no_output_____" ] ], [ [ "def gen_HTSAmpC_figure_main(library):\n score = 'scores'\n size = all_results[library]['size']\n topk = all_results[library]['topk']\n y_min = all_results[library]['y_min']\n split_results = all_results[library]\n \n results_series = make_subplots(\n rows=1, cols=len(SPLITS), shared_yaxes=True,\n x_title='Number of Ligands Explored', y_title=f'Percentage of Top-{topk} Scores Found',\n subplot_titles=[f'{split}%' for split in SPLITS])\n\n for i, split in enumerate(SPLITS):\n retrain_results = split_results[split]['retrain']\n online_results = split_results[split]['retrain']\n random_results = split_results[split]['random']\n xs = [int(size * split/100 * i) for i in range(1, 7)]\n \n ys, y_sds = random_results[score]\n random_trace = go.Scatter(\n x=xs, y=ys, \n error_y=dict(type='data', array=y_sds, visible=True),\n mode='lines+markers',\n marker=dict(symbol='circle'),\n line=dict(color='slategray'),\n name='random', legendgroup='random',\n showlegend=split==0.1,\n )\n results_series.add_trace(random_trace, row=1, col=i+1, )\n \n for j, model in enumerate(MODELS):\n ys, y_sds = retrain_results[model][score]['greedy']\n\n split_trace = go.Scatter(\n x=xs, y=ys,\n error_y=dict(type='data', array=y_sds, visible=True),\n mode='lines+markers',\n marker=dict(symbol=MARKERS[j], color=MODEL_COLORS[j]),\n name=model.upper(), legendgroup=model,\n showlegend=split==0.1\n )\n results_series.add_trace(split_trace, row=1, col=i+1)\n \n results_series.update_xaxes(row=1, col=i+1, \n rangemode='tozero', nticks=10)\n \n border_params = dict(\n showgrid=True, zeroline=False, showticklabels=True, \n visible=True, mirror=True, linewidth=2,\n linecolor='black'\n )\n results_series.update_xaxes(**border_params, tickangle=-35)\n results_series.update_yaxes(**border_params)\n results_series.update_traces(marker=dict(line_width=2, size=7.5))\n \n for i in results_series['layout']['annotations']:\n i['font'] = dict(color='black', size=20, family='sans-serif')\n results_series['layout']['annotations'][-2]['yanchor']='bottom'\n results_series['layout']['annotations'][-2]['y'] = -0.15\n HEIGHT = 500\n results_series.update_layout(\n legend_title_text='Model', legend_traceorder='reversed',\n height=HEIGHT, width=HEIGHT/1.5*3,\n font=dict(color='black', size=20, family='sans-serif')\n )\n \n return results_series", "_____no_output_____" ] ], [ [ "This cell will generate the HTS and AmpC figures from the main text.", "_____no_output_____" ] ], [ [ "fig = gen_HTSAmpC_figure_main('AmpC')\nfig.show()\nfig = gen_HTSAmpC_figure_main('HTS')\nfig.show()", "_____no_output_____" ] ], [ [ "### Recreating the Single-iteration and Convergence Figures\n\nthe two cells below contain the functions to generate these figures, and the third cell will actually create the figures", "_____no_output_____" ] ], [ [ "def generate_one_shot_figure():\n fig = go.Figure()\n fig.update_layout(legend_title_text='Model')\n size = 2141514\n\n for i, model in enumerate(MODELS):\n # 0.2/0.04\n xs = [size*0.02, size*0.024]\n ys, y_sds = one_shot_results[2][model]['scores']['greedy']\n fig.add_trace(go.Scatter(\n x=xs, y=ys, \n error_y=dict(type='data', array=y_sds, visible=True),\n line=dict(dash='dash'),\n marker=dict(symbol=MARKERS[i], color=MODEL_COLORS[i]),\n mode='lines+markers',\n name=model.upper(), legendgroup=model\n ))\n \n # 0.04/0.2\n xs = [size*0.004, size*0.024]\n ys, y_sds = one_shot_results[0.4][model]['scores']['greedy']\n fig.add_trace(go.Scatter(\n x=xs, y=ys, \n error_y=dict(type='data', array=y_sds, visible=True),\n line=dict(dash='solid'),\n marker=dict(symbol=MARKERS[i], color=MODEL_COLORS[i]),\n mode='lines+markers', showlegend=False,\n name=model, legendgroup=model\n ))\n # AL trace\n xs = [int(size * 0.4/100 * i) for i in range(1, 7)]\n ys, y_sds = HTS_004_retrain[model]['scores']['greedy']\n fig.add_trace(go.Scatter(\n x=xs, y=ys, opacity=0.5,\n line=dict(dash='solid'),\n marker=dict(symbol=MARKERS[i], color=MODEL_COLORS[i]),\n mode='lines+markers', showlegend=False,\n name=model, legendgroup=model\n ))\n \n \n border_params = dict(\n showgrid=True, zeroline=False, showticklabels=True, \n visible=True, mirror=True, linewidth=2,\n linecolor='black'\n )\n fig.update_yaxes(title_text=f'Percentage of Top-1000 Scores Found',\n tickfont=dict(color='black', size=18, family='sans-serif'),\n **border_params)\n fig.update_xaxes(title_text='Number of Ligands Explored',\n rangemode='tozero', nticks=10, tickangle=-35,\n tickfont=dict(color='black', size=18, family='sans-serif'),\n **border_params)\n fig.update_traces(mode='lines+markers', marker_line_width=2, marker_size=7.5)\n \n HEIGHT = 500\n fig.update_layout(\n legend_title_text='Model', legend_traceorder='reversed',\n height=HEIGHT, width=HEIGHT/1.25,\n font=dict(color='black', size=16, family='sans-serif')\n )\n \n return fig", "_____no_output_____" ], [ "def generate_convergence_figure(score): \n fig = go.Figure()\n fig.update_layout(legend_title_text='Model')\n \n size = 2141514\n \n for i, model in enumerate(MODELS):\n ys = HTS_convergence[model][score]\n xs = [int(size * 0.001 * i) for i in range(1, len(ys)+1)]\n fig.add_trace(go.Scatter(\n x=xs, y=ys, \n mode='lines+markers',\n marker=dict(symbol=MARKERS[i], color=MODEL_COLORS[i]),\n name=model.upper()\n ))\n \n border_params = dict(\n showgrid=True, zeroline=False, showticklabels=True, \n visible=True, mirror=True, linewidth=2,\n linecolor='black'\n )\n fig.update_yaxes(title_text=f'Percentage of Top-1000 {score.capitalize()} Found',\n tickfont=dict(color='black', size=18, family='sans-serif'),\n **border_params)\n fig.update_xaxes(title_text='Number of Ligands Explored',\n rangemode='tozero', tickangle=-35, nticks=11,\n tickfont=dict(color='black', size=18, family='sans-serif'),\n **border_params)\n fig.update_traces(mode='lines+markers', marker_line_width=2, marker_size=7.5)\n \n HEIGHT = 500\n fig.update_layout(\n legend_title_text='Model', legend_traceorder='reversed',\n height=HEIGHT, width=HEIGHT/1.25,\n font=dict(color='black', size=16, family='sans-serif')\n )\n \n return fig", "_____no_output_____" ], [ "fig = generate_one_shot_figure()\nfig.show()\n\nfig = generate_convergence_figure('scores')\nfig.show()", "_____no_output_____" ] ], [ [ "## Generating SI Figures\n\nthe cell below will generate the 10k, 50k, HTS, and AmpC SI figures", "_____no_output_____" ] ], [ [ "fig = gen_10k50k_figure('10k', False)\nfig.show()\nfig = gen_10k50k_figure('50k', False)\nfig.show()\n\nfor split in [0.4, 0.2, 0.1]:\n digit = str(split).split('.')[1]\n fig = gen_metrics_figure('AmpC', split, True)\n fig.write_image(f'figures/SI/AmpC_00{digit}_model_by_metric_online_with_retrain_faded.pdf')\n fig = gen_metrics_figure('HTS', split, True)\n fig.write_image(f'figures/SI/HTS_00{digit}_model_by_metric_online_with_retrain_faded.pdf')", "_____no_output_____" ] ], [ [ "### Union Plots\n\nthe two functions below were used to generate the union plots in the SI. The mode can also be set to `intersection`, although these figures were not used in the paper. Any metric may also be specificed, but only greedy data was shown in the paper.", "_____no_output_____" ] ], [ [ "def gen_10k50k_smis_figure(mode='union', metric='greedy'):\n \n if mode == 'union':\n y_title = 'Total Number of Molecules Explored Among Runs'\n elif mode == 'intersection':\n y_title = 'Number of Molecules Shared Among Runs'\n else:\n raise ValueError('Unsupported mode!')\n \n results_series = make_subplots(\n rows=1, cols=2,\n x_title='Iteration', y_title=y_title,\n subplot_titles=['10k', '50k'])\n\n xs = list(range(6))\n \n lib_results = smis_results['10k'][mode]\n size = 10560\n for j, model in enumerate(MODELS):\n ys = lib_results[model][metric]\n\n model_trace = go.Scatter(\n x=xs, y=ys,\n mode='lines+markers',\n marker=dict(symbol=MARKERS[j], color=MODEL_COLORS[j]),\n name=model.upper(), legendgroup=model,\n showlegend=True\n )\n results_series.add_trace(model_trace, row=1, col=1)\n if mode =='union':\n ys_upper = [int(5*(x+1)*1/100*size) for x in xs]\n results_series.add_trace(go.Scatter(\n x=xs, y=ys_upper,\n mode='lines', line=dict(color='black'),\n name='upper_bound', showlegend=False\n ), row=1, col=1)\n\n ys_lower = [int((5+x)*1/100*size) for x in xs]\n results_series.add_trace(go.Scatter(\n x=xs, y=ys_lower,\n mode='lines', line=dict(color='black'),\n name='lower_bound', showlegend=False\n ), row=1, col=1)\n\n results_series.update_yaxes(row=1, col=1, range=[0, ys_upper[-1]])\n elif mode == 'intersection':\n results_series.update_yaxes(row=1, col=1, range=[0, 100])\n results_series.update_xaxes(row=1, col=1, range=[-0, 5], dtick=1)\n \n lib_results = smis_results['50k'][mode]\n size = 50240\n for j, model in enumerate(MODELS):\n ys = lib_results[model][metric]\n\n model_trace = go.Scatter(\n x=xs, y=ys,\n mode='lines+markers',\n marker=dict(symbol=MARKERS[j], color=MODEL_COLORS[j]),\n name=model.upper(), legendgroup=model,\n showlegend=False\n )\n results_series.add_trace(model_trace, row=1, col=2)\n if mode =='union':\n ys_upper = [int(5*(x+1)*1/100*size) for x in xs]\n results_series.add_trace(go.Scatter(\n x=xs, y=ys_upper,\n mode='lines', line=dict(color='black'),\n name='upper_bound', showlegend=False\n ), row=1, col=2)\n\n ys_lower = [int((5+x)*1/100*size) for x in xs]\n results_series.add_trace(go.Scatter(\n x=xs, y=ys_lower,\n mode='lines', line=dict(color='black'),\n name='lower_bound', showlegend=False\n ), row=1, col=2)\n\n results_series.update_yaxes(row=1, col=2, range=[0, ys_upper[-1]])\n \n results_series.update_xaxes(row=1, col=2, range=[-0, 5], dtick=1)\n border_params = dict(\n showgrid=True, zeroline=False, showticklabels=True, \n visible=True, mirror=True, linewidth=2,\n linecolor='black'\n )\n results_series.update_xaxes(**border_params)\n results_series.update_yaxes(**border_params)\n results_series.update_traces(marker=dict(line_width=2, size=7.5))\n \n for i in results_series['layout']['annotations']:\n i['font'] = dict(color='black', size=20, family='sans-serif')\n results_series['layout']['annotations'][-2]['yanchor']='bottom'\n results_series['layout']['annotations'][-2]['y'] = -0.15\n results_series['layout']['annotations'][-1].x = -0.02\n HEIGHT = 500\n results_series.update_layout(\n legend_title_text='Model', legend_traceorder='reversed',\n height=HEIGHT, width=HEIGHT/1.5*2,\n font=dict(color='black', size=20, family='sans-serif')\n )\n \n return results_series", "_____no_output_____" ], [ "def gen_HTS_smis_figure(mode='union', training='retrain', metric='greedy'):\n size = 2.1E6\n results = smis_results['HTS'][mode][training]\n if mode == 'union':\n y_title = 'Total Number of Molecules Explored Among Runs'\n elif mode == 'intersection':\n y_title = 'Number of Molecules Shared Among Runs'\n else:\n raise ValueError('Unsupported mode!')\n \n results_series = make_subplots(\n rows=1, cols=len(SPLITS), shared_yaxes=True,\n x_title='Iteration', y_title=y_title,\n subplot_titles=[f'{split}%' for split in SPLITS])\n\n for i, split in enumerate(SPLITS):\n split_results = results[split]\n xs = list(range(6))\n # add actual data\n for j, model in enumerate(MODELS):\n ys = split_results[model][metric]\n\n split_trace = go.Scatter(\n x=xs, y=ys,\n mode='lines+markers',\n marker=dict(symbol=MARKERS[j], color=MODEL_COLORS[j]),\n name=model.upper(), legendgroup=model,\n showlegend=split==0.1\n )\n results_series.add_trace(split_trace, row=1, col=i+1)\n \n # add standards\n if mode =='union':\n ys_upper = [int(5*(x+1)*split/100*size) for x in xs]\n results_series.add_trace(go.Scatter(\n x=xs, y=ys_upper,\n mode='lines', line=dict(color='black'),\n name='upper_bound', showlegend=False\n ), row=1, col=i+1)\n \n ys_lower = [int((5+x)*split/100*size) for x in xs]\n results_series.add_trace(go.Scatter(\n x=xs, y=ys_lower,\n mode='lines', line=dict(color='black'),\n name='lower_bound', showlegend=False\n ), row=1, col=i+1)\n \n results_series.update_yaxes(row=1, col=i+1, range=[0, ys_upper[-1]])\n elif mode == 'intersection':\n results_series.update_yaxes(row=1, col=i+1, range=[0, 1000])\n \n results_series.update_xaxes(row=1, col=i+1, range=[-0, 5], dtick=1)\n #rangemode='tozero', nticks=6)\n border_params = dict(\n showgrid=True, zeroline=False, showticklabels=True, \n visible=True, mirror=True, linewidth=2,\n linecolor='black'\n )\n results_series.update_xaxes(**border_params)\n results_series.update_yaxes(**border_params)\n results_series.update_traces(marker=dict(line_width=2, size=7.5))\n \n for i in results_series['layout']['annotations']:\n i['font'] = dict(color='black', size=20, family='sans-serif')\n results_series['layout']['annotations'][-2]['yanchor']='bottom'\n results_series['layout']['annotations'][-2]['y'] = -0.15\n results_series['layout']['annotations'][-1].x = -0.02\n HEIGHT = 500\n results_series.update_layout(\n legend_title_text='Model', legend_traceorder='reversed',\n height=HEIGHT, width=HEIGHT/1.5*3,\n font=dict(color='black', size=20, family='sans-serif')\n )\n \n return results_series", "_____no_output_____" ] ], [ [ "run the cell below to generate the plots", "_____no_output_____" ] ], [ [ "fig = gen_10k50k_smis_figure('union', 'greedy')\nfig.show()\n\nfig = gen_HTS_smis_figure('union', 'retrain', 'greedy')\nfig.show()", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ] ]

ec7b9a9e9a79341e465f553186233e073b03be5f

20,457

ipynb

Jupyter Notebook

examples/mp/jupyter/Benders_decomposition.ipynb

raineydavid/docplex-examples

53cb00ab26c1200138a827960b0309e2633a035b

[ "Apache-2.0" ]

2

2020-05-15T14:35:17.000Z

2020-06-29T02:05:52.000Z

examples/mp/jupyter/Benders_decomposition.ipynb

raineydavid/docplex-examples

53cb00ab26c1200138a827960b0309e2633a035b

[ "Apache-2.0" ]

null

examples/mp/jupyter/Benders_decomposition.ipynb

raineydavid/docplex-examples

53cb00ab26c1200138a827960b0309e2633a035b

[ "Apache-2.0" ]

1

2019-05-17T01:46:25.000Z

34.49747

348

0.573349

[ [ [ "# Benders decomposition with decision optimization\n\nThis tutorial includes everything you need to set up decision optimization engines, build a mathematical programming model, then use the benders decomposition on it.\n\n\nWhen you finish this tutorial, you'll have a foundational knowledge of _Prescriptive Analytics_.\n\n>This notebook is part of **[Prescriptive Analytics for Python](http://ibmdecisionoptimization.github.io/docplex-doc/)**\n>\n>It requires either an [installation of CPLEX Optimizers](http://ibmdecisionoptimization.github.io/docplex-doc/getting_started.html) or it can be run on [IBM Watson Studio Cloud](https://www.ibm.com/cloud/watson-studio/) (Sign up for a [free IBM Cloud account](https://dataplatform.cloud.ibm.com/registration/stepone?context=wdp&apps=all>)\nand you can start using Watson Studio Cloud right away).\n\nTable of contents:\n\n- [Describe the business problem](#Describe-the-business-problem:--Games-Scheduling-in-the-National-Football-League)\n* [How decision optimization (prescriptive analytics) can help](#How--decision-optimization-can-help)\n* [Use decision optimization](#Use-decision-optimization)\n * [Step 1: Import the library](#Step-1:-Import-the-library)\n * [Step 2: Set up the prescriptive model](#Step-2:-Set-up-the-prescriptive-model)\n * [Step 3: Solve the problem with default CPLEX algorithm](#Step-3:-Solve-the-problem-with-default-CPLEX-algorithm)\n * [Step 4: Apply a Benders strategy](#Step-4:-Apply-a-Benders-strategy)\n * [Step 5: Use the CPLEX annotations to guide CPLEX in your Benders decomposition](#Step-5:-Use-the-CPLEX-annotations-to-guide-CPLEX-in-your-Benders-decomposition)\n* [Summary](#Summary)\n****", "_____no_output_____" ], [ "Benders decomposition is an approach to solve mathematical programming problems with a decomposable structure.\n\nStarting with 12.7, CPLEX can decompose the model into a single master and (possibly multiple) subproblems. \n\nTo do so, CPLEX can use of annotations that you supply for your model or can automatically do the decomposition. \nThis approach can be applied to mixed-integer linear programs (MILP). For certain types of problems, this approach can offer significant performance improvements.\n\n**Note**:\nIf your problem does not match such decomposition, CPLEX will raise an error at solve time.\n\nCPLEX will produce an error CPXERR_BAD_DECOMPOSITION if the annotated decomposition does not yield disjoint subproblems", "_____no_output_____" ], [ "#### Learn more bout Benders decomposition", "_____no_output_____" ], [ "Directs a reader to more sources about Benders algorithm.\n\nThe popular acceptance of the original paper suggesting a decomposition or partitioning of a model to support solution of mixed integer programs gave rise to \"Benders algorithm\" as the name.\n\n* *J. Benders*. <i>Partitioning procedures for solving mixed-variables programming problems in Numerische Mathematik, volume 4, issue 1, pages 238–252, 1962</i>\n\nOther researchers developed the theory of cut-generating linear programs (CGLP) to further this practice.\n* *M. Fischetti, D. Salvagnin, A. Zanette*. <i>A note on the selection of Benders’ cuts in Mathematical Programming, series B, volume 124, pages 175-182, 2010</i>\n\nStill others applied the practice to practical operations research. This technical report describes Benders algorithm in \"modern\" terms and offers implementation hints.\n* *M. Fischetti, I. Ljubic, M. Sinnl*. <i>Benders decomposition without separability: a computational study for capacitated facility location problems in Technical Report University of Padova, 2016</i>", "_____no_output_____" ], [ "## How decision optimization can help\n\n* Prescriptive analytics (decision optimization) technology recommends actions that are based on desired outcomes. It takes into account specific scenarios, resources, and knowledge of past and current events. With this insight, your organization can make better decisions and have greater control of business outcomes. \n\n* Prescriptive analytics is the next step on the path to insight-based actions. It creates value through synergy with predictive analytics, which analyzes data to predict future outcomes. \n\n* Prescriptive analytics takes that insight to the next level by suggesting the optimal way to handle that future situation. Organizations that can act fast in dynamic conditions and make superior decisions in uncertain environments gain a strong competitive advantage. \n<br/>\n\n<u>With prescriptive analytics, you can:</u> \n\n* Automate the complex decisions and trade-offs to better manage your limited resources.\n* Take advantage of a future opportunity or mitigate a future risk.\n* Proactively update recommendations based on changing events.\n* Meet operational goals, increase customer loyalty, prevent threats and fraud, and optimize business processes.\n\n", "_____no_output_____" ], [ "## Use decision optimization", "_____no_output_____" ], [ "### Step 1: Import the library\n\nRun the following code to import Decision Optimization CPLEX Modeling library. The *DOcplex* library contains the two modeling packages, Mathematical Programming and Constraint Programming, referred to earlier.", "_____no_output_____" ] ], [ [ "import sys\ntry:\n import docplex.mp\nexcept:\n raise Exception('Please install docplex. See https://pypi.org/project/docplex/')", "_____no_output_____" ] ], [ [ "A restart of the kernel might be needed.", "_____no_output_____" ], [ "### Step 2: Set up the prescriptive model", "_____no_output_____" ], [ "We will write a toy model just in order to show how to use the annotation api.\n\nThis model is not important: it just matche a benders decomposition, that is CPLEX can apply its new algorithm without any error.\n\nThe aim of this notebook is to discover and learn how to successfully apply Benders, not to see huge performance differences between a standard solve and a Benders based solve.", "_____no_output_____" ] ], [ [ "d1 = 25\nd2 = 35\n\nCosts = [[20, 15, 22, 27, 13, 4, 15, 6, 15, 22, 25, 13, 7, 28, 14, 5, 8, 1, 17, 3, 19, 17, 22, 12, 14],\n [2, 15, 16, 16, 10, 13, 4, 2, 6, 29, 10, 8, 20, 11, 8, 11, 28, 17, 10, 29, 3, 24, 12, 11, 11],\n [13, 14, 6, 17, 14, 13, 8, 29, 19, 26, 22, 0, 8, 29, 15, 20, 5, 20, 26, 17, 24, 10, 24, 9, 1],\n [7, 27, 24, 3, 4, 23, 11, 9, 18, 1, 29, 24, 16, 9, 8, 3, 18, 24, 10, 12, 1, 3, 15, 29, 3],\n [25, 26, 29, 6, 24, 8, 2, 10, 17, 0, 4, 7, 2, 17, 2, 27, 24, 20, 18, 5, 5, 2, 21, 26, 20],\n [29, 5, 15, 5, 4, 26, 18, 8, 2, 14, 13, 6, 14, 28, 16, 28, 23, 8, 5, 8, 10, 28, 17, 0, 23],\n [12, 16, 10, 16, 17, 10, 29, 11, 28, 22, 25, 8, 27, 12, 10, 28, 7, 5, 3, 9, 18, 10, 15, 16, 2],\n [12, 9, 14, 23, 26, 4, 3, 3, 22, 12, 11, 9, 19, 5, 6, 16, 1, 1, 9, 20, 23, 23, 27, 4, 11],\n [18, 13, 28, 29, 3, 28, 16, 11, 9, 2, 7, 20, 13, 23, 6, 10, 3, 16, 14, 2, 15, 17, 1, 19, 27],\n [29, 17, 17, 14, 21, 18, 8, 21, 9, 20, 14, 6, 29, 24, 24, 4, 18, 16, 21, 24, 26, 0, 26, 9, 5],\n [27, 24, 21, 28, 17, 18, 10, 10, 26, 25, 13, 18, 2, 9, 16, 26, 10, 22, 5, 17, 15, 0, 9, 0, 16],\n [13, 15, 17, 21, 25, 9, 22, 13, 20, 15, 1, 17, 18, 10, 2, 27, 19, 21, 14, 26, 29, 13, 28, 28, 15],\n [16, 12, 2, 2, 9, 27, 11, 14, 12, 2, 14, 29, 3, 12, 18, 6, 7, 9, 1, 5, 19, 14, 11, 29, 4],\n [1, 15, 27, 29, 16, 17, 10, 10, 17, 19, 6, 10, 20, 20, 19, 10, 19, 26, 15, 7, 20, 19, 13, 3, 22],\n [22, 14, 12, 3, 22, 6, 15, 3, 6, 10, 9, 13, 11, 21, 6, 19, 29, 4, 5, 21, 7, 12, 13, 11, 22],\n [9, 27, 22, 29, 11, 14, 1, 19, 21, 2, 4, 13, 17, 9, 10, 17, 13, 8, 24, 13, 26, 27, 23, 4, 21],\n [3, 14, 26, 18, 17, 3, 1, 11, 13, 8, 22, 3, 18, 26, 17, 15, 22, 10, 19, 23, 13, 14, 17, 18, 27],\n [21, 14, 1, 28, 28, 0, 0, 29, 12, 23, 22, 17, 19, 2, 10, 19, 4, 18, 28, 13, 27, 12, 9, 29, 22],\n [29, 3, 20, 5, 5, 23, 28, 16, 1, 8, 26, 23, 11, 11, 21, 17, 13, 21, 3, 8, 6, 15, 18, 6, 24],\n [14, 20, 26, 10, 17, 20, 5, 9, 25, 20, 14, 22, 5, 12, 0, 18, 7, 0, 8, 15, 21, 12, 26, 7, 21],\n [7, 7, 1, 9, 24, 29, 0, 3, 29, 24, 1, 6, 14, 0, 11, 5, 21, 12, 15, 1, 25, 4, 7, 17, 16],\n [8, 18, 15, 6, 1, 22, 26, 13, 19, 20, 12, 15, 19, 27, 13, 3, 22, 22, 22, 20, 0, 4, 24, 13, 25],\n [14, 6, 29, 23, 8, 5, 4, 18, 21, 29, 18, 2, 2, 3, 7, 13, 12, 9, 2, 18, 26, 3, 18, 7, 7],\n [5, 8, 4, 8, 25, 4, 6, 20, 14, 21, 18, 16, 15, 11, 7, 8, 20, 27, 22, 7, 5, 8, 24, 11, 8],\n [0, 8, 29, 25, 29, 0, 12, 25, 19, 9, 19, 25, 27, 21, 2, 23, 2, 25, 17, 6, 0, 6, 15, 2, 15],\n [23, 24, 10, 26, 7, 5, 5, 26, 1, 16, 22, 8, 24, 9, 16, 17, 1, 26, 20, 23, 18, 20, 23, 2, 19],\n [16, 3, 9, 21, 15, 29, 8, 26, 20, 12, 18, 27, 29, 15, 24, 9, 17, 24, 3, 5, 21, 28, 7, 1, 12],\n [1, 11, 21, 1, 13, 14, 16, 14, 17, 25, 18, 9, 19, 26, 1, 13, 15, 6, 14, 10, 12, 19, 0, 15, 7],\n [20, 14, 7, 5, 8, 16, 12, 0, 5, 14, 18, 16, 24, 27, 20, 7, 11, 3, 16, 8, 2, 2, 4, 0, 3],\n [26, 19, 27, 29, 8, 9, 8, 10, 18, 4, 6, 0, 5, 17, 12, 18, 17, 17, 13, 0, 16, 12, 18, 19, 16],\n [3, 12, 11, 28, 3, 2, 14, 14, 17, 29, 18, 14, 19, 24, 9, 27, 4, 19, 6, 24, 19, 3, 28, 20, 4],\n [2, 0, 21, 14, 21, 12, 27, 6, 20, 29, 13, 21, 23, 0, 20, 4, 11, 27, 3, 11, 21, 11, 21, 4, 17],\n [20, 26, 5, 8, 18, 14, 12, 12, 24, 3, 8, 0, 25, 16, 19, 21, 7, 4, 23, 21, 20, 28, 6, 21, 19],\n [16, 18, 9, 1, 9, 7, 14, 6, 28, 26, 3, 14, 27, 4, 9, 9, 1, 9, 24, 3, 14, 13, 18, 3, 27],\n [1, 19, 7, 20, 26, 27, 0, 7, 4, 0, 13, 8, 10, 17, 14, 19, 21, 21, 14, 15, 22, 14, 5, 27, 0]];", "_____no_output_____" ], [ "R1 = range(1,d1)\nR2 = range(1,d2);\n\ndim = range(1,d1*d2+1)", "_____no_output_____" ] ], [ [ "Create one model instance, with a name. We set the log output to true such that we can see when CPLEX enables the Benders algorithm.", "_____no_output_____" ] ], [ [ "# first import the Model class from docplex.mp\nfrom docplex.mp.model import Model\n\nm = Model(name='benders', log_output=True)", "_____no_output_____" ], [ "X = m.continuous_var_dict([(i,j) for i in R2 for j in R1])\nY = m.integer_var_dict(R1, 0, 1)\n\n\nbendersPartition = {(i,j) : i for i in R2 for j in R1}", "_____no_output_____" ], [ "m.minimize( m.sum( Costs[i][j]*X[i,j] for i in R2 for j in R1) + sum(Y[i] for i in R1) )\n\n\nm.add_constraints( m.sum( X[i,j] for j in R1) ==1 for i in R2)\n \nm.add_constraints( X[i,j] - Y[j] <= 0 for i in R2 for j in R1)", "_____no_output_____" ] ], [ [ "#### Solve with Decision Optimization \n\nIf you're using a Community Edition of CPLEX runtimes, depending on the size of the problem, the solve stage may fail and will need a paying subscription or product installation.\n\nYou will get the best solution found after ***n*** seconds, thanks to a time limit parameter.", "_____no_output_____" ] ], [ [ "m.print_information()", "_____no_output_____" ] ], [ [ "### Step 3: Solve the problem with default CPLEX algorithm", "_____no_output_____" ] ], [ [ "msol = m.solve()\nassert msol is not None, \"model can't solve\"\nm.report()", "_____no_output_____" ] ], [ [ "#### Inspect the CPLEX Log.", "_____no_output_____" ], [ "If you inspect the CPLEX, you will see that it is a very standard log.\nCPLEX needed 63 iterations to solve it.", "_____no_output_____" ] ], [ [ "obj1 = m.objective_value", "_____no_output_____" ] ], [ [ "### Step 4: Apply a Benders strategy", "_____no_output_____" ], [ "CPLEX implements a default Benders decomposition in certain situations.\n\nIf you want CPLEX to apply a Benders strategy as it solves your problem, but you do not specify cpxBendersPartition annotations yourself, CPLEX puts all integer variables in master and continuous variables into subproblems. \nIf there are no integer variables in your model, or if there are no continuous variables in your model, CPLEX raises an error stating that it cannot automatically decompose the model to apply a Benders strategy.", "_____no_output_____" ], [ "You just need to set the Benders strategy parameter.", "_____no_output_____" ], [ "CPLEX supports 4 values for this parameter, from -1 to 3:\n* OFF (default value) will ignore Benders.\n* AUTO, USER, WORKERS, FULL will enable Benders.\n\nRefer to the CPLEX documentation to understand the differences between the 4 values that trigger it.", "_____no_output_____" ] ], [ [ "m.parameters.benders.strategy = 3", "_____no_output_____" ], [ "m.print_information()", "_____no_output_____" ] ], [ [ "We call cplex solve, but with the <i>clean_before_solve</i> flag because we want it to forget everything about previous solve and solution.", "_____no_output_____" ] ], [ [ "msol = m.solve(clean_before_solve=True)\nassert msol is not None, \"model can't solve\"\nm.report()", "_____no_output_____" ] ], [ [ "#### Inspect the CPLEX Log.", "_____no_output_____" ], [ "Inspect the CPLEX log: you can now see that the log are different and you can see the message\n<code>\nBenders cuts applied: 3\n</code>\nwhich proves CPLEX applied successfully\n\nYou can see that CPLEX needed only 61 cumulative iterations, while it needed 63 previously.", "_____no_output_____" ] ], [ [ "obj2 = m.objective_value", "_____no_output_____" ] ], [ [ "### Step 5: Use the CPLEX annotations to guide CPLEX in your Benders decomposition", "_____no_output_____" ] ], [ [ "m.parameters.benders.strategy = 1", "_____no_output_____" ] ], [ [ "Settings benders annotation in docplex is very simple.\nYou just need to use the <i>benders_annotation</i> property available on variables and constraints to state which worker they belong to.", "_____no_output_____" ] ], [ [ "for i in R2:\n for j in R1:\n X[i,j].benders_annotation = i%2", "_____no_output_____" ], [ "m.print_information()", "_____no_output_____" ], [ "msol = m.solve(clean_before_solve=True)\nassert msol is not None, \"model can't solve\"\nm.report()", "_____no_output_____" ] ], [ [ "#### Inspect the CPLEX Log.", "_____no_output_____" ], [ "Inspect the CPLEX log: you can see that you now need only 57 cumulative iterations instead of 61 with default Benders and 63 with no Benders.\nIf you look at the <i>Best Bound</i> column, you will also see that the listed sub problems are not the same as CPLEX applied the decomposition provided by the annotations.", "_____no_output_____" ] ], [ [ "obj3 = m.objective_value", "_____no_output_____" ], [ "assert (obj1 == obj2) and (obj2 == obj3)", "_____no_output_____" ] ], [ [ "## Summary\n\n\nYou learned how to set up and use the IBM Decision Optimization CPLEX Modeling for Python to formulate a Mathematical Programming model and apply a Benders decomposition.", "_____no_output_____" ], [ "#### References\n* [Decision Optimization CPLEX Modeling for Python documentation](http://ibmdecisionoptimization.github.io/docplex-doc/)\n* [Decision Optimization on Cloud](https://developer.ibm.com/docloud/)\n* Need help with DOcplex or to report a bug? Please go [here](https://stackoverflow.com/questions/tagged/docplex)\n* Contact us at [email protected]\"\n", "_____no_output_____" ], [ "Copyright © 2017-2019 IBM. Sample Materials.", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown", "markdown", "markdown" ] ]

ec7b9c8223ec773cb6edbd283986eb884b02e472

2,938

ipynb

Jupyter Notebook

2020-11/beginner/beginners_soln.ipynb

enzo-bc/user_group

b81a38cc0c1bb0bb066dc0a0a350917967cde291

[ "Unlicense" ]

11

2019-09-13T15:15:44.000Z

2021-04-15T18:18:24.000Z

2020-11/beginner/beginners_soln.ipynb

enzo-bc/user_group

b81a38cc0c1bb0bb066dc0a0a350917967cde291

[ "Unlicense" ]

3

2020-07-08T18:59:52.000Z

2022-02-12T14:13:26.000Z

2020-11/beginner/beginners_soln.ipynb

enzo-bc/user_group

b81a38cc0c1bb0bb066dc0a0a350917967cde291

[ "Unlicense" ]

25

2019-09-13T16:34:58.000Z

2022-03-04T13:34:47.000Z

22.427481

73

0.494554

[ [ [ "from tqdm.notebook import trange, tqdm", "_____no_output_____" ], [ "def process_row(line):\n line = line.strip()\n row = line.split('\\t')\n return [int(e) for e in row]", "_____no_output_____" ], [ "def rect_sum(data, top, bottom, left, right):\n total = 0\n for row_num in range(top, bottom+1):\n row = data[row_num]\n total += sum(row[left:right+1])\n return total", "_____no_output_____" ], [ "def print_rect(data, top, bottom, left, right):\n for row_num in range(top, bottom+1):\n row = data[row_num]\n print('\\t'.join([str(e) for e in row[left:right+1]]))", "_____no_output_____" ], [ "with open('rect2.txt') as file:\n data = [process_row(line) for line in file]", "_____no_output_____" ], [ "height = len(data)\nwidth = len(data[0])\n\nbest_total = None\nbest_range = None\nfor top in trange(height):\n for bottom in range(top, height):\n for left in range(width):\n for right in range(left, width):\n total = rect_sum(data, top, bottom, left, right)\n if best_total is None or total > best_total:\n best_total = total\n best_range = top, bottom, left, right", "_____no_output_____" ], [ "top, bottom, left, right = best_range", "_____no_output_____" ], [ "print(f'total = {best_total}')\nprint()\nprint_rect(data, top, bottom, left, right)", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7ba07afbca8b057c1b9742c009145a92410cef

16,482

ipynb

Jupyter Notebook

SEMANA 2/main.ipynb

jonatascs/Codenation

5338090222a952b0772adc53ae666bc09c4367ad

[ "MIT" ]

null

SEMANA 2/main.ipynb

jonatascs/Codenation

5338090222a952b0772adc53ae666bc09c4367ad

[ "MIT" ]

2

2021-02-02T22:52:12.000Z

2021-02-02T22:55:11.000Z

SEMANA 2/main.ipynb

jonatascs/Codenation

5338090222a952b0772adc53ae666bc09c4367ad

[ "MIT" ]

null

23.715108

187

0.42598

[ [ [ "# Desafio 1\n\nPara esse desafio, vamos trabalhar com o data set [Black Friday](https://www.kaggle.com/mehdidag/black-friday), que reúne dados sobre transações de compras em uma loja de varejo.\n\nVamos utilizá-lo para praticar a exploração de data sets utilizando pandas. Você pode fazer toda análise neste mesmo notebook, mas as resposta devem estar nos locais indicados.\n\n> Obs.: Por favor, não modifique o nome das funções de resposta.", "_____no_output_____" ], [ "## _Set up_ da análise", "_____no_output_____" ] ], [ [ "import pandas as pd\nimport numpy as np", "_____no_output_____" ], [ "#carregar dataframe\nblack_friday = pd.read_csv(\"black_friday.csv\")", "_____no_output_____" ] ], [ [ "## Inicie sua análise a partir daqui", "_____no_output_____" ] ], [ [ "#mostrar as 5 primeiras linhas\nblack_friday.head(5)", "_____no_output_____" ], [ "#sumario de nulos\nblack_friday.isna().sum()", "_____no_output_____" ], [ "#mostrar nome das colunas\nblack_friday.columns", "_____no_output_____" ], [ "#mostrar Tipos\nblack_friday.dtypes", "_____no_output_____" ] ], [ [ "## Questão 1\n\nQuantas observações e quantas colunas há no dataset? Responda no formato de uma tuple `(n_observacoes, n_colunas)`.", "_____no_output_____" ] ], [ [ "def q1():\n return black_friday.shape\n ", "_____no_output_____" ], [ "q1()", "_____no_output_____" ] ], [ [ "## Questão 2\n\nHá quantas mulheres com idade entre 26 e 35 anos no dataset? Responda como um único escalar.", "_____no_output_____" ] ], [ [ "def q2():\n return black_friday[(black_friday.Gender == 'F') & (black_friday.Age == '26-35')].shape[0]\n ", "_____no_output_____" ], [ "q2()", "_____no_output_____" ] ], [ [ "## Questão 3\n\nQuantos usuários únicos há no dataset? Responda como um único escalar.", "_____no_output_____" ] ], [ [ "def q3():\n return black_friday.User_ID.unique().shape[0]", "_____no_output_____" ], [ " q3()", "_____no_output_____" ] ], [ [ "## Questão 4\n\nQuantos tipos de dados diferentes existem no dataset? Responda como um único escalar.", "_____no_output_____" ] ], [ [ "def q4():\n return black_friday.dtypes.unique().shape[0]", "_____no_output_____" ], [ "q4()", "_____no_output_____" ] ], [ [ "## Questão 5\n\nQual porcentagem dos registros possui ao menos um valor null (`None`, `ǸaN` etc)? Responda como um único escalar entre 0 e 1.", "_____no_output_____" ] ], [ [ "def q5():\n return np.float(black_friday.isnull().sum().max() / black_friday.shape[0] )", "_____no_output_____" ], [ "q5()", "_____no_output_____" ] ], [ [ "## Questão 6\n\nQuantos valores null existem na variável (coluna) com o maior número de null? Responda como um único escalar.", "_____no_output_____" ] ], [ [ "def q6():\n return np.int(black_friday.isnull().sum().max())", "_____no_output_____" ], [ "q6()", "_____no_output_____" ] ], [ [ "## Questão 7\n\nQual o valor mais frequente (sem contar nulls) em `Product_Category_3`? Responda como um único escalar.", "_____no_output_____" ] ], [ [ "def q7():\n return black_friday['Product_Category_3'].mode()[0]\n", "_____no_output_____" ], [ "q7()", "_____no_output_____" ] ], [ [ "## Questão 8\n\nQual a nova média da variável (coluna) `Purchase` após sua normalização? Responda como um único escalar.", "_____no_output_____" ] ], [ [ "def q8():\n return np.float(((black_friday['Purchase'] - black_friday['Purchase'].min())/\n (black_friday['Purchase'].max()-black_friday['Purchase'].min())).mean())", "_____no_output_____" ], [ "q8()", "_____no_output_____" ] ], [ [ "## Questão 9\n\nQuantas ocorrências entre -1 e 1 inclusive existem da variáel `Purchase` após sua padronização? Responda como um único escalar.", "_____no_output_____" ], [ "## Questão 10\n\nPodemos afirmar que se uma observação é null em `Product_Category_2` ela também o é em `Product_Category_3`? Responda com um bool (`True`, `False`).", "_____no_output_____" ] ], [ [ "def q10(): \n return bool((((black_friday.Product_Category_2.isnull() == True) & (black_friday.Product_Category_3.isnull() == True)).sum()) == black_friday.Product_Category_2.isnull().sum())", "_____no_output_____" ], [ "q10()", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code", "code" ] ]

ec7bbac46b7968fbe7fd7eaa407f6ba27c06880b

218,758

ipynb

Jupyter Notebook

Flexx POCs.ipynb

siowyisheng/flexx-template

df7eb04c58a34b883b204ae64671da6f77336200

[ "MIT" ]

null

Flexx POCs.ipynb

siowyisheng/flexx-template

df7eb04c58a34b883b204ae64671da6f77336200

[ "MIT" ]

null

Flexx POCs.ipynb

siowyisheng/flexx-template

df7eb04c58a34b883b204ae64671da6f77336200

[ "MIT" ]

null

48.85172

3,144

0.499438

[ [ [ "%gui asyncio\n\nfrom flexx import flx\n\nclass Example(flx.Widget):\n \n def init(self):\n with flx.HBox():\n flx.Button(text='hello', flex=1) # flex is the ratio of width used\n flx.Button(text='world', flex=5) ", "[W 14:03:31 flexx.app] Re-defining app class 'Example'\n" ], [ "from flexx import flx\n\nclass Example(flx.Widget):\n\n def init(self):\n with flx.HSplit():\n flx.Button(text='foo')\n with flx.VBox():\n flx.Widget(style='background:red;', flex=1)\n flx.Widget(style='background:blue;', flex=1)", "[I 14:06:56 flexx.app] Asset store collected 2 new modules.\n" ], [ "from flexx import flx\n\nclass Example(flx.Widget):\n\n counter = flx.IntProp(3, settable=True)\n\n def init(self):\n super().init()\n\n with flx.HBox():\n self.but1 = flx.Button(text='reset')\n self.but2 = flx.Button(text='increase')\n self.label = flx.Label(text='', flex=1) # take all remaining space\n\n @flx.action\n def increase(self):\n self._mutate_counter(self.counter + 1)\n\n @flx.reaction('but1.pointer_click')\n def but1_clicked(self, *events):\n self.set_counter(0)\n\n @flx.reaction('but2.pointer_click')\n def but2_clicked(self, *events):\n self.increase(0)\n\n @flx.reaction\n def update_label(self, *events):\n self.label.set_text('count is ' + str(self.counter))", "_____no_output_____" ], [ "from flexx import flx\n\nclass Example(flx.Widget):\n\n def init(self):\n super().init()\n with flx.VBox():\n with flx.HBox():\n self.but = flx.Button(text='add')\n self.label = flx.Label(flex=1)\n with flx.HBox() as self.box:\n flx.Button(text='x')\n\n @flx.reaction('but.pointer_click')\n def add_widget(self, *events):\n flx.Button(parent=self.box, text='x')\n\n @flx.reaction('box.children*.pointer_click')\n def a_button_was_pressed(self, *events):\n ev = events[-1] # only care about last event\n self.label.set_text(ev.source.id + ' was pressed')", "_____no_output_____" ], [ "from flexx import flx\n\nclass StaticDisplay(flx.Widget):\n def init(self):\n flx.Label(text='Hello world',flex=1)\n flx.Label(text='Second line',flex=2)", "_____no_output_____" ], [ "from flexx import flx\n\n# Display from variable\nclass Root(flx.Widget):\n name = flx.StringProp('HELLOW')\n \n def init(self):\n flx.Label(text=self.name,flex=1)", "_____no_output_____" ], [ "from flexx import flx\n\n# Change the variable display with textfield\nclass Root(flx.Widget):\n name = flx.StringProp('HELLOW', settable=True)\n \n def init(self):\n with flx.HBox():\n self.name_input = flx.LineEdit(flex=1)\n self.name_label = flx.Label(text=lambda:self.name,flex=1)\n\n @flx.reaction('name_input.user_done')\n def update_name(self, *events):\n self.set_name(self.name_input.text)", "_____no_output_____" ], [ "from flexx import flx\n\n# Application of variable input using a template\nclass Root(flx.Widget):\n name = flx.StringProp(settable=True)\n gender = flx.StringProp(settable=True)\n tone = flx.StringProp(settable=True)\n \n def init(self):\n with flx.VBox():\n with flx.HBox():\n flx.Label(text='Name: ')\n self.name_input = flx.LineEdit(flex=1)\n flx.Label(text='Gender: ')\n self.gender_input = flx.LineEdit(flex=1)\n flx.Label(text='Tone: ')\n self.tone_input = flx.LineEdit(flex=1)\n self.email = flx.Label(text=self.write_email,flex=1) \n \n @flx.reaction('name_input.user_done')\n def update_name(self, *events):\n self.set_name(self.name_input.text)\n\n @flx.reaction('gender_input.user_done')\n def update_gender(self, *events):\n self.set_gender(self.gender_input.text)\n\n @flx.reaction('tone_input.user_done')\n def update_tone(self, *events):\n self.set_tone(self.tone_input.text)\n \n def write_email(self):\n return 'Dear {},\\nI\\'m pretty sure you\\'re a {}. I am now using my {} tone. Be pleased'.format(self.name, self.gender, self.tone)", "_____no_output_____" ], [ "from flexx import flx\n\n# Application of variable input using a template and component abstraction\nclass TextInput(flx.Widget):\n text = flx.StringProp(settable=True)\n name = flx.StringProp('Pass in something here')\n \n def init(self):\n flx.Label(text=lambda:'{}: '.format(self.name))\n self.text_input = flx.LineEdit(flex=1)\n\n @flx.reaction('text_input.user_done')\n def update_text(self, *events):\n self.set_text(self.text_input.text)\n \n\nclass Root(flx.Widget):\n def init(self):\n with flx.VBox():\n with flx.HBox():\n self.name_input = TextInput(name='Name')\n self.gender_input = TextInput(name='Gender')\n self.tone_input = TextInput(name='Tone')\n self.email = flx.Label(text=self.write_email,flex=1, wrap=True) \n \n def write_email(self):\n return 'Dear {},\\nI\\'m pretty sure you\\'re a {}. I am now using my {} tone. Be pleased'.format(self.name_input.text, self.gender_input.text, self.tone_input.text)", "_____no_output_____" ], [ "from flexx import flx\n\n# Application of variable input using a template and component abstraction and set_HTML\nclass TextInput(flx.Widget):\n text = flx.StringProp(settable=True)\n name = flx.StringProp('Pass in something here')\n \n def init(self):\n flx.Label(text=lambda:'{}: '.format(self.name))\n self.text_input = flx.LineEdit(flex=1)\n\n @flx.reaction('text_input.user_done')\n def update_text(self, *events):\n self.set_text(self.text_input.text)\n \nclass Root(flx.Widget):\n def init(self):\n with flx.VBox(padding=20):\n with flx.HBox(flex=1):\n self.name_input = TextInput(name='Name')\n self.gender_input = TextInput(name='Gender')\n self.tone_input = TextInput(name='Tone')\n with flx.HBox(flex=5):\n self.email = flx.Label(html=self.write_email,flex=10, wrap=True) \n \n def write_email(self):\n return 'Dear {},<br><br>I\\'m pretty sure you\\'re a {}. I am now using my {} tone. Be pleased.'.format(self.name_input.text, self.gender_input.text, self.tone_input.text)", "_____no_output_____" ], [ "from flexx import flx\n\n# Add two numbers together when you press a button, but not before\nclass Root(flx.Widget):\n text = flx.StringProp(settable=True)\n number = flx.IntProp(settable=True)\n \n def init(self):\n with flx.VSplit(padding=20):\n with flx.VBox(flex=1): \n with flx.HBox(flex=1):\n self.first = flx.LineEdit(title='1st number: ')\n self.second = flx.LineEdit(title='2nd number: ')\n self.button = flx.Button(text='Calculate')\n with flx.HBox(flex=5):\n flx.Label(html=lambda:'<h1>{}</h1>'.format(self.number)) \n \n @flx.reaction('button.pointer_click')\n def update_number(self, *events):\n self.set_number(int(self.first.text) + int(self.second.text))", "[I 04:03:44 flexx.app] Asset store collected 2 new modules.\n" ], [ "app = flx.App(Root)\napp.launch('app')\nflx.init_notebook()", "_____no_output_____" ], [ "app.export('example2.html',link=0)", "_____no_output_____" ], [ "# other emitters to play with\n\n# key_down, key_press, key_up, pointer_cancel, pointer_click, \n# pointer_double_click, pointer_down, pointer_move, pointer_up, pointer_wheel", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7bbdc7812d2ac00d8d9926c06441d8a5fcf7c1

12,696

ipynb

Jupyter Notebook

likelihood_count_vectorizer.ipynb

yeonghwanchoi/fc16

27ca7b890155a93b8a08fdbb1a783439aff71d6f

[ "MIT" ]

null

likelihood_count_vectorizer.ipynb

yeonghwanchoi/fc16

27ca7b890155a93b8a08fdbb1a783439aff71d6f

[ "MIT" ]

null

likelihood_count_vectorizer.ipynb

yeonghwanchoi/fc16

27ca7b890155a93b8a08fdbb1a783439aff71d6f

[ "MIT" ]

null

23.909605

79

0.433522

[ [ [ "from sklearn.feature_extraction.text import CountVectorizer\nvectorizer = CountVectorizer(min_df = 1)", "_____no_output_____" ], [ "contents= ['상처받은 아이들은 너무 일찍 커버려',\n '내가 상처받은 거 아는 사람 불편해',\n '잘 사는 사람들은 좋은 사람 되기 쉬워',\n '아무 일도 아니야 괜찮아']\n", "_____no_output_____" ], [ "#형태소 분석\nfrom konlpy.tag import Twitter\nt=Twitter()\ncontents_tokens = [t.morphs(row) for row in contents]", "_____no_output_____" ], [ "contents_tokens", "_____no_output_____" ], [ "contents_for_vectorize=[]\nfor content in contents_tokens:\n sentence=''\n for word in content:\n sentence = sentence + ' ' + word\n contents_for_vectorize.append(sentence)\ncontents_for_vectorize", "_____no_output_____" ], [ "#백터라이즈\nX= vectorizer.fit_transform(contents_for_vectorize)\nnum_samples, num_features = X.shape\nnum_samples, num_features", "_____no_output_____" ], [ "vectorizer.get_feature_names()", "_____no_output_____" ], [ "X.toarray().transpose()", "_____no_output_____" ], [ "new_post = ['상처받기 싫어 괜찮아']\nnew_post_tokens = [t.morphs(row) for row in new_post]\nnew_post_for_vectorize = []\n\nfor content in new_post_tokens:\n sentence = ''\n for word in content:\n sentence = sentence + ' ' + word\n new_post_for_vectorize.append(sentence)\nnew_post_for_vectorize", "_____no_output_____" ], [ "new_post_vec = vectorizer.transform(new_post_for_vectorize)\nnew_post_vec.toarray()", "_____no_output_____" ], [ "import scipy as sp\ndef dist_raw(v1, v2):\n delta = v1 - v2\n return sp.linalg.norm(delta.toarray())", "_____no_output_____" ], [ "dist = [dist_raw(each, new_post_vec) for each in X]\ndist", "_____no_output_____" ], [ "print(\"Best post is\", dist.index(min(dist)), ',dist=', min(dist))\nprint('test post is --->',new_post)\nprint('vest dist post is ---->', contents[dist.index(min(dist))])", "Best post is 3 ,dist= 2.0\ntest post is ---> ['상처받기 싫어 괜찮아']\nvest dist post is ----> 아무 일도 아니야 괜찮아\n" ], [ "#유클리드 유사도 \nfor i in range(0,len(contents)):\n print(X.getrow(i).toarray())\nprint('--------------------')\nprint(new_post_vec.toarray())", "[[0 1 0 1 0 0 0 1 0 0 0 0 1 0 1 0 1]]\n[[0 0 0 1 1 0 1 1 0 1 0 0 0 0 0 0 0]]\n[[0 0 1 0 0 1 2 0 1 0 0 0 0 0 0 1 0]]\n[[1 0 0 0 0 0 0 0 0 0 1 1 0 1 0 0 0]]\n--------------------\n[[1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0]]\n" ], [ "# TfidfVectorizer\nfrom sklearn.feature_extraction.text import TfidfVectorizer\nvectorizer = TfidfVectorizer(min_df=1, decode_error='ignore')\n\nX = vectorizer.fit_transform(contents_for_vectorize)\nnum_samples, num_features = X.shape\nnum_samples, num_features ", "_____no_output_____" ], [ "X.toarray().transpose()", "_____no_output_____" ], [ "#문장 적용\nnew_post_vec = vectorizer.transform(new_post_for_vectorize)\nnew_post_vec.toarray()", "_____no_output_____" ], [ "#거리 구하는 법\ndef dist_norm(v1,v2):\n v1_normalized = v1 / sp.linalg.norm(v1.toarray())\n v2_normalized = v2 / sp.linalg.norm(v1.toarray())\n \n delta = v1_normalized - v2_normalized\n \n return sp.linalg.norm(delta.toarray())", "_____no_output_____" ], [ "dist = [dist_norm(each, new_post_vec) for each in X]\nprint(\"Best post is\", dist.index(min(dist)), ',dist=', min(dist))\nprint('test post is --->',new_post)\nprint('vest dist post is ---->', contents[dist.index(min(dist))])", "Best post is 3 ,dist= 1.1021396119773588\ntest post is ---> ['상처받기 싫어 괜찮아']\nvest dist post is ----> 아무 일도 아니야 괜찮아\n" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7bbebb8965ef827db94eb5757eddf4e6112dbb

140,205

ipynb

Jupyter Notebook

_build/html/_sources/rise/05-python-intermediario-rise.ipynb

gcpeixoto/ICD

bae7d02cd467240649c89b0ba4440966fba18cc7

[ "CC0-1.0" ]

2

2021-09-09T01:56:40.000Z

2021-11-10T01:56:56.000Z

_build/html/_sources/rise/05-python-intermediario-rise.ipynb

gcpeixoto/ICD

bae7d02cd467240649c89b0ba4440966fba18cc7

[ "CC0-1.0" ]

null

_build/html/_sources/rise/05-python-intermediario-rise.ipynb

gcpeixoto/ICD

bae7d02cd467240649c89b0ba4440966fba18cc7

[ "CC0-1.0" ]

1

2021-11-23T14:24:03.000Z

22.753165

1,177

0.51802

[ [ [ "# Python Intermediário: Parte 1", "_____no_output_____" ], [ "## Visão geral sobre sequencias\n\n\nSequencias podem ser classificadas como: \n\n- _Sequências contêinerizadas_ (_container sequences_): guardam referências aos objetos que elas contêm, que podem ser de qualquer tipo. Exemplos são: `list`, `tuple` e `collections.deque`.\n\n- _Sequências rasas_ (_flat sequences_): armazenam fisicamente o valor de cada item dentro de seu próprio espaço na memória, e não como objetos distintos. Exemplos são: `str`,`bytes` e `array.array`.", "_____no_output_____" ], [ "Outra forma de agrupar tipos de sequências é por mutabilidade: \n\n- _Sequências mutáveis_: podem ter seus elementos alterados. Exemplos: `list`,`bytearray`,`memoryview`.\n\n- _Sequências imutáveis_: seus elementos são inalteráveis. Exemplos: `tuple`,`str`,`bytes`.", "_____no_output_____" ], [ "## Tuplas \n\nTuplas são sequências imutáveis de comprimento fixo.", "_____no_output_____" ], [ "**Exemplo:** Crie uma tupla com 4 elementos e desempacote-a em 4 variáveis.", "_____no_output_____" ] ], [ [ "tup1 = -1,'a',12.23,'$'\na,b,c,d = tup1\n\ntup2 = (-1,'a',12.23,'$')\ne,f,g,h = tup2\n\nprint(tup1)\nprint(tup2) ", "(-1, 'a', 12.23, '$')\n(-1, 'a', 12.23, '$')\n" ] ], [ [ "**Exemplo:** Crie uma tupla a partir de objetos existentes.", "_____no_output_____" ] ], [ [ "tuple(tup1)", "_____no_output_____" ] ], [ [ "**Exemplo:** Construa tuplas aninhadas.", "_____no_output_____" ] ], [ [ "nums = (0,2,4,6,8),(1,3,5,7,9)\nnums", "_____no_output_____" ] ], [ [ "**Exemplo:** Altere elementos de uma tupla _in-place_.", "_____no_output_____" ] ], [ [ "t = tuple(['a',[1,2,3],False])\n\n# imutabilidade!\nt[2] = True", "_____no_output_____" ], [ "# alteração in-place\nt[1].append(4) \nt ", "_____no_output_____" ] ], [ [ "**Exemplo:** Crie tuplas por concatenação.", "_____no_output_____" ] ], [ [ "('a',1) + ('b',2) + ('c',3)", "_____no_output_____" ], [ "('+','-')*3", "_____no_output_____" ] ], [ [ "**Exemplo:** Imprima tabela de valores usando desempacotamento e iteração.", "_____no_output_____" ] ], [ [ "s = [(1,2,3),(4,5,6),(7,8,9)]\ns\n\nfor (m,n,p) in s:\n print('a={0}, b={1}, c={2}'.format(m,n,p))", "a=1, b=2, c=3\na=4, b=5, c=6\na=7, b=8, c=9\n" ] ], [ [ "**Exemplo:** Desempacote elementos em iteráveis de comprimento arbitrário.", "_____no_output_____" ] ], [ [ "# desempacota por star expression e ignora primeiro item\n_,*restante = tup1\nrestante", "_____no_output_____" ] ], [ [ "Comentário: \n\n- Outros casos em que _star expressions_ são úteis envolvem dados que possuem mais de um valor atribuível a um mesmo registro (p.ex.: 1 pessoa com 2 números de telefone), ou quando se quer quebrar iteráveis em comprimentos arbitrários. \n\n- _Star expressions_ produzem listas.\n\nPor exemplo: ", "_____no_output_____" ] ], [ [ "_,*m4,_ = (3,4,8,12,16,10)\nm4 # múltiplos de 4", "_____no_output_____" ], [ "_,_,_,*m5 = (4,8,12,5,10 )\nm5 # múltiplos de 5\n", "_____no_output_____" ], [ "*m6,_,_ = (6,12,5,10)\nm6 # múltiplos de 6", "_____no_output_____" ], [ "# 2 star expressions não são permitidas\n*m6,*m5 = (6,12,5,10)", "_____no_output_____" ] ], [ [ ">Ao usar uma _star expression_, certifique-se que o número de variáveis usadas no desempacotamento é consistente com os seus objetivos.", "_____no_output_____" ], [ "### Métodos de tupla\n\nMétodos são funções atribuídas a alguns objetos. Embora tuplas sejam imutáveis, elas possuem métodos bastante úteis. Aqui, consideraremos dois:\n\n- `count`: conta quantas vezes um dado valor aparece na tupla.\n- `index`: localiza o índice de um valor dado.", "_____no_output_____" ], [ "**Exemplo:** Contando valores em tupla.", "_____no_output_____" ] ], [ [ "tupla = (4,3,2,1,1,2,3,4,2,2,2,3,3,3,1,3,1,2,3,4,2,1,3,4,2,2)\n\nprint(tupla.count(0))\nprint(tupla.count(3))\nprint(tupla.count(4))", "0\n8\n4\n" ] ], [ [ "**Exemplo:** Localizando posição.", "_____no_output_____" ] ], [ [ "tupla.index(3)", "_____no_output_____" ] ], [ [ "### _Swap_ de variáveis\n\nO modo Pythônico de realizar uma troca de variáveis (_swap_) é o seguinte:", "_____no_output_____" ] ], [ [ "x,y = 1,2\ny,x = x,y\nx,y", "_____no_output_____" ] ], [ [ "### Tuplas como registros\n\nTuplas servem para armazenar registros: cada item na tupla é o dado de um campo e a posição do item dá o significado. ", "_____no_output_____" ], [ "**Exemplo:** Tuplas como registros. ", "_____no_output_____" ] ], [ [ "# latitute e longitude da Torre Eiffel\neiffel_ll = (48.85844772530444, 2.2948031631859886)\n\n# capitais do mundo\ncaps = [('Afeganistão','Cabul','Ásia'),\n ('Barbados','Bridgetown','América'),\n ('Chade','Jamena','África')]\n\n# população indígena no Brasil por decênio\npopind = [(1991,294131),(2000,734127),(2010,817963)]\n\nfor a,p in popind:\n print(f'Em {a}, o Brasil possuía {p} habitantes indígenas.')", "Em 1991, o Brasil possuía 294131 habitantes indígenas.\nEm 2000, o Brasil possuía 734127 habitantes indígenas.\nEm 2010, o Brasil possuía 817963 habitantes indígenas.\n" ], [ "print('{:14} | {:14} | {:14}'.format('PAÍS','CAPITAL','CONTINENTE'))\nprint('-'*(14*3+2))\nfmt = '{:14} | {:14} | {:14}' \nfor pais, cap, cont in caps:\n print(fmt.format(pais,cap,cont))", "PAÍS | CAPITAL | CONTINENTE \n--------------------------------------------\nAfeganistão | Cabul | Ásia \nBarbados | Bridgetown | América \nChade | Jamena | África \n" ] ], [ [ "Comentários: \n\n- Neste exemplo, simulamos uma tabela usando um formato-base e criamos uma linha de cabeçalho por concatenação de string.", "_____no_output_____" ], [ "## Dicionários \n\nUm dicionário (_dict_) é uma estrutura conhecida como _tabela hash_ ou _array associativo_. É uma coleção de tamanho flexível composta de pares do tipo _chave:valor_. Criamos um `dict` por diversas formas. A mais simples é usar chaves e pares explícitos.", "_____no_output_____" ] ], [ [ "d = {} # dict vazio\nd", "_____no_output_____" ] ], [ [ "Os pares chave-valor incorporam quaisquer tipos de dados.", "_____no_output_____" ] ], [ [ "d = {'par': [0,2,4,6,8], 'ímpar': [1,3,5,7,9], 'nome':'Meu dict', 'teste': True}\nd", "_____no_output_____" ] ], [ [ "### Acesso a conteúdo", "_____no_output_____" ], [ "Para acessar o conteúdo de uma chave, indexamos pelo seu nome.", "_____no_output_____" ] ], [ [ "d['par'] ", "_____no_output_____" ], [ "d['nome']", "_____no_output_____" ] ], [ [ "**Exemplo:** construindo soma e multiplicação especial.", "_____no_output_____" ] ], [ [ "# dict\nop = {'X' :[1,2,3], 'delta' : 0.1}\n\n# função\ndef sp(op): \n s = [x + op['delta'] for x in op['X']]\n p = [x * op['delta'] for x in op['X']]\n \n return (s,p) # retorna tupla\n\nsoma, prod = sp(op) # desempacota\n\nfor i,s in enumerate(soma):\n print(f'pos({i}) | Soma = {s} | Prod = {prod[i]}')", "pos(0) | Soma = 1.1 | Prod = 0.1\npos(1) | Soma = 2.1 | Prod = 0.2\npos(2) | Soma = 3.1 | Prod = 0.30000000000000004\n" ] ], [ [ "### Inserção de conteúdo", "_____no_output_____" ] ], [ [ "# apensa variáveis\nop[1] = 3 \nop['novo'] = (3,4,1) \nop", "_____no_output_____" ] ], [ [ "### Alteração de conteúdo", "_____no_output_____" ] ], [ [ "op['novo'] = [2,1,4] # sobrescreve\nop", "_____no_output_____" ] ], [ [ "### Deleção de conteúdo com `del` e `pop`", "_____no_output_____" ] ], [ [ "del op[1] # deleta chave \nop ", "_____no_output_____" ], [ "novo = op.pop('novo') # retorna e simultaneamente deleta\nnovo", "_____no_output_____" ], [ "op ", "_____no_output_____" ] ], [ [ "### Listagem de chaves e valores\n\nUsamos os métodos `keys()` e `values()` para listar chaves e valores.", "_____no_output_____" ] ], [ [ "arit = {'soma': '+', 'subtr': '-', 'mult': '*', 'div': '/'} # dict\n\nk = list(arit.keys())\nprint(k)\nval = list(arit.values())\nprint(val)\nfor v in range(len(arit)):\n print(f'A operação \\'{k[v]}\\' de \"arit\" usa o símbolo \\'{val[v]}\\'.') ", "['soma', 'subtr', 'mult', 'div']\n['+', '-', '*', '/']\nA operação 'soma' de \"arit\" usa o símbolo '+'.\nA operação 'subtr' de \"arit\" usa o símbolo '-'.\nA operação 'mult' de \"arit\" usa o símbolo '*'.\nA operação 'div' de \"arit\" usa o símbolo '/'.\n" ] ], [ [ "### Combinando dicionários\n\nUsamos `update` para combinar dicionários. Este método possui um resultado similar a `extend`, usado em listas.", "_____no_output_____" ] ], [ [ "pot = {'pot': '**'}\narit.update(pot)\narit", "_____no_output_____" ] ], [ [ "### Dicionários a partir de sequencias\n\nPodemos criar dicionários a partir de sequencias existentes usando `zip`.", "_____no_output_____" ] ], [ [ "arit = ('soma', 'subtr', 'mult', 'div', 'pot')\nops = ('+', '-', '*', '/', '**')\n\ndict_novo = {}\n\nfor chave,valor in zip(arit,ops):\n dict_novo[chave] = valor\n \ndict_novo", "_____no_output_____" ] ], [ [ "Visto que um `dict` é composto de várias tuplas de 2, podemos criar um de maneira ainda mais simples.", "_____no_output_____" ] ], [ [ "dict_novo = dict(zip(arit,ops)) # visto que dicts\ndict_novo", "_____no_output_____" ] ], [ [ "## Conjuntos\n\nUm conjunto (_set_) é uma coleção não ordenada de elementos únicos. Eles podem ser criados através da função `set` ou de uma expressão literal com um par de chaves `{}`. Eles são parecidos com _dicts_, porém não possuem chaves (_keys_).\n", "_____no_output_____" ], [ "**Exemplo:** Crie o conjunto dos números pares positivos e menores do que 15.", "_____no_output_____" ] ], [ [ "pq1 = set(range(0,15,2))\npq2 = {0,2,4,6,8,10,12,14}", "_____no_output_____" ] ], [ [ "Discussão: \n\n- Em `pq1`, criamos o conjunto por função, usando `range` para gerar os números.\n- Em `pq2`, criamos o conjunto literalmente, por extensão.", "_____no_output_____" ] ], [ [ "{1,2,2,3,3,4,4,4} # 'set' possui unicidade de elementos", "_____no_output_____" ] ], [ [ "### Operações com conjuntos \n\nA seguir mostraremos uma série de operações com conjuntos. Considere os seguintes conjuntos.", "_____no_output_____" ] ], [ [ "A = {1,2,3}\nB = {3,4,5}\nC = {6}", "_____no_output_____" ] ], [ [ "#### União de conjuntos", "_____no_output_____" ] ], [ [ "A.union(B) # união", "_____no_output_____" ], [ "A | B # união com operador alternativo ('ou')", "_____no_output_____" ] ], [ [ "#### Atualização de conjuntos (união)\n\nA união *in-place* de dois conjuntos pode ser feita com `update`.", "_____no_output_____" ] ], [ [ "C", "_____no_output_____" ], [ "C.update(B) # C é atualizado com elementos de B\nC ", "_____no_output_____" ], [ "C.union(A) # conjunto união com A", "_____no_output_____" ], [ "C # os elementos de A não foram atualizados em C", "_____no_output_____" ] ], [ [ "A atualização da união possui a seguinte forma alternativa com `|=`.", "_____no_output_____" ] ], [ [ "C |= A # elementos de A atualizados em C\nC", "_____no_output_____" ] ], [ [ "#### Interseção de conjuntos", "_____no_output_____" ] ], [ [ "A.intersection(B) # interseção", "_____no_output_____" ], [ "A & B # interseção com operador alternativo ('e')", "_____no_output_____" ] ], [ [ "#### Atualização de conjuntos (interseção)\n\nA interseção *in-place* de dois conjuntos pode ser feita com `intersection_update`.", "_____no_output_____" ] ], [ [ "D = {1, 2, 3, 4}\nE = {2, 3, 4, 5}", "_____no_output_____" ], [ "D.intersection(E) # interseção com E", "_____no_output_____" ], [ "D # D inalterado", "_____no_output_____" ], [ "D.intersection_update(E) \nD # D alterado", "_____no_output_____" ] ], [ [ "A atualização da interseção possui a seguinte forma alternativa com `&=`.", "_____no_output_____" ] ], [ [ "D &= E\nD", "_____no_output_____" ] ], [ [ "#### Diferença entre conjuntos", "_____no_output_____" ] ], [ [ "A", "_____no_output_____" ], [ "D", "_____no_output_____" ], [ "A.difference(D) # apenas elementos de A", "_____no_output_____" ], [ "D.difference(A) # apenas elementos de D", "_____no_output_____" ], [ "A - D # operador alternativo ", "_____no_output_____" ], [ "D - A ", "_____no_output_____" ] ], [ [ "#### Atualização de conjuntos (diferença)\n\nA interseção *in-place* de dois conjuntos pode ser feita com `difference_update`.", "_____no_output_____" ] ], [ [ "D = {1, 2, 3, 4}\nE = {1, 2, 3, 5}", "_____no_output_____" ], [ "D", "_____no_output_____" ], [ "D.difference(E)\nD", "_____no_output_____" ], [ "D.difference_update(E)\nD", "_____no_output_____" ] ], [ [ "A atualização da diferença possui a seguinte forma alternativa com `-=`.", "_____no_output_____" ] ], [ [ "D -= E\nD", "_____no_output_____" ] ], [ [ "#### Adição ou remoção de elementos", "_____no_output_____" ] ], [ [ "A", "_____no_output_____" ], [ "A.add(4) # adiciona 4 a A\nA", "_____no_output_____" ], [ "B", "_____no_output_____" ], [ "B.remove(3) # remove 3 de B\nB", "_____no_output_____" ] ], [ [ "#### Reinicialização de um conjunto (vazio)\n\nPodemos remover todos os elementos de um conjunto com `clear`, deixando-o em um estado vazio.", "_____no_output_____" ] ], [ [ "A", "_____no_output_____" ], [ "A.clear()\nA # A é vazio", "_____no_output_____" ], [ "len(A) # 0 elementos", "_____no_output_____" ] ], [ [ "#### Continência\n\nPodemos verificar se um conjunto $A$ é subconjunto de (está contido em) outro conjunto $B$ ($A \\subseteq B$) ou se $B$ é um superconjunto para (contém) $A$ ($B \\supseteq A$) com `issubset` e `issuperset`. ", "_____no_output_____" ] ], [ [ "B", "_____no_output_____" ], [ "C", "_____no_output_____" ], [ "B.issubset(C) # B está contido em C", "_____no_output_____" ], [ "C.issuperset(B) # C contém B", "_____no_output_____" ] ], [ [ "#### Subconjuntos e subconjuntos próprios\n\nPodemos usar operadores de comparação entre conjuntos para verificar continência.\n\n- $A \\subseteq B$: $A$ é subconjunto de $B$\n- $A \\subset B$: $A$ é subconjunto próprio de $B$ ($A$ possui elementos que não estão em $B$)", "_____no_output_____" ] ], [ [ "{1,2,3} <= {1,2,3} # subconjunto", "_____no_output_____" ], [ "{1,2} < {1,2,3} # subconjunto próprio", "_____no_output_____" ], [ "{1,2,3} > {1,2}", "_____no_output_____" ], [ "{1,2} >= {1,2,3}", "_____no_output_____" ] ], [ [ "#### Disjunção\n\nDois conjuntos são disjuntos se sua interseção é vazia. Podemos verificar a disjunção com `isdisjoint`", "_____no_output_____" ] ], [ [ "E", "_____no_output_____" ], [ "G = {1,2,3,4}\nG", "_____no_output_____" ], [ "E.isdisjoint(G) # 1,2,3 são comuns", "_____no_output_____" ] ], [ [ "#### Igualdade entre conjuntos\n\nDois conjuntos são iguais se contém os mesmos elementos.", "_____no_output_____" ] ], [ [ "H = {3,'a', 2}\nI = {'a',2, 3}\nJ = {1,'a'}", "_____no_output_____" ], [ "H == I", "_____no_output_____" ], [ "H == J", "_____no_output_____" ], [ "{1,2,2,3} == {3,3,3,2,1} # lembre-se da unicidade", "_____no_output_____" ] ], [ [ "## Compreensões de lista (_listcomps_)", "_____no_output_____" ], [ "**Exemplo:** Construa uma lista dos códigos _Unicode_ correspondentes dos caracteres de uma _string_.", "_____no_output_____" ] ], [ [ "palavra = '!@#$%'\ncods = []\nfor c in palavra:\n cods.append(ord(c))\ncods ", "_____no_output_____" ] ], [ [ "Discussão:\n\n- `ord` produz o número ordinal correspondente de um caracter segundo o padrão Unicode. Para saber o código hexadecimal correspondente, pode-se fazer `hex(ord(c))`.", "_____no_output_____" ], [ "**Exemplo:** Construa uma _listcomp_ correspondente.", "_____no_output_____" ] ], [ [ "palavra = '!@#$%'\ncods = [ord(c) for c in palavra]\ncods", "_____no_output_____" ], [ "# listcomp com quebra de linha sem `\\`\n[ord(c) \n for c \n in palavra]", "_____no_output_____" ] ], [ [ "**Exemplo:** Construa um produto cartesiano de pares $(a,b)$, onde $a$ é um tipo de sangue de acordo com o sistema ABO e $b$ é o Fator Rh.", "_____no_output_____" ] ], [ [ "A = ['A','B','O','AB']\nRh = ['+','-']\nsangue = [(a,b) for a in A for b in Rh]\nsangue", "_____no_output_____" ] ], [ [ "**Exemplo:** Construa um produto cartesiano de pares $(a,b)$, onde $a$ é o FatorRh e $b$ é um tipo de sangue de acordo com o sistema ABO.", "_____no_output_____" ] ], [ [ "[(b,a) for a in A for b in Rh]", "_____no_output_____" ] ], [ [ "**Exemplo:** Use _listcomps_ com condicionais para criar filtros.", "_____no_output_____" ] ], [ [ "# equipes da fórmula 1\nf1 = ['Red Bull Racing','Mercedes',\n 'McLaren','Ferrari','AlphaTauri',\n 'Aston Martin','Alpine','Alfa Romeo Racing',\n 'Williams','Haas F1 Team']\n\n# equipes que começam com A, em maiúsculas\n[team.upper() for team in f1 if team[0] == 'A']", "_____no_output_____" ], [ "# equipes cujo nome não inicia com A e tem menos do que 10 caracteres\n[team for team in f1 \n if team[0] != 'A' and len(team) < 10]", "_____no_output_____" ] ], [ [ "## Compreensões de conjuntos\n\nSemelhantemente a _listcomps_, podemos realizar \"_setcomps_\".", "_____no_output_____" ] ], [ [ "# comprimentos de nomes de equipes (únicos)\n{len(team) for team in f1}", "_____no_output_____" ], [ "[len(team) for team in f1]", "_____no_output_____" ], [ "# comprimento de nomes de equipes que terminam com \"ing\"\n{len(team) for team in f1 if team.endswith('ing')}", "_____no_output_____" ] ], [ [ "## Compreensões de dicionários\n\nCompreensões de dicionários, ou \"_dictcomps_\", possuem resultados similares às anteriores e são convenientes para criar _dicts_.", "_____no_output_____" ] ], [ [ "{k:v for k,v in enumerate(f1)}", "_____no_output_____" ], [ "# enumera a partir do índice 7, mas percorre toda a lista\n{k:v for k,v in enumerate(f1,7)}", "_____no_output_____" ] ], [ [ "## _Sorted_ \n\n- Listas em Python possuem um método chamado `sort` com o qual podemos realizar uma ordenação _in-place_ de uma lista. \n\n- A função `sorted` permite que façamos o mesmo criando uma nova lista e, dessa maneira, não alterando a lista original.", "_____no_output_____" ] ], [ [ "# copia lista f1\nf1c = f1.copy()\nf1c", "_____no_output_____" ], [ "# lista ordenada\nf1c.sort()\nf1c", "_____no_output_____" ], [ "# lista f1 ordenada\nf1c2 = sorted(f1)\nf1c2", "_____no_output_____" ], [ "# lista original intacta\nf1", "_____no_output_____" ], [ "# ordenação re\nsorted(f1,reverse=True)", "_____no_output_____" ] ], [ [ "## Aleatoriedade\n\n- Números aleatórios, ou também chamados de randômicos, são úteis para fabricar experimentos estatísticos, realizar amostragem e desenvolver uma série de estudos envolvendo probabilidade. \n\n- Em Python, podemos usar o módulo `random` para atingir uma diversidade de propósitos em ciência de dados. Vamos explorar algumas funções para geração aleatória: `choice`, `random`, `randint`, `shuffle` e `sample`.", "_____no_output_____" ] ], [ [ "import random", "_____no_output_____" ] ], [ [ "**Exemplo:** Crie um experimento de lançamento de dados e realize uma escolha aleatória em $n$ lançamentos.", "_____no_output_____" ] ], [ [ "dado = list(range(1,7))\n\nn = 4\nfor _ in range(n):\n print(random.choice(dado)) ", "5\n5\n1\n6\n" ] ], [ [ "Comentários:\n\n- O _loop_ será repetido $n$ vezes.\n- O método `choice` escolhe um valor ao acaso entre os disponíveis na lista. ", "_____no_output_____" ], [ "**Exemplo:** Crie um experimento de probabilidade aleatório que determine a salinidade presente em 1 litro de água marinha (3 - 5%). ", "_____no_output_____" ] ], [ [ "na = 5\nsaly = []\nfor _ in range(na):\n saly.append(0.03 + (0.05-0.03)*random.random())\nsaly ", "_____no_output_____" ] ], [ [ "Comentários: \n\n- Este experimento equivale a coletar 5 amostras de água marinha com salinidade variável entre 3 e 5%.\n- O método `random` determina um número aleatório no intervalo [0,1).", "_____no_output_____" ], [ "**Exemplo:** Crie um experimento que ordena 15 arremessadores para cestas de 3 pontos durante um treinamento do time do Los Angeles Lakers (0 a 99) excluindo a lista de números reservados. Considere que se o número sorteado não estiver atribuído a nenhum jogador atual ou se já tiver sido sorteado, o técnico escolherá o próximo da lista.\n\nLista de números reservados: 8,13,17,19,22,24,25,32,33,34,42,44,52,99.", "_____no_output_____" ] ], [ [ "reservados = [8,13,17,19,22,24,25,32,33,34,42,44,52,99]\n\nshooter = []\nwhile len(shooter) != 15:\n x = random.randint(0,99)\n if x not in reservados and x not in shooter:\n shooter.append(x) \n\nprint('Los Angeles Lakers :: 3-point shooters (Training Schedule):')\nfor s in shooter:\n print(s,end=' ') \n ", "Los Angeles Lakers :: 3-point shooters (Training Schedule):\n89 70 40 56 4 91 48 5 3 0 98 59 58 75 81 " ] ], [ [ "Comentários: \n\n- O _loop_ _while_ encerrará quando uma lista com 15 arremessadores estiver completa.\n- A estrutura condicional previne que números já sorteados entrem na lista.\n- O método `randint` escolhe um número inteiro aleatório entre 0 e 99.", "_____no_output_____" ], [ "**Exemplo:** Crie uma alteração no schedule de treinamento anterior a partir da última lista disponível.", "_____no_output_____" ] ], [ [ "print(random.shuffle(shooter))\nshooter ", "None\n" ] ], [ [ "Comentário: \n\n- O método `shuffle` reordena a lista `shooter` randomicamente e produz um valor `None` como saída. Dessa forma, geramos uma nova lista com os mesmos arremessadores pré-selecionados, mas em ordem distinta.", "_____no_output_____" ], [ "**Exemplo:** Construa uma tabela aleatória de 3 jogos de beisebol entre times da MLB.", "_____no_output_____" ] ], [ [ "# lê arquivo com nome de times da MLB\nwith open('../database/mlb-teams.csv') as f:\n mlb = f.read().splitlines()\n \nfor _ in range(3):\n a,b = random.sample(mlb,2)\n print(f'Game {_+1}: {a} x {b}') ", "Game 1: Pittsburgh Pirates x Texas Rangers\nGame 2: Philadelphia Phillies x Chicago Cubs\nGame 3: St. Louis Cardinals x Tampa Bay Rays\n" ] ], [ [ "Comentários: \n\n- O método `sample` realiza uma amostragem de $k$ valores disponíveis na lista. Aqui, escolhemos $k=2$, de modo que 2 times são escolhidos ao acaso para comporem 3 jogos.", "_____no_output_____" ], [ "## Expressões regulares\n\nExpressões regulares (_regular expressions_), também conhecidas como _RE_, _regexes_, ou _regex pattern_ são regras estabelecidas para localizar padrões desejados em strings. REs, na verdade, formam uma pequena linguagem de programação dentro de uma linguagem maior. ", "_____no_output_____" ], [ "### Padrões simples, caracterese ordinários e metacaracteres\n\n- A tarefa mais simples que podemos executar com REs é combinar caracteres.\n\n_Caracteres ordinários_ são basicamente todos os caracteres ASCII alfanuméricos, tais como `'b'`, `'B'` e `'1'`.\n\n_Caracteres especiais_ são todos os metacaracteres. A lista dos metacaracteres é a seguinte: \n\n```\n. ^ $ * + ? { } [ ] \\ | ( )\n``` ", "_____no_output_____" ] ], [ [ "# imprime caracteres ASCII mais comuns \n# end = ' ' altera a terminação padrão '\\n'\nfor i in range(33,127):\n print(chr(i), end = ' ') ", "! \" # $ % & ' ( ) * + , - . / 0 1 2 3 4 5 6 7 8 9 : ; < = > ? @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \\ ] ^ _ ` a b c d e f g h i j k l m n o p q r s t u v w x y z { | } ~ " ] ], [ [ "### O módulo `re`\n\n- Em Python, temos à disposição o módulo `re` para trabalhar com REs. \n\n- Vamos ler o arquivo bras-cubas.txt e salvar o texto em uma _string_.", "_____no_output_____" ] ], [ [ "with open('../database/bras-cubas.txt', 'r') as f:\n cubas = f.read()\ncubas \n", "_____no_output_____" ], [ "# importa o módulo\nimport re", "_____no_output_____" ] ], [ [ "As principais funções do módulo `re` podem ser resumidas ao seguinte modelo:\n\n```python\nre.f(pattern,string)\n```\nonde _f_ é o nome da função, _pattern_ é o padrão, isto é, a _regex_, e _string_ é a string na qual o padrão é procurado. Exploraremos algumas delas a seguir compreendendo, simultaneamente, o papel desempenhado pelos metacaracteres. ", "_____no_output_____" ], [ "**Exemplo:** Procure pelo padrão \"que\" no texto-base e conte quantas vezes ele aparece.", "_____no_output_____" ] ], [ [ "# busca todas as combinações\nque = re.findall('que',cubas)\nprint(que)\nprint(len(que))", "['que', 'que', 'que', 'que', 'que', 'que', 'que']\n7\n" ], [ "# padrão presente, mas não é palavra inteligível\nre.findall('xou',cubas)", "_____no_output_____" ] ], [ [ "Buscas por caracteres ordinários são sempre exatas, no sentido de que o caracter é combinado como ele é. Vejamos um exemplo que conta as vogais que aparecem no texto.", "_____no_output_____" ] ], [ [ "for vogal in ['a','e','i','o','u']: \n print(f'A vogal \\'{vogal}\\' foi \\\nlocalizada {len(re.findall(vogal,cubas))} vezes no texto.')", "A vogal 'a' foi localizada 143 vezes no texto.\nA vogal 'e' foi localizada 104 vezes no texto.\nA vogal 'i' foi localizada 65 vezes no texto.\nA vogal 'o' foi localizada 83 vezes no texto.\nA vogal 'u' foi localizada 43 vezes no texto.\n" ] ], [ [ "**Exemplo:** Busque pelo padrão \"Tijuca\" no texto e determine suas posições de início e término.", "_____no_output_____" ] ], [ [ "# busca por combinação retornado posição\np = re.search('Tijuca',cubas)\np.start(), p.end(), p.group()", "_____no_output_____" ] ], [ [ "Discussão: \n\n- O método `search` procura pelo padrão `'Tijuca'` em `cubas` e retorna três informações:\n - `start`: a posição inicial do padrão\n - `end`: a posição final do padrão\n - `group`: o padrão localizado\n- O padrão aparece apenas uma vez no texto na fatia `cubas[1042:1048]`.", "_____no_output_____" ], [ "Comentários: \n\n- Caso houvesse repetição do padrão, apenas os índices da primeira aparição seriam retornados. Vejamos: ", "_____no_output_____" ] ], [ [ "p2 = re.search('que',cubas)\np2.start(), p2.end(), p2.group()", "_____no_output_____" ] ], [ [ "Vimos anteriormente que há 7 aparições do padrão \"que\" no texto-base. Porém, apenas a primeira é retornada por `search`. Vejamos:", "_____no_output_____" ] ], [ [ "# uma maneira de mostrar que há apenas\n# um 'que' até o caracter 182 de 'cubas'\n# por meio de split e interseção de conjuntos\nset(cubas[0:182].split(' ')).intersection({'que'}) ", "_____no_output_____" ] ], [ [ "Como este conjunto é unitário, há apenas um padrão 'que' até a posição 182 de `cubas`.", "_____no_output_____" ], [ "**Exemplo:** Busque pelo padrão \"Não\" no texto e verifique se ele aparece no início da string.", "_____no_output_____" ] ], [ [ "p3 = re.match('Não',cubas)\np3.start()", "_____no_output_____" ] ], [ [ "**Exemplo:** Substitua o padrão \"Virgília\" por \"Ofélia\" em toda a string do texto.", "_____no_output_____" ] ], [ [ "cubas_novo = re.sub('Virgília','Ofélia',cubas) \ncubas_novo", "_____no_output_____" ] ], [ [ "**Exemplo:** Substitua o padrão \"Virgília\" por \"Ofélia\" apenas nas duas primeiras aparições", "_____no_output_____" ] ], [ [ "cubas_novo2 = re.sub('Virgília','Ofélia',cubas,count=2)\ncubas_novo2 ", "_____no_output_____" ], [ "# original\nprint(len(re.findall('Virgília',cubas)))\n\n# \"Virgília\" -> \"Ofélia\" integralmente\nprint(len(re.findall('Virgília',cubas_novo)))\n\n# \"Virgília\" -> \"Ofélia\" em 2 ocorrências apenas\nprint(len(re.findall('Virgília',cubas_novo2)))", "4\n0\n2\n" ] ], [ [ "**Exemplo:** Seccione o texto em várias substrings usando a vírgula (\",\") como ponto de quebra:", "_____no_output_____" ] ], [ [ "cubas_sep = re.split(',',cubas) \ncubas_sep", "_____no_output_____" ] ], [ [ "**Exemplo:** Separe a string do texto em no máximo 3 substrings usando o padrão \"ão\" como ponto de quebra. ", "_____no_output_____" ] ], [ [ "# maxsplit = n; logo, n + 1 substrings\ncubas_sep3 = re.split(\"ão\",cubas,maxsplit=2)\ncubas_sep3", "_____no_output_____" ] ], [ [ "### Sintaxes especiais\n\nOs metacaracteres podem ser usados para criar REs das formas mais variadas (e criativas) possíveis. Vejamos exemplos.", "_____no_output_____" ], [ "#### Metacaracter `.`\n\nServe para combinar qualquer caracter, exceto o _newline_ (`\\n`), se usado no modo padrão. Em uso não padrão, pode também combinar-se com `\\n`.", "_____no_output_____" ], [ "**Exemplo:** Localize padrões com 5 caracteres contendo \"lh\" na 3a. e 4a. posições.", "_____no_output_____" ] ], [ [ "re.findall('..lh.',cubas)", "_____no_output_____" ] ], [ [ "**Exemplo:** Localize padrões iniciando com a letra T maiúscula e contendo, ao todo, 3 caracteres.", "_____no_output_____" ] ], [ [ "re.findall('T...',cubas) ", "_____no_output_____" ] ], [ [ "#### Metacaracter `\\`\n\nServe para criar escapes de caracteres especiais ou sinalizar uma sequência especial.", "_____no_output_____" ], [ "**Exemplo:** Quebre a string usando o ponto final (\".\") como ponto de quebra.", "_____no_output_____" ] ], [ [ "# escape do metacaracter '.'\nmt1 = re.split('\\.',cubas)\nmt1", "_____no_output_____" ], [ "# considera um espaço no padrão\nmt2 = re.split('\\. ',cubas)\nmt2 ", "_____no_output_____" ] ], [ [ "#### Metacaracter `^`\n\nTambém chamado _caret_, serve para buscar padrões no início de uma string. ", "_____no_output_____" ], [ "**Exemplo:** Use as substrings do exemplo anterior para buscar quais delas iniciam-se por `'T'`.", "_____no_output_____" ] ], [ [ "for k,s in enumerate(mt2): \n aux = re.match('^T',s)\n if aux: # se lista não for vazia\n print(f'---> Padrão detectado na \\\nsubstring {k}:\\n\\\"{s}\\\"')", "---> Padrão detectado na substring 1:\n\"Talvez eu exponha ao leitor, em algum canto deste livro, a minha teoria das edições humanas\"\n" ] ], [ [ "#### Metacaracter `$`\n\nVisto em sentido oposto a `^`, `$` serve para buscar padrões no final de uma string, ou logo antes de um caracter _newline_.", "_____no_output_____" ], [ "**Exemplo:** Use as substrings do exemplo anterior para buscar quais delas terminam por `'o'`.", "_____no_output_____" ] ], [ [ "for k,s in enumerate(mt2):\n aux = re.findall('o$',s)\n if aux: # se lista não for vazia\n print(f'---> Padrão detectado na \\\nsubstring {k}:\\n\\\"{s}\\\"')", "---> Padrão detectado na substring 0:\n\"Não durou muito a evocação; a realidade dominou logo; o presente expeliu o passado\"\n---> Padrão detectado na substring 2:\n\"O que por agora importa saber é que Virgília — chamava-se Virgília — entrou na alcova, firme, com a gravidade que lhe davam as roupas e os anos, e veio até o meu leito\"\n---> Padrão detectado na substring 4:\n\"Era um sujeito, que me visitava todos os dias para falar do câmbio, da colonização e da necessidade de desenvolver a viação férrea; nada mais interessante para um moribundo\"\n" ], [ "# busca substring terminando com 'undo'\nfor k,s in enumerate(mt2):\n aux = re.findall('undo$',s)\n if aux: # se lista nã o for vazia\n print(f'---> Padrão detectado na \\\nsubstring {k}:\\n\\\"{s}\\\"')", "---> Padrão detectado na substring 4:\n\"Era um sujeito, que me visitava todos os dias para falar do câmbio, da colonização e da necessidade de desenvolver a viação férrea; nada mais interessante para um moribundo\"\n" ] ], [ [ "Comentário:\n\n- Note que para buscar o caracter literal '$' em uma string devemos usar o caracter de escape. Vejamos o próximo exemplo.", "_____no_output_____" ] ], [ [ "real = 'João me pagou R$ 150,00.'\nv = re.search('\\$',real)\nprint(f'Caracter $ identificado na posição {v.start()}.')", "Caracter $ identificado na posição 15.\n" ] ], [ [ "#### Metacaracter `*`\n\nServe para buscar 0 ou mais repetições de um caracter precedente. Este caracter especial é de _repetição_.", "_____no_output_____" ], [ "**Exemplo:**", "_____no_output_____" ] ], [ [ "re.findall('hav*.',cubas)", "_____no_output_____" ] ], [ [ "Discussão: \n\n- A RE `'hav*.'` procura por padrões que atendem às seguintes restrições: \n - o caracter 'v' aparece 0 ou mais vezes após 'ha'.\n - quando não houver 'v', o terceiro caracter passa a ser qualquer um que vier na sequência por causa do metacaracter '.'. \n - quando houver um 'v', o quarto caracter é qualquer um\n - quando houver n repetições de 'v' após 'a', todos os n são considerados, e o n+1-ésimo caracter é qualquer um.", "_____no_output_____" ], [ "Comentários: \n\n- Descrever em palavras o que a expressão regular faz exatamente não é sempre muito fácil. Vejamos mais um exemplo.\n ", "_____no_output_____" ] ], [ [ "c = 'ha hahaa hav havv havva-havvva; haiiaa havia'\n\nre.findall('hav*.',c)", "_____no_output_____" ] ], [ [ "#### Metacaracter `+`\n\nServe para buscar 1 ou mais repetições de um caracter precedente. Assim como `*`, `+` também é um caracter especial de _repetição_.", "_____no_output_____" ], [ "**Exemplo:**", "_____no_output_____" ] ], [ [ "re.findall('hav+.',cubas)", "_____no_output_____" ] ], [ [ "Discussão: \n\n- A RE `'hav+.'` é similar à anterior. Porém, pelo fato de ser obrigatória a ocorrência de pelo menos 1 caracter 'v' após 'a', os padrões 'ha ', 'ham' e 'har' são excluídos.\n- Se encontradas n ocorrências de 'v', o padrão será encerrado pelo n+1-ésimo caracter arbitrário\n- Como no texto existe tal padrão com apenas uma ocorrência de 'v', o quarto caracter, sendo arbitrário, é encontrado como 'i'.\n- O padrão 'havia', que forma uma palavra compreensível em Português, é ignorado por conter 5 caracteres.", "_____no_output_____" ], [ "#### Metacaracter `?`\n\nServe para buscar 0 ou 1 repetição de um caracter precedente. Assim como `*`, `+` também é um caracter especial de _repetição_.", "_____no_output_____" ], [ "**Exemplo:**", "_____no_output_____" ] ], [ [ "re.findall('s?an.',cubas) ", "_____no_output_____" ] ], [ [ "Comentários: \n\n- Esta RE procurará por 0 ou 1 repetição de 's' seguida por 'an' e qualquer outro caracter. \n\n- Neste caso, a substring 'ssan', por exemplo, existente na palavra 'interessante' não seria coberta por `?`.", "_____no_output_____" ], [ "#### Metacaracteres `[]`\n\nOs colchetes devem ser usados, juntos, em par, para indicar um conjunto de caracteres sobre os quais a expressão regular deve operar. Em geral, usa-se o par `[]` para os seguintes propósitos: ", "_____no_output_____" ], [ "- Listar caracteres individualmente\n\nA RE `[vnd]` combinará os caracteres `'v'`, `'n'` ou `'d'`.", "_____no_output_____" ], [ "**Exemplo:**", "_____no_output_____" ] ], [ [ "# localiza todas as ocorrências \n# de 'v', 'n', ou 'd'\nprint(re.findall('[vnd]',cubas))", "['d', 'v', 'd', 'd', 'd', 'n', 'n', 'd', 'v', 'n', 'n', 'd', 'v', 'n', 'd', 'd', 'n', 'v', 'n', 'n', 'v', 'v', 'd', 'd', 'd', 'v', 'n', 'v', 'n', 'v', 'n', 'v', 'v', 'd', 'd', 'd', 'd', 'n', 'd', 'n', 'd', 'd', 'd', 'd', 'n', 'v', 'v', 'v', 'n', 'd', 'n', 'n', 'n', 'd', 'd', 'd', 'd', 'n', 'v', 'd', 'd', 'n', 'd', 'n', 'd', 'd', 'd', 'n', 'd', 'v', 'v', 'n', 'n', 'd', 'v', 'n', 'd', 'd', 'v', 'd', 'v', 'd', 'd', 'd', 'n', 'd', 'n', 'd', 'n', 'd', 'v', 'n', 'v', 'n', 'd', 'n', 'd', 'v', 'v', 'n', 'd', 'n', 'd', 'v', 'n', 'n', 'd']\n" ] ], [ [ "- Localizar um intervalo de caracteres\n\nUma RE do tipo `[a-d]` combinará os caracteres de `'a'` a `'d'`, ao passo que `[a-z]` combinará todas as letras minúsculas da tabela ASCII (e não Unicode!), ou seja, de `'a'` a `'z'`.\n\nDo mesmo modo, `[0-4]` combinará os dígitos de 0 a 4, e `[0-9]` todos os dígitos decimais. Por outro lado, a expressão `[0-3][3-8]` combinará todos os números de dois dígitos de 03 a 38.\n\nSe o símbolo `-` for escapado (p.ex. `[a\\-d]`), ou se for colocado como primeiro ou último caracter (p.ex. `[-b]` ou `[b-]`), ele significará o hífen literal ('-').", "_____no_output_____" ], [ "**Exemplo:**", "_____no_output_____" ] ], [ [ "re.findall('[a-z]um',cubas)", "_____no_output_____" ] ], [ [ "Discussão: \n\n- Esta RE procurará todos os padrões que iniciam por qualquer letra minúscula (de 'a' a 'z') seguidas por 'um'.", "_____no_output_____" ], [ "**Exemplo:**", "_____no_output_____" ] ], [ [ "re.findall('[v-z][a-].',cubas)", "_____no_output_____" ] ], [ [ "Discussão: \n\n- Esta RE procurará todos os padrões que iniciam por uma letra minúscula entre 'v' e 'z' seguidas por 'a', '-', ou 'a-' e um caracter qualquer adicional.", "_____no_output_____" ], [ "**Exemplo:**", "_____no_output_____" ] ], [ [ "# mesmo efeito que anterior\nre.findall('[v-z][-a].',cubas)", "_____no_output_____" ] ], [ [ "- Localizar caracteres especiais como literais\n\nOs caracteres especiais perdem seu status quando aparecem entre colchetes. Por exemplo, a RE `[(+*)]` procurará por qualquer um dos literais `(`, `+`, `*` ou `)`.", "_____no_output_____" ] ], [ [ "ex = 'Quanto é (22 + 3)*1 - 6*1?'\nre.findall('[0-9][)*]',ex)", "_____no_output_____" ] ], [ [ "Discussão:\n\n- Esta RE busca padrões em que um dígito de 0 a 9 precede um ')' ou '*' literalmente.", "_____no_output_____" ] ], [ [ "re.findall('[0-9][?]',ex)", "_____no_output_____" ] ], [ [ "Discussão:\n\n- Esta RE busca padrões em que um dígito de 0 a 9 precede um '?' literal.", "_____no_output_____" ], [ "#### Metacaracteres `{}`\n\nAs chaves, assim como os colchetes, devem ser usadas juntas, em par, para localizar cópias ou repetições de uma expressão regular. Em geral, ela é usada para os seguintes objetivos:", "_____no_output_____" ], [ "- Localizar exatamente _m_ cópias de uma RE anterior", "_____no_output_____" ], [ "**Exemplo:**", "_____no_output_____" ] ], [ [ "# busca pelo padrão 'rr'\nre.search('r{2}',cubas).group()", "_____no_output_____" ] ], [ [ "**Exemplo:**", "_____no_output_____" ] ], [ [ "ex2 = '0110011000110011001010110100110111111000'\nre.findall('0{2}',ex2)", "_____no_output_____" ], [ "re.findall('1{3}',ex2)", "_____no_output_____" ], [ "# busca 000000 - 111111\n# com variações de 0,1 intermediárias\nre.findall('[0-1]{6}',ex2)", "_____no_output_____" ] ], [ [ "- Localizar entre _m_ e _n_ repetições de uma RE anterior", "_____no_output_____" ], [ "**Exemplos:**", "_____no_output_____" ] ], [ [ "set(re.findall('10{2,3}',ex2))", "_____no_output_____" ], [ "set(re.findall('01{2,3}',ex2))", "_____no_output_____" ], [ "set(re.findall('[0-1]{2,3}',ex2))", "_____no_output_____" ] ], [ [ "#### Metacaracter `|`\n\nUsamos o _pipe_ para significar \"ou\" com o objetivo de localizar mais de uma RE. Por exemplo, `A|B` buscará padrões determinados pela RE `A` **ou** por `B`. _Pipes_ adicionais constroem sintaxes conjuntivas. Ou seja, `A|B|C|D...|Z` significaria \"A ou B ou C ou ... D\".", "_____no_output_____" ], [ "**Exemplos:**", "_____no_output_____" ] ], [ [ "# busca por 'rre' ou 'ssa'\nre.findall('r{2}e|s{2}a',cubas)", "_____no_output_____" ], [ "# busca por 'cha', 'cho' ou\n# 'lha', 'lhi' ou \n# 'pra', 'pre', 'pri', 'pro', 'pru'\nre.findall('ch[ao]|lh[ai]|pr[aeiou]',cubas)", "_____no_output_____" ] ], [ [ "#### Metacaracteres `()`\n\nServem para criar _grupos de captura_. Um grupo de captura é formado por uma RE confinada entre parênteses. Grupos de captura podem ser aplicados para extrair padrões que possuam uma certa estrutura. ", "_____no_output_____" ], [ "**Exemplo**:", "_____no_output_____" ] ], [ [ "table = ['alfa00-LAX','beta22-PET', 'zeta92-XIR']\n\nfor t in table: \n pat = re.match(\"([a-z][a-z][a-z][a-z][0-9][0-9])-([A-Z]{3})\", t)\n if pat:\n print(pat.groups())", "('alfa00', 'LAX')\n('beta22', 'PET')\n('zeta92', 'XIR')\n" ] ], [ [ "Discussão: \n\n- A RE acima é composta de dois grupos de captura que são separados pelo `-` literal. \n\n- O primeiro grupo de captura é `([a-z][a-z][a-z][a-z][0-9][0-9])`\n\n- O segundo grupo de captura é `([A-Z]{3})`\n\n- Tuplas com dois elementos são impressas com `groups()`.", "_____no_output_____" ], [ "Comentários: \n\n- Como se vê, os grupos de captura permitiram que identificássemos padrões separados em uma lista de strings que continha uma estrutura predefinida. \n\n- Entretanto, as REs anteriores criadas como grupos de captura poderiam ser definidas de uma forma muito mais concisa através de _identificadores_.", "_____no_output_____" ], [ "#### Identificadores e _tokens_ gerais\n\nAo trabalhar com REs, podemos utilizar identificadores e _tokens_ diversos para significar padrões de maneira concisa. \n\nO exemplo anterior sobre grupos de captura, por exemplo, poderia ser escrito de outra forma:", "_____no_output_____" ] ], [ [ "for t in table:\n pat = re.match(\"(\\w+)\\W(\\w+)\", t)\n if pat:\n print(pat.groups())", "('alfa00', 'LAX')\n('beta22', 'PET')\n('zeta92', 'XIR')\n" ] ], [ [ "Neste caso, `\\w` e `\\W` são chamados _identificadores_.", "_____no_output_____" ], [ "A tabela a seguir resume os principais identificadores e seu significado.\n\n|Identificador|Significado|\n|---|---|\n|`\\d`|qualquer caracter que é um dígito decimal. Equivalente a `[0-9]`|\n|`\\D`|qualquer caracter exceto dígido (não-dígito). Equivalente a `[^0-9]`|\n|`\\w`|qualquer caracter Unicode. Se a flag `ASCII` for utilizada, equivale ao conjunto `[a-zA-Z0-9_]`|\n|`\\W`|o oposto de `\\w`. Se a flag `ASCII` for utilizada, equivale ao conjunto `[^a-zA-Z0-9_]`|\n|`\\s`|qualquer caracter que é um espaço.|\n|`\\S`|qualquer caracter que não seja espaço.|", "_____no_output_____" ], [ "A tabela a seguir resume _tokens_ gerais e seu significado.\n\n|_Token_|Significado|\n|---|---|\n|`\\n`|quebra de linha. _Newline_.|\n|`\\r`|_carriage return_|\n|`\\t`|TAB|\n|`\\0`|caracter nulo|", "_____no_output_____" ], [ "**Exemplo:** Identificando padrões de data e hora. ", "_____no_output_____" ] ], [ [ "dt = ['2021-04-06 10:32:00', '2020-12-03 01:12:58'] \nfor m in dt:\n pat = re.match('(\\d+)-(\\d+)-(\\d+)\\s(\\d+):(\\d+):(\\d+)',m)\n if pat:\n print(pat.groups())", "('2021', '04', '06', '10', '32', '00')\n('2020', '12', '03', '01', '12', '58')\n" ] ], [ [ "**Exemplo:** Identificando cores em padrão hexadecimal.", "_____no_output_____" ] ], [ [ "cores = ['(1,23,43)','#3ed4f4', '#ffcc00','(C,M,Y,K)','#9999ff'] \n\nfor c in cores:\n ok = re.match('#[a-fA-F0-9]{6}',c)\n if ok: \n print(ok.group())", "#3ed4f4\n#ffcc00\n#9999ff\n" ] ], [ [ "**Exemplo:** Identificando URLs no domínio Wikipedia Português.", "_____no_output_____" ] ], [ [ "urls = ['http://pt.wikipedia.org/wiki/Bras',\n 'http://wikipedia.org/wiki/Bras',\n 'https://pt.wikipedia.org/wiki/Bras',\n 'http://pt.wikipedia.org/wik/Bras',\n 'https://pt.wikipedia.org/wiki/Cubas']\n\nfor u in urls:\n ok = re.match('((http|https)://)?(pt.wikipedia.org/wiki/).+',u)\n if ok:\n print(ok.group())", "http://pt.wikipedia.org/wiki/Bras\nhttps://pt.wikipedia.org/wiki/Bras\nhttps://pt.wikipedia.org/wiki/Cubas\n" ] ], [ [ "#### Compilação e literais\n\nAté agora abordamos REs de uma maneira direta, mostrando como podemos localizar padrões pela utilização de alguns métodos. Vale comentar que REs podem ser compiladas como objetos mais abstratos. \n\nPara compilar expressões, podemos utilizar o método `compile`, no qual REs são manipuladas como strings. ", "_____no_output_____" ], [ "**Exemplo:**", "_____no_output_____" ] ], [ [ "rex = re.compile('[p-t]a[b-d]a?')\nrex.findall(cubas)", "_____no_output_____" ] ], [ [ "Aqui, `rex` é um objeto abstrato.", "_____no_output_____" ] ], [ [ "print(type(rex))", "<class 're.Pattern'>\n" ] ], [ [ "Se imprimirmos `rex`, veremos o seguinte:", "_____no_output_____" ] ], [ [ "rex", "_____no_output_____" ] ], [ [ "O prefixo `r` torna a string literal e serve para tratar problemas com escape. Todavia, deixaremos esta discussão para um momento posterior. Para saber mais, leia sobre a [Praga do Backspace](https://docs.python.org/3/howto/regex.html#the-backslash-plague).", "_____no_output_____" ], [ "#### Recursos de aprendizagem\n\nPara aprofundar seu entendimento sobre expressões regulares, recomendamos os seguintes sites: \n\n- [regex101.com](https://regex101.com)\n- [regexr.com](https://regexr.com)\n- [Regular Expression HOWTO](https://docs.python.org/3/howto/regex.html)\n- [Documentação do módulo `re`](https://docs.python.org/3/library/re.html)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown", "markdown" ], [ "code", "code", "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ] ]

ec7bc75d8befe1ceb403121c2aaa4adda5358229

15,629

ipynb

Jupyter Notebook

examples/example.ipynb

ruber0id/gaapi4py

fbb0221b7cff3c0945204f34e71d8db6006364c9

[ "MIT" ]

21

2019-08-23T15:53:33.000Z

2021-11-21T05:47:02.000Z

examples/example.ipynb

ruber0id/gaapi4py

fbb0221b7cff3c0945204f34e71d8db6006364c9

[ "MIT" ]

5

2019-08-11T09:53:19.000Z

2021-04-30T20:51:08.000Z

examples/example.ipynb

ruber0id/gaapi4py

fbb0221b7cff3c0945204f34e71d8db6006364c9

[ "MIT" ]

16

2019-08-19T17:23:22.000Z

2022-01-13T18:40:43.000Z

28.211191

91

0.399578

[ [ [ "### Import libraries", "_____no_output_____" ] ], [ [ "from datetime import date, timedelta\nimport pandas as pd\nfrom gaapi4py import GAClient", "_____no_output_____" ] ], [ [ "### Define constants", "_____no_output_____" ] ], [ [ "PATH_TO_SERVICE_ACCOUNT = 'PATH/TO/SERVICE_ACCOUNT.json'\nVIEW_ID = '<YOUR_VIEW_ID>'\n\nSESSION_ID_CD_INDEX = '1'\nHIT_ID_CD_INDEX = '2'", "_____no_output_____" ] ], [ [ "### Instantiate the client", "_____no_output_____" ] ], [ [ "c = GAClient(PATH_TO_SERVICE_ACCOUNT)\nc.set_view_id(VIEW_ID)", "_____no_output_____" ], [ "request_body = {\n 'view_id': VIEW_ID,\n 'start_date': '2019-01-01',\n 'end_date': '2019-01-31',\n 'dimensions': {\n 'ga:sourceMedium',\n 'ga:date'\n },\n 'metrics': {\n 'ga:sessions'\n }\n}\nresponse = c.get_all_data(request_body)", "_____no_output_____" ], [ "response['info']", "_____no_output_____" ], [ "response['data'].head(2) # Pandas dataframe that contains data from GA", "_____no_output_____" ] ], [ [ "### Get data daily to avoid sampling", "_____no_output_____" ] ], [ [ "start_date = date(2019,7,1)\nend_date = date(2019,7,14)\n\ndf_list = []\niter_date = start_date\nwhile iter_date <= end_date:\n c.set_dateranges(iter_date, iter_date)\n response = c.get_all_data({\n 'dimensions': {\n 'ga:sourceMedium',\n 'ga:deviceCategory'\n },\n 'metrics': {\n 'ga:sessions'\n }\n })\n df = response['data']\n df['date'] = iter_date\n df_list.append(response['data'])\n iter_date = iter_date + timedelta(days=1)\n \nall_data = pd.concat(df_list, ignore_index=True)", "_____no_output_____" ] ], [ [ "### Get per-session data (using session_id custom dimension)", "_____no_output_____" ] ], [ [ "one_day = date(2019,7,1)\nc.set_dateranges(one_day, one_day)\n\nsession_id = 'dimension' + SESSION_ID_CD_INDEX\nhit_id = 'dimension' + HIT_ID_CD_INDEX\n\nresponse_1 = c.get_all_data({\n 'dimensions': {\n 'ga:' + session_id,\n 'ga:sourceMedium',\n 'ga:campaign',\n 'ga:keyword',\n 'ga:adContent',\n 'ga:userType',\n 'ga:deviceCategory'\n },\n 'metrics': {\n 'ga:sessions'\n }\n})\n\nresponse_2 = c.get_all_data({\n 'dimensions': {\n 'ga:' + session_id,\n 'ga:landingPagePath',\n 'ga:secondPagePath',\n 'ga:exitPagePath',\n 'ga:pageDepth',\n 'ga:daysSinceLastSession',\n 'ga:sessionCount'\n },\n 'metrics': {\n 'ga:hits',\n 'ga:totalEvents',\n 'ga:bounces',\n 'ga:sessionDuration'\n }\n})\nall_data = response_1['data'].merge(response_2['data'], on=session_id, how='left')\nall_data.rename(index=str, columns={\n session_id: 'session_id'\n}, inplace=True)", "_____no_output_____" ], [ "all_data.head(2)", "_____no_output_____" ] ], [ [ "### Get hit-level data (using hit_id custom dimension)", "_____no_output_____" ] ], [ [ "hit_id = 'dimension' + HIT_ID_CD_INDEX\n\none_day = date(2019,7,1)\nc.set_dateranges(one_day, one_day)\n\nhits_response_1 = c.get_all_data({\n 'dimensions': {\n 'ga:' + session_id,\n 'ga:' + hit_id,\n 'ga:pagePath',\n 'ga:previousPagePath',\n 'ga:dateHourMinute'\n },\n 'metrics': {\n 'ga:hits',\n 'ga:totalEvents',\n 'ga:pageviews'\n }\n})\n\nhits_response_2 = c.get_all_data({\n 'dimensions': {\n 'ga:' + session_id,\n 'ga:' + hit_id,\n 'ga:eventCategory',\n 'ga:eventAction',\n 'ga:eventLabel'\n },\n 'metrics': {\n 'ga:totalEvents'\n }\n})\nall_hits_data = hits_response_1['data'].merge(hits_response_2['data'], \n on=[session_id, hit_id], \n how='left')\nall_hits_data.rename(index=str, columns={\n session_id: 'session_id',\n hit_id: 'hit_id'\n}, inplace=True)\nall_hits_data.head(2)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ] ]

ec7bccf6fa179f11a60f189a3202fa246b91b860

892,195

ipynb

Jupyter Notebook

Video game sales/vgsales.ipynb

TheBrownViking20/DSstuff

d915f48e88c22baa81cf8f6114615c6d1bd3faa9

[ "MIT" ]

3

2018-02-12T14:18:50.000Z

2018-05-31T19:06:54.000Z

Video game sales/vgsales.ipynb

TheBrownViking20/DSstuff

d915f48e88c22baa81cf8f6114615c6d1bd3faa9

[ "MIT" ]

null

Video game sales/vgsales.ipynb

TheBrownViking20/DSstuff

d915f48e88c22baa81cf8f6114615c6d1bd3faa9

[ "MIT" ]

1

2018-05-31T19:06:56.000Z

356.307907

261,972

0.911151

[ [ [ "# Importing the libraries and data", "_____no_output_____" ] ], [ [ "import numpy as np\nimport matplotlib.pyplot as plt\nimport pandas as pd\n%matplotlib inline \nimport seaborn as sns\nsns.set_style('dark')\n\ndataset = pd.read_csv('vgsales.csv')", "_____no_output_____" ], [ "dataset.head()", "_____no_output_____" ], [ "dataset.info()", "<class 'pandas.core.frame.DataFrame'>\nRangeIndex: 16598 entries, 0 to 16597\nData columns (total 11 columns):\nRank 16598 non-null int64\nName 16598 non-null object\nPlatform 16598 non-null object\nYear 16327 non-null float64\nGenre 16598 non-null object\nPublisher 16540 non-null object\nNA_Sales 16598 non-null float64\nEU_Sales 16598 non-null float64\nJP_Sales 16598 non-null float64\nOther_Sales 16598 non-null float64\nGlobal_Sales 16598 non-null float64\ndtypes: float64(6), int64(1), object(4)\nmemory usage: 1.4+ MB\n" ], [ "dataset.describe()", "_____no_output_____" ] ], [ [ "# Dropping 'Rank' column because we have an index thanks to pd.read_csv", "_____no_output_____" ] ], [ [ "dataset = dataset.drop(['Rank'],axis=1)\ndataset.head()", "_____no_output_____" ] ], [ [ "# frequency of platforms", "_____no_output_____" ] ], [ [ "dataset['Platform'].value_counts()", "_____no_output_____" ] ], [ [ "# frequency of genres", "_____no_output_____" ] ], [ [ "dataset['Genre'].value_counts()", "_____no_output_____" ] ], [ [ "# frequency of publishers (There are too many of them)", "_____no_output_____" ] ], [ [ "dataset['Publisher'].value_counts()", "_____no_output_____" ] ], [ [ "# Mean of sales in North America, Europe, Japan, Others and the entire world", "_____no_output_____" ] ], [ [ "li = ['NA_Sales','EU_Sales','JP_Sales','Other_Sales','Global_Sales']\n\nfor i in li:\n print(i,end=\" -->> \")\n print(dataset[i].mean(),end=\" million\")\n print(\"\\n\")", "NA_Sales -->> 0.26466742981084057 million\n\nEU_Sales -->> 0.1466520062658483 million\n\nJP_Sales -->> 0.07778166044101108 million\n\nOther_Sales -->> 0.048063019640913515 million\n\nGlobal_Sales -->> 0.53744065550074 million\n\n" ] ], [ [ "# For better analysis and visualisation, it would be better to get rid of games with publishers who have published less than 80 games and platforms with less than 90 games. Publishers with less than 80 games will be renamed with 'Other'. Same will be done for platforms", "_____no_output_____" ] ], [ [ "for i in dataset['Publisher'].unique():\n if dataset['Publisher'][dataset['Publisher'] == i].count() < 50:\n dataset['Publisher'][dataset['Publisher'] == i] = 'Other'\n \nfor i in dataset['Platform'].unique():\n if dataset['Platform'][dataset['Platform'] == i].count() < 90:\n dataset['Platform'][dataset['Platform'] == i] = 'Other'", "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:3: SettingWithCopyWarning: \nA value is trying to be set on a copy of a slice from a DataFrame\n\nSee the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n This is separate from the ipykernel package so we can avoid doing imports until\nC:\\ProgramData\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:7: SettingWithCopyWarning: \nA value is trying to be set on a copy of a slice from a DataFrame\n\nSee the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n import sys\n" ] ], [ [ "# Now let's see the change", "_____no_output_____" ] ], [ [ "dataset['Publisher'].value_counts()", "_____no_output_____" ], [ "dataset['Platform'].value_counts()", "_____no_output_____" ] ], [ [ "# Now we plot them", "_____no_output_____" ] ], [ [ "dataset['Publisher'].value_counts(sort=False).plot(kind='barh',figsize=(10,10))", "_____no_output_____" ], [ "dataset['Platform'].value_counts(sort=False).plot(kind='barh',figsize=(10,10),color='green')", "_____no_output_____" ], [ "sns.factorplot('Genre',data=dataset,kind='count',size=8)", "_____no_output_____" ] ], [ [ "# Line plot of frequencies of genres for different platforms", "_____no_output_____" ] ], [ [ "for i in dataset['Platform'].unique():\n dataset['Genre'][dataset['Platform'] == i].value_counts().plot(kind='line',label=i, figsize=(20,10), grid=True)\nplt.legend(bbox_to_anchor=(0., 1.02, 1., .102), loc=3, ncol=20, borderaxespad=0.)\nplt.xticks(np.arange(12), tuple(dataset['Genre'].unique()))\nplt.tight_layout()\nplt.show() ", "_____no_output_____" ], [ "dataset.index.name = 'Index'", "_____no_output_____" ] ], [ [ "# DataFrame of games having more than 1 million copies sold in North America grouped by platforms", "_____no_output_____" ] ], [ [ "platform_name = []\nplatform_frequency = []\nfor i in dataset['Platform'].unique():\n platform_name.append(i)\n platform_frequency.append(dataset['Name'][dataset['NA_Sales'] > 1.0][dataset['Platform'] == i].count())\nplat = pd.DataFrame()\nplat[\"Name\"] = platform_name\nplat[\"Frequency\"] = platform_frequency\nplat", "_____no_output_____" ] ], [ [ "# Now the above DataFrame is plotted as a pie chart", "_____no_output_____" ] ], [ [ "plat.plot(kind='pie',y='Frequency',labels=plat['Name'],legend=False,figsize=(16,16))", "_____no_output_____" ] ], [ [ "# Scatter plots of different sales columns", "_____no_output_____" ] ], [ [ "ax = dataset.plot(kind='scatter',x='NA_Sales',y='EU_Sales',color='red',label='EU_Sales',figsize=(16,16))\ndataset.plot(kind='scatter',x='NA_Sales',y='JP_Sales',ax=ax,color='blue',label='JP_Sales')\ndataset.plot(kind='scatter',x='NA_Sales',y='Other_Sales',ax=ax,color='green',label='Other_Sales')\nplt.xlabel('North America Sales')\nplt.ylabel('Other Markets')\nplt.show()", "_____no_output_____" ], [ "ax = dataset.plot(kind='scatter',x='EU_Sales',y='NA_Sales',color='red',label='NA_Sales',figsize=(16,16))\ndataset.plot(kind='scatter',x='EU_Sales',y='JP_Sales',ax=ax,color='blue',label='JP_Sales')\ndataset.plot(kind='scatter',x='EU_Sales',y='Other_Sales',ax=ax,color='green',label='Other_Sales')\nplt.xlabel('Europe Sales')\nplt.ylabel('Other Markets')\nplt.show()", "_____no_output_____" ], [ "ax = dataset.plot(kind='scatter',x='JP_Sales',y='EU_Sales',color='red',label='EU_Sales',figsize=(16,16))\ndataset.plot(kind='scatter',x='JP_Sales',y='NA_Sales',ax=ax,color='blue',label='NA_Sales')\ndataset.plot(kind='scatter',x='JP_Sales',y='Other_Sales',ax=ax,color='green',label='Other_Sales')\nplt.xlabel('Japan Sales')\nplt.ylabel('Other Markets')\nplt.show()", "_____no_output_____" ] ], [ [ "# Correlation coefficients of different sales columns", "_____no_output_____" ] ], [ [ "corr_1 = []\ncorr_2 = []\ncorr_res = []\nsales_list = ['NA_Sales','EU_Sales','JP_Sales','Other_Sales','Global_Sales']\nfor i in sales_list:\n for j in sales_list:\n corr_1.append(i)\n corr_2.append(j)\n corr_res.append(dataset[i].corr(dataset[j]))\ncorr_data = pd.DataFrame(\n {'Corr_1': corr_1,\n 'Corr_2': corr_2,\n 'Correlation': corr_res\n })\ncorr_data", "_____no_output_____" ] ], [ [ "# Pivoting the correlation table for good", "_____no_output_____" ] ], [ [ "corr_data = corr_data.pivot(values='Correlation',index='Corr_1',columns='Corr_2')", "_____no_output_____" ], [ "corr_data", "_____no_output_____" ] ], [ [ "# Heatmap for the pivot table", "_____no_output_____" ] ], [ [ "sns.heatmap(corr_data)", "_____no_output_____" ] ], [ [ "# Jointplot for Global Sales and North America Sales", "_____no_output_____" ] ], [ [ "sns.jointplot(x=\"NA_Sales\", y=\"Global_Sales\", data=dataset, kind='reg')", "_____no_output_____" ] ], [ [ "# Now, we have linear regression to predict Global Sales using sales in North America Sales", "_____no_output_____" ], [ "# Preparing Data", "_____no_output_____" ] ], [ [ "X = dataset.iloc[:, 5].values\ny = dataset.iloc[:, 9].values", "_____no_output_____" ] ], [ [ "# Splitting the dataset into the Training set and Test set", "_____no_output_____" ] ], [ [ "from sklearn.cross_validation import train_test_split\nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 1/10, random_state = 0)", "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\cross_validation.py:44: DeprecationWarning: This module was deprecated in version 0.18 in favor of the model_selection module into which all the refactored classes and functions are moved. Also note that the interface of the new CV iterators are different from that of this module. This module will be removed in 0.20.\n \"This module will be removed in 0.20.\", DeprecationWarning)\n" ], [ "X_train = X_train.reshape((14938,1))\ny_train = y_train.reshape((14938,1))\nX_test = X_test.reshape((1660,1))\ny_test = y_test.reshape((1660,1))", "_____no_output_____" ] ], [ [ "# Fitting Simple Linear Regression into the Training set", "_____no_output_____" ] ], [ [ "from sklearn.linear_model import LinearRegression\nregressor = LinearRegression()\nregressor.fit(X_test, y_test)", "_____no_output_____" ] ], [ [ "# Predicting the Test set results", "_____no_output_____" ] ], [ [ "y_pred = regressor.predict(X_test)", "_____no_output_____" ], [ "# Visualising the Training set results\nplt.scatter(X_train, y_train,color='red')\nplt.plot(X_train, regressor.predict(X_train), color='blue')\nplt.title('(Training set)')\nplt.xlabel('North America Sales')\nplt.ylabel('Global Sales')\nplt.show()", "_____no_output_____" ], [ "# Visualising the Test set results\nplt.scatter(X_test, y_test,color='red')\nplt.plot(X_train, regressor.predict(X_train), color='blue')\nplt.title('(Test set)')\nplt.xlabel('North America Sales')\nplt.ylabel('Global Sales')\nplt.show()", "_____no_output_____" ] ], [ [ "# Checking Score", "_____no_output_____" ] ], [ [ "print(\"Training set score: {:.2f}\".format(regressor.score(X_train,y_train)))\nprint(\"Test set score: {:.2f}\".format(regressor.score(X_test,y_test)))", "Training set score: 0.88\nTest set score: 0.90\n" ] ], [ [ "# It is only justified to say that the score is beautiful for this regressor", "_____no_output_____" ], [ "\n\n# Decision Tree Regression", "_____no_output_____" ], [ "\n\n# Fitting the Decision Tree Regression to the dataset", "_____no_output_____" ] ], [ [ "X = X.reshape((16598,1))\ny = y.reshape((16598,1))\n\nfrom sklearn.tree import DecisionTreeRegressor\nDregressor = DecisionTreeRegressor(random_state=0)\nDregressor.fit(X,y)", "_____no_output_____" ] ], [ [ "# Predicting a new result", "_____no_output_____" ] ], [ [ "y_pred = Dregressor.predict(6.5)\ny_pred", "_____no_output_____" ] ], [ [ "# Visualising the Regression results", "_____no_output_____" ] ], [ [ "plt.scatter(X, y, color = 'red')\nplt.plot(X, Dregressor.predict(X), color = 'blue')\nplt.title('(Decision Tree Regression)')\nplt.xlabel('North America Sales')\nplt.ylabel('Global Sales')\nplt.show()", "_____no_output_____" ] ], [ [ "# Visualising the Regression results (for higher resolution and smoother curve)", "_____no_output_____" ] ], [ [ "X_grid = np.arange(min(X), max(X), 0.1)\nX_grid = X_grid.reshape((len(X_grid), 1))\nplt.scatter(X, y, color = 'red')\nplt.plot(X_grid, Dregressor.predict(X_grid), color = 'blue')\nplt.title('(Decision Tree Regression)')\nplt.xlabel('North America Sales')\nplt.ylabel('Global Sales')\nplt.show()", "_____no_output_____" ] ], [ [ "# Checking Score", "_____no_output_____" ] ], [ [ "print(\"Decision tree score: {:.2f}\".format(Dregressor.score(X,y)))", "Decision tree score: 0.95\n" ] ], [ [ "# This one gives better results", "_____no_output_____" ], [ "\n\n# Random Forest Regression", "_____no_output_____" ], [ "\n\n# Fitting Random Forest Regression to the dataset", "_____no_output_____" ] ], [ [ "from sklearn.ensemble import RandomForestRegressor\nRregressor = RandomForestRegressor(n_estimators=300, random_state=0)\nRregressor.fit(X,y)", "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:3: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n This is separate from the ipykernel package so we can avoid doing imports until\n" ] ], [ [ "# Predicting a new result", "_____no_output_____" ] ], [ [ "y_pred = Rregressor.predict(6.5)\ny_pred", "_____no_output_____" ] ], [ [ "# Visualising the Random Forest Regression results (for higher resolution and smoother curve)", "_____no_output_____" ] ], [ [ "X_grid = np.arange(min(X), max(X), 0.01)\nX_grid = X_grid.reshape((len(X_grid), 1))\nplt.scatter(X, y, color = 'red')\nplt.plot(X_grid, Rregressor.predict(X_grid), color = 'blue')\nplt.title('(Random Forest Regression)')\nplt.xlabel('North America Sales')\nplt.ylabel('Global Sales')\nplt.show()", "_____no_output_____" ] ], [ [ "# Checking Score", "_____no_output_____" ] ], [ [ "print(\"Random Forest score: {:.2f}\".format(Rregressor.score(X,y)))", "Random Forest score: 0.93\n" ] ], [ [ "# Support Vector Regression", "_____no_output_____" ], [ "\n\n# Fitting the SVR to the dataset", "_____no_output_____" ] ], [ [ "from sklearn.svm import SVR\nSVRregressor = SVR(kernel = 'rbf')\nSVRregressor.fit(X,y)", "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py:526: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n y = column_or_1d(y, warn=True)\n" ] ], [ [ "# Feature Scaling", "_____no_output_____" ] ], [ [ "from sklearn.preprocessing import StandardScaler\nsc_X = StandardScaler()\nsc_y = StandardScaler()\nX = sc_X.fit_transform(X)\ny = sc_y.fit_transform(y)", "_____no_output_____" ] ], [ [ "# Predicting a new result", "_____no_output_____" ] ], [ [ "y_pred = sc_y.inverse_transform(SVRregressor.predict(sc_X.transform(np.array([[6.5]]))))\ny_pred", "_____no_output_____" ] ], [ [ "# Visualising the SVR results", "_____no_output_____" ] ], [ [ "plt.scatter(X, y, color = 'red')\nplt.plot(X, SVRregressor.predict(X), color = 'blue')\nplt.title('(SVR)')\nplt.xlabel('North America Sales')\nplt.ylabel('Global Sales')\nplt.show()", "_____no_output_____" ] ], [ [ "# Visualising the Regression results (for higher resolution and smoother curve)", "_____no_output_____" ] ], [ [ "X_grid = np.arange(min(X), max(X), 0.1)\nX_grid = X_grid.reshape((len(X_grid), 1))\nplt.scatter(X, y, color = 'red')\nplt.plot(X_grid, SVRregressor.predict(X_grid), color = 'blue')\nplt.title('(SVR)')\nplt.xlabel('North America Sales')\nplt.ylabel('Global Sales')\nplt.show()", "_____no_output_____" ] ], [ [ "# Checking Score", "_____no_output_____" ] ], [ [ "print(\"SVR score: {:.2f}\".format(SVRregressor.score(X,y)))", "SVR score: 0.31\n" ] ], [ [ "# So, SVR is probably not a good choice for this data but linear regression, decision tree regression and random forest regression are good choices especially Decision Tree regression.", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ] ]

ec7bceedca7fe27e073b612577cd252e93539179

19,946

ipynb

Jupyter Notebook

2016-labtech-julia/julia-for-pythonistas.ipynb

kbarbary/talks

632296ca81d31f1186c8f39b5ea76e04f09d9434

[ "MIT" ]

2

2019-07-22T08:44:30.000Z

2021-05-28T02:30:34.000Z

2016-labtech-julia/julia-for-pythonistas.ipynb

kbarbary/talks

632296ca81d31f1186c8f39b5ea76e04f09d9434

[ "MIT" ]

null

2016-labtech-julia/julia-for-pythonistas.ipynb

kbarbary/talks

632296ca81d31f1186c8f39b5ea76e04f09d9434

[ "MIT" ]

1

2021-09-29T08:20:55.000Z

20.229209

304

0.482854

[ [ [ "empty" ] ] ]

[ "empty" ]

[ [ "empty" ] ]

ec7bd549ec677d4ebc3b5d2c6dc16c1484ebb389

127,029

ipynb

Jupyter Notebook

demo.ipynb

zzeng13/DISC

7542fcd67dfbea448c0cfa76ac34353ae6ab171d

[ "MIT", "Unlicense" ]

1

2022-02-24T10:51:52.000Z

demo.ipynb

zzeng13/DISC

7542fcd67dfbea448c0cfa76ac34353ae6ab171d

[ "MIT", "Unlicense" ]

null

demo.ipynb

zzeng13/DISC

7542fcd67dfbea448c0cfa76ac34353ae6ab171d

[ "MIT", "Unlicense" ]

null

184.90393

3,297

0.67424

[ [ [ "# Demo\nThis demo shows how to use DISC model with your own sentence input. \n\nThe script will first load the model and data processing module. \nThen, it will run inference on the input sentence and output the detected idiom from the sentence. \n", "_____no_output_____" ] ], [ [ "from IPython.display import display, HTML\nimport torch\nimport numpy as np\nfrom tqdm import tqdm\nfrom src.utils.model_util import load_model_from_checkpoint\nfrom src.model.read_comp_triflow import ReadingComprehensionDetector as DetectorMdl\nfrom config import Config as config\nfrom demo_helper.data_processor import DataHandler\nfrom demo_helper.visualize_helper import simple_scoring_viz", "_____no_output_____" ] ], [ [ "## 1. Load model", "_____no_output_____" ] ], [ [ "data_handler = DataHandler(config)\ndetector_model= load_model_from_checkpoint(DetectorMdl, data_handler.config)\n", "Loading Pre-trained Glove Embeddings...\n" ] ], [ [ "## 2. Set and prepare input sentences", "_____no_output_____" ] ], [ [ "sentences = [\n # The following examples are idioms\n 'If you’re head over heels, you’re completely in love.',\n 'If you keep someone at arm’s length, you avoid becoming friendly with them.',\n 'If you’re a chip off the old block, you’re similar in some distinct way to your father or mother.',\n 'He must face the music for his transgression.',\n 'Getting fired turned out to be a blessing in disguise.',\n 'I’m sorry but I just can’t seem to wrap my head around it.',\n 'At the end of the day, it is you who will take the heat.',\n 'At the end of the day, it is you who will take the responsibility.',\n 'I don’t want to be Hayley’s friend anymore, she stabbed me in the back!',\n 'Why not go to the post office on your way to the mall and kill two birds with one stone?',\n 'Hey, I’m feeling pretty angry right now. I’m going to go blow off some steam.',\n 'As a rule of thumb, you should usually pay for your date’s dinner, too.',\n 'If you burn the candle at both ends, you work excessively hard, say, by keeping two jobs or by leading a busy social life in the evening.',\n # The following examples are similes\n 'You were as brave as a lion.',\n 'This house is as clean as a whistle.',\n \"Sometimes you feel like a nut, sometimes you don't.\",\n # Negative examples (no idioms)\n \"We will also see which library is recommended to use on each occasion and the unique capabilities of each library.\"\n]", "_____no_output_____" ], [ "data = data_handler.prepare_input(sentences)", "_____no_output_____" ] ], [ [ "## 3. Model inference", "_____no_output_____" ] ], [ [ "with torch.no_grad():\n ys_ = detector_model(data)\n probs = torch.nn.functional.softmax(ys_, dim=-1)\nys_ = ys_.cpu().detach().numpy()\nprobs = probs.cpu().detach().numpy()\nidiom_class_probs = probs[:, :, -1].tolist()\npredicts = np.argmax(ys_, axis=2)", "/home/zzeng/miniconda3/envs/pytorch_py37/lib/python3.7/site-packages/torch/nn/functional.py:1639: UserWarning: nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.\n warnings.warn(\"nn.functional.sigmoid is deprecated. Use torch.sigmoid instead.\")\n" ] ], [ [ "## 4. Extract output", "_____no_output_____" ] ], [ [ "ys_.shape", "_____no_output_____" ], [ "sentences_tkns = data['xs_bert'].cpu().detach().numpy().tolist()\nsentences_tkns = [data_handler.tokenizer.convert_ids_to_tokens(s) for s in sentences_tkns]", "_____no_output_____" ], [ "print('Visualize Results by Scoring: ')\nfor i in range(len(sentences_tkns)):\n s = simple_scoring_viz(sentences_tkns[i], idiom_class_probs[i], 'YlGn')\n display(HTML(s))", "Visualize Results by Scoring: \n" ], [ "predicts = predicts == 4\npredicts = predicts.astype(float)\nprint('Visualize Results by Classification: ')\nfor i in range(len(sentences_tkns)):\n s = simple_scoring_viz(sentences_tkns[i], predicts[i], 'YlGn')\n display(HTML(s))", "Visualize Results by Classification: \n" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code" ] ]

ec7bd6cc172cb7b002ad52003340bd600edb98f4

3,069

ipynb

Jupyter Notebook

ExploratoryAnalisis.ipynb

ojopiojo/TwoEchellon

ca8dc66873e53b8758fc9dafb7aee517440c2851

[ "MIT" ]

null

ExploratoryAnalisis.ipynb

ojopiojo/TwoEchellon

ca8dc66873e53b8758fc9dafb7aee517440c2851

[ "MIT" ]

null

ExploratoryAnalisis.ipynb

ojopiojo/TwoEchellon

ca8dc66873e53b8758fc9dafb7aee517440c2851

[ "MIT" ]

null

43.225352

806

0.610297

[ [ [ "\nfrom scripts.milp import *\nimport os\n\ndatadir = os.path.join(os.path.pardir, 'data')\nsoldir = os.path.join(os.path.pardir, 'solutions')\nsolfile = \"solution milp v147-city-n15-f2-d1-s4-c8-p1-v1.xlsx\"\ndatafile = 'v147-city-n15-f2-d1-s4-c8-p1-v1.xlsx'\n#plotdir = os.path.join(os.path.pardir,'plots')\n\nExecuteFromInitial(datadir, datafile, solfile, soldir)", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code" ] ]

ec7bfc13e600602b0db23d83653e41d6dc0449e7

4,361

ipynb

Jupyter Notebook

mne/viz/_brain/tests/test.ipynb

kimcoco/mne-python

df227e7a9f67f61cf1322686308a78627d2289f4

[ "BSD-3-Clause" ]

1

2021-03-13T04:41:45.000Z

mne/viz/_brain/tests/test.ipynb

kimcoco/mne-python

df227e7a9f67f61cf1322686308a78627d2289f4

[ "BSD-3-Clause" ]

12

2020-07-23T15:41:38.000Z

2021-02-24T09:38:41.000Z

mne/viz/_brain/tests/test.ipynb

kimcoco/mne-python

df227e7a9f67f61cf1322686308a78627d2289f4

[ "BSD-3-Clause" ]

null

35.455285

100

0.571887

[ [ [ "import mne\nfrom mne.datasets import testing\ndata_path = testing.data_path()\nraw_fname = data_path + '/MEG/sample/sample_audvis_trunc_raw.fif'\nsubjects_dir = data_path + '/subjects'\nsubject = 'sample'\ntrans = data_path + '/MEG/sample/sample_audvis_trunc-trans.fif'\ninfo = mne.io.read_info(raw_fname)\nmne.viz.set_3d_backend('notebook')\nfig = mne.viz.plot_alignment(info, trans, subject=subject, dig=True,\n meg=['helmet', 'sensors'], subjects_dir=subjects_dir,\n surfaces=['head-dense'])\nassert fig.display is not None", "_____no_output_____" ], [ "from contextlib import contextmanager\nimport os\nfrom numpy.testing import assert_allclose\nfrom ipywidgets import Button\nimport matplotlib.pyplot as plt\nimport mne\nfrom mne.datasets import testing\ndata_path = testing.data_path()\nsample_dir = os.path.join(data_path, 'MEG', 'sample')\nsubjects_dir = os.path.join(data_path, 'subjects')\nfname_stc = os.path.join(sample_dir, 'sample_audvis_trunc-meg')\nstc = mne.read_source_estimate(fname_stc, subject='sample')\ninitial_time = 0.13\nmne.viz.set_3d_backend('notebook')\nbrain_class = mne.viz.get_brain_class()\n\n\n@contextmanager\ndef interactive(on):\n old = plt.isinteractive()\n plt.interactive(on)\n try:\n yield\n finally:\n plt.interactive(old)\n\nwith interactive(False):\n brain = stc.plot(subjects_dir=subjects_dir, initial_time=initial_time,\n clim=dict(kind='value', pos_lims=[3, 6, 9]),\n time_viewer=True,\n show_traces=True,\n hemi='lh', size=300)\n assert isinstance(brain, brain_class)\n assert brain.notebook\n assert brain._renderer.figure.display is not None\n brain._update()\n total_number_of_buttons = len([k for k in brain.actions.keys() if '_field' not in k])\n number_of_buttons = 0\n for action in brain.actions.values():\n if isinstance(action, Button):\n action.click()\n number_of_buttons += 1\n assert number_of_buttons == total_number_of_buttons\n img_nv = brain.screenshot()\n assert img_nv.shape == (300, 300, 3), img_nv.shape\n img_v = brain.screenshot(time_viewer=True)\n assert img_v.shape[1:] == (300, 3), img_v.shape\n # XXX This rtol is not very good, ideally would be zero\n assert_allclose(img_v.shape[0], img_nv.shape[0] * 1.25, err_msg=img_nv.shape, rtol=0.1)\n brain.close()", "_____no_output_____" ], [ "import mne\nmne.viz.set_3d_backend('notebook')\nrend = mne.viz.create_3d_figure(size=(100, 100), scene=False)\nfig = rend.scene()\nmne.viz.set_3d_title(fig, 'Notebook testing')\nmne.viz.set_3d_view(fig, 200, 70, focalpoint=[0, 0, 0])\nassert fig.display is None\nrend.show()\ntotal_number_of_buttons = len([k for k in rend.actions.keys() if '_field' not in k])\nnumber_of_buttons = 0\nfor action in rend.actions.values():\n if isinstance(action, Button):\n action.click()\n number_of_buttons += 1\nassert number_of_buttons == total_number_of_buttons\nassert fig.display is not None", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code" ] ]

ec7bfdd51c712386ebd3e1500077e5e01ba99bf6

51,585

ipynb

Jupyter Notebook

image-classification/dlnd_image_classification.ipynb

pieapple/deep-learning

08a7a0645fc6894e89620fdf72477c7f079c6f4f

[ "MIT" ]

null

image-classification/dlnd_image_classification.ipynb

pieapple/deep-learning

08a7a0645fc6894e89620fdf72477c7f079c6f4f

[ "MIT" ]

null

image-classification/dlnd_image_classification.ipynb

pieapple/deep-learning

08a7a0645fc6894e89620fdf72477c7f079c6f4f

[ "MIT" ]

null

60.052386

1,217

0.644684

[ [ [ "# Image Classification\nIn this project, you'll classify images from the [CIFAR-10 dataset](https://www.cs.toronto.edu/~kriz/cifar.html). The dataset consists of airplanes, dogs, cats, and other objects. You'll preprocess the images, then train a convolutional neural network on all the samples. The images need to be normalized and the labels need to be one-hot encoded. You'll get to apply what you learned and build a convolutional, max pooling, dropout, and fully connected layers. At the end, you'll get to see your neural network's predictions on the sample images.\n## Get the Data\nRun the following cell to download the [CIFAR-10 dataset for python](https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz).", "_____no_output_____" ] ], [ [ "\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n\"\"\"\nfrom urllib.request import urlretrieve\nfrom os.path import isfile, isdir\nfrom tqdm import tqdm\nimport problem_unittests as tests\nimport tarfile\n\ncifar10_dataset_folder_path = 'cifar-10-batches-py'\n\n# Use Floyd's cifar-10 dataset if present\nfloyd_cifar10_location = '/cifar/cifar-10-python.tar.gz'\nif isfile(floyd_cifar10_location):\n tar_gz_path = floyd_cifar10_location\nelse:\n tar_gz_path = 'cifar-10-python.tar.gz'\n\nclass DLProgress(tqdm):\n last_block = 0\n\n def hook(self, block_num=1, block_size=1, total_size=None):\n self.total = total_size\n self.update((block_num - self.last_block) * block_size)\n self.last_block = block_num\n\nif not isfile(tar_gz_path):\n with DLProgress(unit='B', unit_scale=True, miniters=1, desc='CIFAR-10 Dataset') as pbar:\n urlretrieve(\n 'https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz',\n tar_gz_path,\n pbar.hook)\n\nif not isdir(cifar10_dataset_folder_path):\n with tarfile.open(tar_gz_path) as tar:\n tar.extractall()\n tar.close()\n\n\ntests.test_folder_path(cifar10_dataset_folder_path)", "/Users/zhiholiu/.pyenv/versions/3.6.6/lib/python3.6/importlib/_bootstrap.py:219: RuntimeWarning: compiletime version 3.5 of module 'tensorflow.python.framework.fast_tensor_util' does not match runtime version 3.6\n return f(*args, **kwds)\nCIFAR-10 Dataset: 0.00B [00:00, ?B/s]\n" ] ], [ [ "## Explore the Data\nThe dataset is broken into batches to prevent your machine from running out of memory. The CIFAR-10 dataset consists of 5 batches, named `data_batch_1`, `data_batch_2`, etc.. Each batch contains the labels and images that are one of the following:\n* airplane\n* automobile\n* bird\n* cat\n* deer\n* dog\n* frog\n* horse\n* ship\n* truck\n\nUnderstanding a dataset is part of making predictions on the data. Play around with the code cell below by changing the `batch_id` and `sample_id`. The `batch_id` is the id for a batch (1-5). The `sample_id` is the id for a image and label pair in the batch.\n\nAsk yourself \"What are all possible labels?\", \"What is the range of values for the image data?\", \"Are the labels in order or random?\". Answers to questions like these will help you preprocess the data and end up with better predictions.", "_____no_output_____" ] ], [ [ "%matplotlib inline\n%config InlineBackend.figure_format = 'retina'\n\nimport helper\nimport numpy as np\n\n# Explore the dataset\nbatch_id = 1\nsample_id = 5\nhelper.display_stats(cifar10_dataset_folder_path, batch_id, sample_id)", "_____no_output_____" ] ], [ [ "## Implement Preprocess Functions\n### Normalize\nIn the cell below, implement the `normalize` function to take in image data, `x`, and return it as a normalized Numpy array. The values should be in the range of 0 to 1, inclusive. The return object should be the same shape as `x`.", "_____no_output_____" ] ], [ [ "def normalize(x):\n \"\"\"\n Normalize a list of sample image data in the range of 0 to 1\n : x: List of image data. The image shape is (32, 32, 3)\n : return: Numpy array of normalize data\n \"\"\"\n # TODO: Implement Function\n return None\n\n\n\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n\"\"\"\ntests.test_normalize(normalize)", "_____no_output_____" ] ], [ [ "### One-hot encode\nJust like the previous code cell, you'll be implementing a function for preprocessing. This time, you'll implement the `one_hot_encode` function. The input, `x`, are a list of labels. Implement the function to return the list of labels as One-Hot encoded Numpy array. The possible values for labels are 0 to 9. The one-hot encoding function should return the same encoding for each value between each call to `one_hot_encode`. Make sure to save the map of encodings outside the function.\n\nHint: Don't reinvent the wheel.", "_____no_output_____" ] ], [ [ "def one_hot_encode(x):\n \"\"\"\n One hot encode a list of sample labels. Return a one-hot encoded vector for each label.\n : x: List of sample Labels\n : return: Numpy array of one-hot encoded labels\n \"\"\"\n # TODO: Implement Function\n return None\n\n\n\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n\"\"\"\ntests.test_one_hot_encode(one_hot_encode)", "_____no_output_____" ] ], [ [ "### Randomize Data\nAs you saw from exploring the data above, the order of the samples are randomized. It doesn't hurt to randomize it again, but you don't need to for this dataset.", "_____no_output_____" ], [ "## Preprocess all the data and save it\nRunning the code cell below will preprocess all the CIFAR-10 data and save it to file. The code below also uses 10% of the training data for validation.", "_____no_output_____" ] ], [ [ "\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL\n\"\"\"\n# Preprocess Training, Validation, and Testing Data\nhelper.preprocess_and_save_data(cifar10_dataset_folder_path, normalize, one_hot_encode)", "_____no_output_____" ] ], [ [ "# Check Point\nThis is your first checkpoint. If you ever decide to come back to this notebook or have to restart the notebook, you can start from here. The preprocessed data has been saved to disk.", "_____no_output_____" ] ], [ [ "\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL\n\"\"\"\nimport pickle\nimport problem_unittests as tests\nimport helper\n\n# Load the Preprocessed Validation data\nvalid_features, valid_labels = pickle.load(open('preprocess_validation.p', mode='rb'))", "_____no_output_____" ] ], [ [ "## Build the network\nFor the neural network, you'll build each layer into a function. Most of the code you've seen has been outside of functions. To test your code more thoroughly, we require that you put each layer in a function. This allows us to give you better feedback and test for simple mistakes using our unittests before you submit your project.\n\n>**Note:** If you're finding it hard to dedicate enough time for this course each week, we've provided a small shortcut to this part of the project. In the next couple of problems, you'll have the option to use classes from the [TensorFlow Layers](https://www.tensorflow.org/api_docs/python/tf/layers) or [TensorFlow Layers (contrib)](https://www.tensorflow.org/api_guides/python/contrib.layers) packages to build each layer, except the layers you build in the \"Convolutional and Max Pooling Layer\" section. TF Layers is similar to Keras's and TFLearn's abstraction to layers, so it's easy to pickup.\n\n>However, if you would like to get the most out of this course, try to solve all the problems _without_ using anything from the TF Layers packages. You **can** still use classes from other packages that happen to have the same name as ones you find in TF Layers! For example, instead of using the TF Layers version of the `conv2d` class, [tf.layers.conv2d](https://www.tensorflow.org/api_docs/python/tf/layers/conv2d), you would want to use the TF Neural Network version of `conv2d`, [tf.nn.conv2d](https://www.tensorflow.org/api_docs/python/tf/nn/conv2d). \n\nLet's begin!\n\n### Input\nThe neural network needs to read the image data, one-hot encoded labels, and dropout keep probability. Implement the following functions\n* Implement `neural_net_image_input`\n * Return a [TF Placeholder](https://www.tensorflow.org/api_docs/python/tf/placeholder)\n * Set the shape using `image_shape` with batch size set to `None`.\n * Name the TensorFlow placeholder \"x\" using the TensorFlow `name` parameter in the [TF Placeholder](https://www.tensorflow.org/api_docs/python/tf/placeholder).\n* Implement `neural_net_label_input`\n * Return a [TF Placeholder](https://www.tensorflow.org/api_docs/python/tf/placeholder)\n * Set the shape using `n_classes` with batch size set to `None`.\n * Name the TensorFlow placeholder \"y\" using the TensorFlow `name` parameter in the [TF Placeholder](https://www.tensorflow.org/api_docs/python/tf/placeholder).\n* Implement `neural_net_keep_prob_input`\n * Return a [TF Placeholder](https://www.tensorflow.org/api_docs/python/tf/placeholder) for dropout keep probability.\n * Name the TensorFlow placeholder \"keep_prob\" using the TensorFlow `name` parameter in the [TF Placeholder](https://www.tensorflow.org/api_docs/python/tf/placeholder).\n\nThese names will be used at the end of the project to load your saved model.\n\nNote: `None` for shapes in TensorFlow allow for a dynamic size.", "_____no_output_____" ] ], [ [ "import tensorflow as tf\n\ndef neural_net_image_input(image_shape):\n \"\"\"\n Return a Tensor for a batch of image input\n : image_shape: Shape of the images\n : return: Tensor for image input.\n \"\"\"\n # TODO: Implement Function\n return None\n\n\ndef neural_net_label_input(n_classes):\n \"\"\"\n Return a Tensor for a batch of label input\n : n_classes: Number of classes\n : return: Tensor for label input.\n \"\"\"\n # TODO: Implement Function\n return None\n\n\ndef neural_net_keep_prob_input():\n \"\"\"\n Return a Tensor for keep probability\n : return: Tensor for keep probability.\n \"\"\"\n # TODO: Implement Function\n return None\n\n\n\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n\"\"\"\ntf.reset_default_graph()\ntests.test_nn_image_inputs(neural_net_image_input)\ntests.test_nn_label_inputs(neural_net_label_input)\ntests.test_nn_keep_prob_inputs(neural_net_keep_prob_input)", "_____no_output_____" ] ], [ [ "### Convolution and Max Pooling Layer\nConvolution layers have a lot of success with images. For this code cell, you should implement the function `conv2d_maxpool` to apply convolution then max pooling:\n* Create the weight and bias using `conv_ksize`, `conv_num_outputs` and the shape of `x_tensor`.\n* Apply a convolution to `x_tensor` using weight and `conv_strides`.\n * We recommend you use same padding, but you're welcome to use any padding.\n* Add bias\n* Add a nonlinear activation to the convolution.\n* Apply Max Pooling using `pool_ksize` and `pool_strides`.\n * We recommend you use same padding, but you're welcome to use any padding.\n\n**Note:** You **can't** use [TensorFlow Layers](https://www.tensorflow.org/api_docs/python/tf/layers) or [TensorFlow Layers (contrib)](https://www.tensorflow.org/api_guides/python/contrib.layers) for **this** layer, but you can still use TensorFlow's [Neural Network](https://www.tensorflow.org/api_docs/python/tf/nn) package. You may still use the shortcut option for all the **other** layers.", "_____no_output_____" ] ], [ [ "def conv2d_maxpool(x_tensor, conv_num_outputs, conv_ksize, conv_strides, pool_ksize, pool_strides):\n \"\"\"\n Apply convolution then max pooling to x_tensor\n :param x_tensor: TensorFlow Tensor\n :param conv_num_outputs: Number of outputs for the convolutional layer\n :param conv_ksize: kernal size 2-D Tuple for the convolutional layer\n :param conv_strides: Stride 2-D Tuple for convolution\n :param pool_ksize: kernal size 2-D Tuple for pool\n :param pool_strides: Stride 2-D Tuple for pool\n : return: A tensor that represents convolution and max pooling of x_tensor\n \"\"\"\n # TODO: Implement Function\n return None \n\n\n\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n\"\"\"\ntests.test_con_pool(conv2d_maxpool)", "_____no_output_____" ] ], [ [ "### Flatten Layer\nImplement the `flatten` function to change the dimension of `x_tensor` from a 4-D tensor to a 2-D tensor. The output should be the shape (*Batch Size*, *Flattened Image Size*). Shortcut option: you can use classes from the [TensorFlow Layers](https://www.tensorflow.org/api_docs/python/tf/layers) or [TensorFlow Layers (contrib)](https://www.tensorflow.org/api_guides/python/contrib.layers) packages for this layer. For more of a challenge, only use other TensorFlow packages.", "_____no_output_____" ] ], [ [ "def flatten(x_tensor):\n \"\"\"\n Flatten x_tensor to (Batch Size, Flattened Image Size)\n : x_tensor: A tensor of size (Batch Size, ...), where ... are the image dimensions.\n : return: A tensor of size (Batch Size, Flattened Image Size).\n \"\"\"\n # TODO: Implement Function\n return None\n\n\n\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n\"\"\"\ntests.test_flatten(flatten)", "_____no_output_____" ] ], [ [ "### Fully-Connected Layer\nImplement the `fully_conn` function to apply a fully connected layer to `x_tensor` with the shape (*Batch Size*, *num_outputs*). Shortcut option: you can use classes from the [TensorFlow Layers](https://www.tensorflow.org/api_docs/python/tf/layers) or [TensorFlow Layers (contrib)](https://www.tensorflow.org/api_guides/python/contrib.layers) packages for this layer. For more of a challenge, only use other TensorFlow packages.", "_____no_output_____" ] ], [ [ "def fully_conn(x_tensor, num_outputs):\n \"\"\"\n Apply a fully connected layer to x_tensor using weight and bias\n : x_tensor: A 2-D tensor where the first dimension is batch size.\n : num_outputs: The number of output that the new tensor should be.\n : return: A 2-D tensor where the second dimension is num_outputs.\n \"\"\"\n # TODO: Implement Function\n return None\n\n\n\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n\"\"\"\ntests.test_fully_conn(fully_conn)", "_____no_output_____" ] ], [ [ "### Output Layer\nImplement the `output` function to apply a fully connected layer to `x_tensor` with the shape (*Batch Size*, *num_outputs*). Shortcut option: you can use classes from the [TensorFlow Layers](https://www.tensorflow.org/api_docs/python/tf/layers) or [TensorFlow Layers (contrib)](https://www.tensorflow.org/api_guides/python/contrib.layers) packages for this layer. For more of a challenge, only use other TensorFlow packages.\n\n**Note:** Activation, softmax, or cross entropy should **not** be applied to this.", "_____no_output_____" ] ], [ [ "def output(x_tensor, num_outputs):\n \"\"\"\n Apply a output layer to x_tensor using weight and bias\n : x_tensor: A 2-D tensor where the first dimension is batch size.\n : num_outputs: The number of output that the new tensor should be.\n : return: A 2-D tensor where the second dimension is num_outputs.\n \"\"\"\n # TODO: Implement Function\n return None\n\n\n\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n\"\"\"\ntests.test_output(output)", "_____no_output_____" ] ], [ [ "### Create Convolutional Model\nImplement the function `conv_net` to create a convolutional neural network model. The function takes in a batch of images, `x`, and outputs logits. Use the layers you created above to create this model:\n\n* Apply 1, 2, or 3 Convolution and Max Pool layers\n* Apply a Flatten Layer\n* Apply 1, 2, or 3 Fully Connected Layers\n* Apply an Output Layer\n* Return the output\n* Apply [TensorFlow's Dropout](https://www.tensorflow.org/api_docs/python/tf/nn/dropout) to one or more layers in the model using `keep_prob`. ", "_____no_output_____" ] ], [ [ "def conv_net(x, keep_prob):\n \"\"\"\n Create a convolutional neural network model\n : x: Placeholder tensor that holds image data.\n : keep_prob: Placeholder tensor that hold dropout keep probability.\n : return: Tensor that represents logits\n \"\"\"\n # TODO: Apply 1, 2, or 3 Convolution and Max Pool layers\n # Play around with different number of outputs, kernel size and stride\n # Function Definition from Above:\n # conv2d_maxpool(x_tensor, conv_num_outputs, conv_ksize, conv_strides, pool_ksize, pool_strides)\n \n\n # TODO: Apply a Flatten Layer\n # Function Definition from Above:\n # flatten(x_tensor)\n \n\n # TODO: Apply 1, 2, or 3 Fully Connected Layers\n # Play around with different number of outputs\n # Function Definition from Above:\n # fully_conn(x_tensor, num_outputs)\n \n \n # TODO: Apply an Output Layer\n # Set this to the number of classes\n # Function Definition from Above:\n # output(x_tensor, num_outputs)\n \n \n # TODO: return output\n return None\n\n\n\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n\"\"\"\n\n##############################\n## Build the Neural Network ##\n##############################\n\n# Remove previous weights, bias, inputs, etc..\ntf.reset_default_graph()\n\n# Inputs\nx = neural_net_image_input((32, 32, 3))\ny = neural_net_label_input(10)\nkeep_prob = neural_net_keep_prob_input()\n\n# Model\nlogits = conv_net(x, keep_prob)\n\n# Name logits Tensor, so that is can be loaded from disk after training\nlogits = tf.identity(logits, name='logits')\n\n# Loss and Optimizer\ncost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=y))\noptimizer = tf.train.AdamOptimizer().minimize(cost)\n\n# Accuracy\ncorrect_pred = tf.equal(tf.argmax(logits, 1), tf.argmax(y, 1))\naccuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32), name='accuracy')\n\ntests.test_conv_net(conv_net)", "_____no_output_____" ] ], [ [ "## Train the Neural Network\n### Single Optimization\nImplement the function `train_neural_network` to do a single optimization. The optimization should use `optimizer` to optimize in `session` with a `feed_dict` of the following:\n* `x` for image input\n* `y` for labels\n* `keep_prob` for keep probability for dropout\n\nThis function will be called for each batch, so `tf.global_variables_initializer()` has already been called.\n\nNote: Nothing needs to be returned. This function is only optimizing the neural network.", "_____no_output_____" ] ], [ [ "def train_neural_network(session, optimizer, keep_probability, feature_batch, label_batch):\n \"\"\"\n Optimize the session on a batch of images and labels\n : session: Current TensorFlow session\n : optimizer: TensorFlow optimizer function\n : keep_probability: keep probability\n : feature_batch: Batch of Numpy image data\n : label_batch: Batch of Numpy label data\n \"\"\"\n # TODO: Implement Function\n pass\n\n\n\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE\n\"\"\"\ntests.test_train_nn(train_neural_network)", "_____no_output_____" ] ], [ [ "### Show Stats\nImplement the function `print_stats` to print loss and validation accuracy. Use the global variables `valid_features` and `valid_labels` to calculate validation accuracy. Use a keep probability of `1.0` to calculate the loss and validation accuracy.", "_____no_output_____" ] ], [ [ "def print_stats(session, feature_batch, label_batch, cost, accuracy):\n \"\"\"\n Print information about loss and validation accuracy\n : session: Current TensorFlow session\n : feature_batch: Batch of Numpy image data\n : label_batch: Batch of Numpy label data\n : cost: TensorFlow cost function\n : accuracy: TensorFlow accuracy function\n \"\"\"\n # TODO: Implement Function\n pass", "_____no_output_____" ] ], [ [ "### Hyperparameters\nTune the following parameters:\n* Set `epochs` to the number of iterations until the network stops learning or start overfitting\n* Set `batch_size` to the highest number that your machine has memory for. Most people set them to common sizes of memory:\n * 64\n * 128\n * 256\n * ...\n* Set `keep_probability` to the probability of keeping a node using dropout", "_____no_output_____" ] ], [ [ "# TODO: Tune Parameters\nepochs = None\nbatch_size = None\nkeep_probability = None", "_____no_output_____" ] ], [ [ "### Train on a Single CIFAR-10 Batch\nInstead of training the neural network on all the CIFAR-10 batches of data, let's use a single batch. This should save time while you iterate on the model to get a better accuracy. Once the final validation accuracy is 50% or greater, run the model on all the data in the next section.", "_____no_output_____" ] ], [ [ "\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL\n\"\"\"\nprint('Checking the Training on a Single Batch...')\nwith tf.Session() as sess:\n # Initializing the variables\n sess.run(tf.global_variables_initializer())\n \n # Training cycle\n for epoch in range(epochs):\n batch_i = 1\n for batch_features, batch_labels in helper.load_preprocess_training_batch(batch_i, batch_size):\n train_neural_network(sess, optimizer, keep_probability, batch_features, batch_labels)\n print('Epoch {:>2}, CIFAR-10 Batch {}: '.format(epoch + 1, batch_i), end='')\n print_stats(sess, batch_features, batch_labels, cost, accuracy)", "_____no_output_____" ] ], [ [ "### Fully Train the Model\nNow that you got a good accuracy with a single CIFAR-10 batch, try it with all five batches.", "_____no_output_____" ] ], [ [ "\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL\n\"\"\"\nsave_model_path = './image_classification'\n\nprint('Training...')\nwith tf.Session() as sess:\n # Initializing the variables\n sess.run(tf.global_variables_initializer())\n \n # Training cycle\n for epoch in range(epochs):\n # Loop over all batches\n n_batches = 5\n for batch_i in range(1, n_batches + 1):\n for batch_features, batch_labels in helper.load_preprocess_training_batch(batch_i, batch_size):\n train_neural_network(sess, optimizer, keep_probability, batch_features, batch_labels)\n print('Epoch {:>2}, CIFAR-10 Batch {}: '.format(epoch + 1, batch_i), end='')\n print_stats(sess, batch_features, batch_labels, cost, accuracy)\n \n # Save Model\n saver = tf.train.Saver()\n save_path = saver.save(sess, save_model_path)", "_____no_output_____" ] ], [ [ "# Checkpoint\nThe model has been saved to disk.\n## Test Model\nTest your model against the test dataset. This will be your final accuracy. You should have an accuracy greater than 50%. If you don't, keep tweaking the model architecture and parameters.", "_____no_output_____" ] ], [ [ "\"\"\"\nDON'T MODIFY ANYTHING IN THIS CELL\n\"\"\"\n%matplotlib inline\n%config InlineBackend.figure_format = 'retina'\n\nimport tensorflow as tf\nimport pickle\nimport helper\nimport random\n\n# Set batch size if not already set\ntry:\n if batch_size:\n pass\nexcept NameError:\n batch_size = 64\n\nsave_model_path = './image_classification'\nn_samples = 4\ntop_n_predictions = 3\n\ndef test_model():\n \"\"\"\n Test the saved model against the test dataset\n \"\"\"\n\n test_features, test_labels = pickle.load(open('preprocess_test.p', mode='rb'))\n loaded_graph = tf.Graph()\n\n with tf.Session(graph=loaded_graph) as sess:\n # Load model\n loader = tf.train.import_meta_graph(save_model_path + '.meta')\n loader.restore(sess, save_model_path)\n\n # Get Tensors from loaded model\n loaded_x = loaded_graph.get_tensor_by_name('x:0')\n loaded_y = loaded_graph.get_tensor_by_name('y:0')\n loaded_keep_prob = loaded_graph.get_tensor_by_name('keep_prob:0')\n loaded_logits = loaded_graph.get_tensor_by_name('logits:0')\n loaded_acc = loaded_graph.get_tensor_by_name('accuracy:0')\n \n # Get accuracy in batches for memory limitations\n test_batch_acc_total = 0\n test_batch_count = 0\n \n for test_feature_batch, test_label_batch in helper.batch_features_labels(test_features, test_labels, batch_size):\n test_batch_acc_total += sess.run(\n loaded_acc,\n feed_dict={loaded_x: test_feature_batch, loaded_y: test_label_batch, loaded_keep_prob: 1.0})\n test_batch_count += 1\n\n print('Testing Accuracy: {}\\n'.format(test_batch_acc_total/test_batch_count))\n\n # Print Random Samples\n random_test_features, random_test_labels = tuple(zip(*random.sample(list(zip(test_features, test_labels)), n_samples)))\n random_test_predictions = sess.run(\n tf.nn.top_k(tf.nn.softmax(loaded_logits), top_n_predictions),\n feed_dict={loaded_x: random_test_features, loaded_y: random_test_labels, loaded_keep_prob: 1.0})\n helper.display_image_predictions(random_test_features, random_test_labels, random_test_predictions)\n\n\ntest_model()", "_____no_output_____" ] ], [ [ "## Why 50-80% Accuracy?\nYou might be wondering why you can't get an accuracy any higher. First things first, 50% isn't bad for a simple CNN. Pure guessing would get you 10% accuracy. However, you might notice people are getting scores [well above 80%](http://rodrigob.github.io/are_we_there_yet/build/classification_datasets_results.html#43494641522d3130). That's because we haven't taught you all there is to know about neural networks. We still need to cover a few more techniques.\n## Submitting This Project\nWhen submitting this project, make sure to run all the cells before saving the notebook. Save the notebook file as \"dlnd_image_classification.ipynb\" and save it as a HTML file under \"File\" -> \"Download as\". Include the \"helper.py\" and \"problem_unittests.py\" files in your submission.", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ] ]

ec7c040d8310a2f30721c3fe21066371bda44e14

7,367

ipynb

Jupyter Notebook

docs/tools/engine/sparse_vector_times_sparse_vector.ipynb

seibert/mlir-graphblas

89b811a8800551e5cd3205c70b6a5431f311030b

[ "Apache-2.0", "MIT" ]

7

2021-06-04T19:42:13.000Z

2022-03-25T15:53:23.000Z

docs/tools/engine/sparse_vector_times_sparse_vector.ipynb

seibert/mlir-graphblas

89b811a8800551e5cd3205c70b6a5431f311030b

[ "Apache-2.0", "MIT" ]

204

2021-02-28T22:26:40.000Z

2022-03-29T19:29:21.000Z

docs/tools/engine/sparse_vector_times_sparse_vector.ipynb

seibert/mlir-graphblas

89b811a8800551e5cd3205c70b6a5431f311030b

[ "Apache-2.0", "MIT" ]

8

2021-01-11T17:08:18.000Z

2021-08-13T15:47:15.000Z

24.556667

213

0.495181

[ [ [ "# JIT Engine: Sparse Vector x Sparse Vector\n\nThis example will go over how to compile MLIR code for multiplying sparse vectors in an element-wise fashion. \n\nAccomplishing this task is mostly applying the knowledge from our previous tutorials on sparse tensors and dense tensors. Thus, this will be more of a demonstration or example than it will be a tutorial. \n\nLet’s first import some necessary modules and generate an instance of our JIT engine.", "_____no_output_____" ] ], [ [ "import mlir_graphblas\nimport mlir_graphblas.sparse_utils\nimport numpy as np\n\nengine = mlir_graphblas.MlirJitEngine()", "_____no_output_____" ] ], [ [ "This is the code we'll use to multiply two sparse vectors. ", "_____no_output_____" ] ], [ [ "mlir_text = \"\"\"\n#trait_mul_s = {\n indexing_maps = [\n affine_map<(i) -> (i)>,\n affine_map<(i) -> (i)>,\n affine_map<(i) -> (i)>\n ],\n sparse = [\n [ \"S\" ],\n [ \"S\" ],\n [ \"D\" ]\n ],\n iterator_types = [\"parallel\"],\n doc = \"Sparse Vector Multiply\"\n}\n\n#CV64 = #sparse_tensor.encoding<{\n dimLevelType = [ \"compressed\" ],\n pointerBitWidth = 64,\n indexBitWidth = 64\n}>\n\nfunc @sparse_vector_multiply(%arga: tensor<8xf32, #CV64>, %argb: tensor<8xf32, #CV64>) -> tensor<8xf32> {\n %output_storage = constant dense<0.0> : tensor<8xf32>\n %0 = linalg.generic #trait_mul_s\n ins(%arga, %argb: tensor<8xf32, #CV64>, tensor<8xf32, #CV64>)\n outs(%output_storage: tensor<8xf32>) {\n ^bb(%a: f32, %b: f32, %x: f32):\n %0 = mulf %a, %b : f32\n linalg.yield %0 : f32\n } -> tensor<8xf32>\n return %0 : tensor<8xf32>\n}\n\"\"\"", "_____no_output_____" ] ], [ [ "These are the passes we'll use.", "_____no_output_____" ] ], [ [ "passes = [\n \"--sparsification\",\n \"--sparse-tensor-conversion\",\n \"--linalg-bufferize\",\n \"--func-bufferize\",\n \"--tensor-bufferize\",\n \"--tensor-constant-bufferize\",\n \"--finalizing-bufferize\",\n \"--convert-linalg-to-loops\",\n \"--convert-scf-to-std\",\n \"--convert-memref-to-llvm\",\n \"--convert-std-to-llvm\",\n]", "_____no_output_____" ] ], [ [ "Let's generate our Python function.", "_____no_output_____" ] ], [ [ "engine.add(mlir_text, passes)\nsparse_vector_multiply = engine.sparse_vector_multiply", "_____no_output_____" ] ], [ [ "Let's generate our inputs.", "_____no_output_____" ] ], [ [ "# equivalent to np.array([0.0, 1.1, 2.2, 3.3, 0, 0, 0, 7.7], dtype=np.float32)\n\nindices = np.array([\n [0], \n [1], \n [2], \n [3], \n [7], \n], dtype=np.uint64) # Coordinates\nvalues = np.array([0.0, 1.1, 2.2, 3.3, 7.7], dtype=np.float32)\nsizes = np.array([8], dtype=np.uint64)\nsparsity = np.array([True], dtype=np.bool8)\n\na = mlir_graphblas.sparse_utils.MLIRSparseTensor(indices, values, sizes, sparsity)", "_____no_output_____" ], [ "# equivalent to np.array([0, 0, 0, 3.3, 4.4, 0, 0, 7.7], dtype=np.float32)\n\nindices = np.array([\n [3], \n [4],\n [7],\n], dtype=np.uint64) # Coordinates\nvalues = np.array([3.3, 4.4, 7.7], dtype=np.float32)\nsizes = np.array([8], dtype=np.uint64)\nsparsity = np.array([True], dtype=np.bool8)\n\nb = mlir_graphblas.sparse_utils.MLIRSparseTensor(indices, values, sizes, sparsity)", "_____no_output_____" ] ], [ [ "Let's grab our result.", "_____no_output_____" ] ], [ [ "answer = sparse_vector_multiply(a, b)\nanswer", "_____no_output_____" ] ], [ [ "Let's see if our results match what we would expect. ", "_____no_output_____" ] ], [ [ "a_dense = np.array([0.0, 1.1, 2.2, 3.3, 0, 0, 0, 7.7], dtype=np.float32)\nb_dense = np.array([0, 0, 0, 3.3, 4.4, 0, 0, 7.7], dtype=np.float32)\nnp_result = a_dense * b_dense", "_____no_output_____" ], [ "np_result", "_____no_output_____" ], [ "all(answer == np_result)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ] ]

ec7c0833f231686b0907299a6b48740ba99145ac

4,467

ipynb

Jupyter Notebook

Pandas/.ipynb_checkpoints/Series-checkpoint.ipynb

Pearl6193/Data-Science

6398cd6b14c77126ff0bc02e83e3ffd882a3c21a

[ "MIT" ]

null

Pandas/.ipynb_checkpoints/Series-checkpoint.ipynb

Pearl6193/Data-Science

6398cd6b14c77126ff0bc02e83e3ffd882a3c21a

[ "MIT" ]

null

Pandas/.ipynb_checkpoints/Series-checkpoint.ipynb

Pearl6193/Data-Science

6398cd6b14c77126ff0bc02e83e3ffd882a3c21a

[ "MIT" ]

null

17.313953

75

0.424894

[ [ [ "import pandas as pd\nimport numpy as np", "_____no_output_____" ], [ "labels = ['a','b','c']\nmy_list = [10,20,30]\narr = np.array(my_list)\nd = {\"a\":10,\"b\":20,\"c\":30}", "_____no_output_____" ], [ "pd.Series(my_list)", "_____no_output_____" ], [ "pd.Series(my_list,labels)", "_____no_output_____" ], [ "pd.Series(arr)", "_____no_output_____" ], [ "pd.Series(arr,labels)", "_____no_output_____" ], [ "pd.Series(d)", "_____no_output_____" ], [ "pd.Series(labels)", "_____no_output_____" ], [ "pd.Series([sum,len,print])", "_____no_output_____" ], [ "ser1 = pd.Series([1,2,3,4],[\"USA\",\"Germany\",\"Italy\",\"Japan\"])", "_____no_output_____" ], [ "ser1", "_____no_output_____" ], [ "ser2 = pd.Series([1,2,3,4],[\"USA\",\"Germany\",\"USSR\",\"Japan\"])", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7c112a8acdf94a6af561b70356c402ec4b1737

44,857

ipynb

Jupyter Notebook

snippets/xarray.ipynb

BenSchZA/snippets

16c4a23b343614bdeeb58db07e8146e5af5a3add

[ "MIT" ]

null

snippets/xarray.ipynb

BenSchZA/snippets

16c4a23b343614bdeeb58db07e8146e5af5a3add

[ "MIT" ]

null

snippets/xarray.ipynb

BenSchZA/snippets

16c4a23b343614bdeeb58db07e8146e5af5a3add

[ "MIT" ]

null

44.194089

9,213

0.527432

[ [ [ "# Appending & navigating simulations with given parameters\n\n*Danilo Lessa Bernardineli* \n\n---\n", "_____no_output_____" ], [ "## Dependences", "_____no_output_____" ] ], [ [ "import pandas as pd\nimport matplotlib.pyplot as plt\nimport numpy as np\nfrom cadCAD.configuration import Experiment\nfrom cadCAD.configuration.utils import config_sim\nfrom cadCAD.engine import ExecutionMode, ExecutionContext, Executor", "_____no_output_____" ] ], [ [ "## Definitions", "_____no_output_____" ], [ "### Initial conditions and parameters", "_____no_output_____" ] ], [ [ "initial_conditions = {\n 'prey_population': 100,\n 'predator_population': 15\n }\n\nparams = {\n \"prey_birth_rate\": [1.0],\n \"predator_birth_rate\": [0.01],\n \"predator_death_const\": [1.0],\n \"prey_death_const\": [0.03],\n \"dt\": [0.01, 0.1, 0.05] # Precision of the simulation. Lower is more accurate / slower\n}\n\nsimulation_parameters = {\n 'N': 7,\n 'T': range(200),\n 'M': params\n}", "_____no_output_____" ] ], [ [ "### Policies", "_____no_output_____" ] ], [ [ "def p_predator_births(params, step, sL, s):\n dt = params['dt']\n predator_population = s['predator_population']\n prey_population = s['prey_population']\n birth_fraction = params['predator_birth_rate'] + np.random.random() * 0.0002\n births = birth_fraction * prey_population * predator_population * dt\n return {'add_to_predator_population': births}\n\n\ndef p_prey_births(params, step, sL, s):\n dt = params['dt']\n population = s['prey_population']\n birth_fraction = params['prey_birth_rate'] + np.random.random() * 0.1\n births = birth_fraction * population * dt\n return {'add_to_prey_population': births}\n\n\ndef p_predator_deaths(params, step, sL, s):\n dt = params['dt']\n population = s['predator_population']\n death_rate = params['predator_death_const'] + np.random.random() * 0.005\n deaths = death_rate * population * dt\n return {'add_to_predator_population': -1.0 * deaths}\n\n\ndef p_prey_deaths(params, step, sL, s):\n dt = params['dt']\n death_rate = params['prey_death_const'] + np.random.random() * 0.1\n prey_population = s['prey_population']\n predator_population = s['predator_population']\n deaths = death_rate * prey_population * predator_population * dt\n return {'add_to_prey_population': -1.0 * deaths}", "_____no_output_____" ] ], [ [ "### State update functions", "_____no_output_____" ] ], [ [ "def s_prey_population(params, step, sL, s, _input):\n y = 'prey_population'\n x = s['prey_population'] + _input['add_to_prey_population']\n return (y, x)\n\n\ndef s_predator_population(params, step, sL, s, _input):\n y = 'predator_population'\n x = s['predator_population'] + _input['add_to_predator_population']\n return (y, x)", "_____no_output_____" ] ], [ [ "### State update blocks", "_____no_output_____" ] ], [ [ "partial_state_update_blocks = [\n { \n 'policies': {\n 'predator_births': p_predator_births,\n 'prey_births': p_prey_births,\n 'predator_deaths': p_predator_deaths,\n 'prey_deaths': p_prey_deaths,\n },\n 'variables': {\n 'predator_population': s_prey_population,\n 'prey_population': s_predator_population\n }\n }\n]", "_____no_output_____" ] ], [ [ "### Configuration and Execution", "_____no_output_____" ] ], [ [ "sim_config = config_sim(simulation_parameters)\n\nexp = Experiment()\nexp.append_configs(sim_configs=sim_config, \n initial_state=initial_conditions,\n partial_state_update_blocks=partial_state_update_blocks)\n\n\nfrom cadCAD import configs\nexec_mode = ExecutionMode()\nexec_context = ExecutionContext(exec_mode.local_mode)\nexecutor = Executor(exec_context=exec_context, configs=configs) \n(records, tensor_field, _) = executor.execute() ", "\n ___________ ____\n ________ __ ___/ / ____/ | / __ \\\n / ___/ __` / __ / / / /| | / / / /\n/ /__/ /_/ / /_/ / /___/ ___ |/ /_/ /\n\\___/\\__,_/\\__,_/\\____/_/ |_/_____/\nby cadCAD\n\nExecution Mode: local_proc\nConfiguration Count: 3\nDimensions of the first simulation: (Timesteps, Params, Runs, Vars) = (200, 5, 7, 2)\nExecution Method: local_simulations\nSimIDs : [0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2]\nSubsetIDs: [0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2]\nNs : [0, 1, 2, 3, 4, 5, 6, 0, 1, 2, 3, 4, 5, 6, 0, 1, 2, 3, 4, 5, 6]\nExpIDs : [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]\nExecution Mode: parallelized\nTotal execution time: 1.72s\n" ] ], [ [ "### Results", "_____no_output_____" ] ], [ [ "import plotly.express as px", "_____no_output_____" ], [ "df = pd.DataFrame(records)\n\n# Drop all intermediate substeps\nfirst_ind = (df.substep == 0) & (df.timestep == 0)\nlast_ind = df.substep == max(df.substep)\ninds_to_drop = (first_ind | last_ind)\n#df = df.loc[inds_to_drop].drop(columns=['substep'])\n\n# Attribute parameters to each row\ndf = df.assign(**configs[0].sim_config['M'])\nfor i, (_, n_df) in enumerate(df.groupby(['simulation', 'subset', 'run'])):\n df.loc[n_df.index] = n_df.assign(**configs[i].sim_config['M'])", "_____no_output_____" ], [ "import xarray as xr", "_____no_output_____" ], [ "xr_df = df.set_index([\"timestep\", \"substep\", *configs[0].sim_config['M'].keys()])", "_____no_output_____" ], [ "ds = xr.Dataset(xr_df)", "_____no_output_____" ], [ "ds.unstack()", "_____no_output_____" ], [ "px.scatter(df,\n x='prey_population',\n y='predator_population',\n color='timestep'\n animation_frame=')", "_____no_output_____" ], [ "vdsvds", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7c1d59b2cac367fd2992627391dab33da14588

767,678

ipynb

Jupyter Notebook

nlp/nlp.ipynb

lhy0807/used_car_playground

8a96f13324fb3daf24e3c5c372a21dfe936fc8b2

[ "MIT" ]

1

2020-10-17T17:05:48.000Z

nlp/nlp.ipynb

songxu2022/used_car_playground

f9eefd21f5ee35fb764ad64d4fb1736b02382f2d

[ "MIT" ]

17

2020-09-28T16:37:42.000Z

2021-09-22T19:36:46.000Z

nlp/nlp.ipynb

songxu2022/used_car_playground

f9eefd21f5ee35fb764ad64d4fb1736b02382f2d

[ "MIT" ]

5

2020-10-07T17:02:07.000Z

2020-10-19T04:22:34.000Z

628.728911

483,044

0.938658

[ [ [ "import numpy as np\nimport pandas as pd\nfrom sklearn.model_selection import train_test_split\n", "_____no_output_____" ], [ "df = pd.read_csv(\"usedcar.csv\")", "_____no_output_____" ], [ "df.head()", "_____no_output_____" ], [ "df.isna().sum()", "_____no_output_____" ], [ "df.dtypes", "_____no_output_____" ], [ "df['price'] = df['price'].str.replace(',', '')\ndf['mileage'] = df['mileage'].str.replace(',', '')\ndf['model'] = df['model'].str.replace('c', '')\n\ndf['price'] = df['price'].astype(int)\ndf['mileage'] = df['mileage'].astype(int)\ndf['model'] = df['model'].astype(int)\n\ndf.dtypes", "_____no_output_____" ], [ "x = df[['model','zip','mileage']]\ny = df['price']\n\nx_train, x_test, y_train, y_test = train_test_split(x, y, random_state=0)", "_____no_output_____" ], [ "from sklearn.neighbors import KNeighborsClassifier\n\nknn = KNeighborsClassifier(n_neighbors = 1)", "_____no_output_____" ], [ "knn.fit(x_train, y_train)", "_____no_output_____" ], [ "knn.score(x_test, y_test)", "_____no_output_____" ], [ "car_prediction = knn.predict([[21411, 11010, 40000]])\ncar_prediction[0]", "_____no_output_____" ], [ "from sklearn.linear_model import LinearRegression\nlin_reg=LinearRegression()\nlin_reg.fit(x,y)", "_____no_output_____" ], [ "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False)\n", "_____no_output_____" ], [ "lin_reg.score(x,y)", "_____no_output_____" ], [ "lin_reg.predict([[21411, 11010, 40000]])[0]", "_____no_output_____" ], [ "from sklearn.feature_extraction.text import CountVectorizer\nfrom sklearn.feature_extraction.text import TfidfVectorizer\nfrom sklearn.preprocessing import LabelBinarizer\nfrom nltk.corpus import stopwords\nfrom nltk.stem.porter import PorterStemmer\nfrom wordcloud import WordCloud,STOPWORDS\nfrom nltk.stem import WordNetLemmatizer\nfrom nltk.tokenize import word_tokenize,sent_tokenize\nfrom nltk.tokenize.toktok import ToktokTokenizer\nfrom nltk.stem import LancasterStemmer,WordNetLemmatizer\nfrom nltk import pos_tag\nfrom nltk.corpus import wordnet\nimport string", "_____no_output_____" ], [ "df.drop(columns=['model', 'zip','mileage'])", "_____no_output_____" ], [ "stop = set(stopwords.words('english'))\npunctuation = list(string.punctuation)\nstop.update(punctuation)", "_____no_output_____" ], [ "def get_simple_pos(tag):\n if tag.startswith('J'):\n return wordnet.ADJ\n elif tag.startswith('V'):\n return wordnet.VERB\n elif tag.startswith('N'):\n return wordnet.NOUN\n elif tag.startswith('R'):\n return wordnet.ADV\n else:\n return wordnet.NOUN", "_____no_output_____" ], [ "lemmatizer = WordNetLemmatizer()\ndef lemmatize_words(text):\n final_text = []\n for i in text.split():\n if i.strip().lower() not in stop:\n pos = pos_tag([i.strip()])\n word = lemmatizer.lemmatize(i.strip(),get_simple_pos(pos[0][1]))\n final_text.append(word.lower())\n return \" \".join(final_text)", "_____no_output_____" ], [ "df.text = df.text.apply(lemmatize_words)", "_____no_output_____" ], [ "df.head()", "_____no_output_____" ], [ "df.text", "_____no_output_____" ], [ "from sklearn.feature_extraction.text import CountVectorizer\nfrom sklearn.feature_extraction.text import TfidfVectorizer\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.linear_model import LogisticRegression,SGDClassifier\n", "_____no_output_____" ], [ "x_train,x_test,y_train,y_test = train_test_split(df.text,df.price,test_size = 0.2 , random_state = 0)", "_____no_output_____" ], [ "cv=CountVectorizer(min_df=0,max_df=1,binary=False,ngram_range=(1,3))\n#transformed train reviews\ncv_train_reviews=cv.fit_transform(x_train)\n#transformed test reviews\ncv_test_reviews=cv.transform(x_test)", "_____no_output_____" ], [ "tv=TfidfVectorizer(min_df=0,max_df=1,use_idf=True,ngram_range=(1,3))\n#transformed train reviews\ntv_train_reviews=tv.fit_transform(x_train)\n#transformed test reviews\ntv_test_reviews=tv.transform(x_test)", "_____no_output_____" ], [ "lr=LogisticRegression(penalty='l2',max_iter=500,C=1,random_state=0)\n#Fitting the model for Bag of words\nlr_bow=lr.fit(cv_train_reviews,y_train)\nprint(lr_bow)\n#Fitting the model for tfidf features\nlr_tfidf=lr.fit(tv_train_reviews,y_train)\nprint(lr_tfidf)", "LogisticRegression(C=1, max_iter=500, random_state=0)\nLogisticRegression(C=1, max_iter=500, random_state=0)\n" ], [ "good = x_train[y_train[y_train > 30000].index]\nbad = x_train[y_train[y_train < 5000].index]\nx_train.shape,good.shape,bad.shape", "_____no_output_____" ], [ "import matplotlib.pyplot as plt\n", "_____no_output_____" ], [ "plt.figure(figsize = (20,20)) # Text Reviews with Poor Ratings\nwc = WordCloud(min_font_size = 3, max_words = 3000 , width = 1600 , height = 800).generate(\" \".join(bad))\nplt.imshow(wc,interpolation = 'bilinear')", "_____no_output_____" ], [ "plt.figure(figsize = (20,20)) # # Text Reviews with Good Ratings\nwc = WordCloud(min_font_size = 3, max_words = 3000 , width = 1600 , height = 800).generate(\" \".join(good))\nplt.imshow(wc,interpolation = 'bilinear')", "_____no_output_____" ], [ "from sklearn.naive_bayes import MultinomialNB\nfrom sklearn.metrics import classification_report,confusion_matrix,accuracy_score", "_____no_output_____" ], [ "#training the model\nmnb=MultinomialNB()\n#fitting the nb for bag of words\nmnb_bow=mnb.fit(cv_train_reviews,y_train)\nprint(mnb_bow)\n#fitting the nb for tfidf features\nmnb_tfidf=mnb.fit(tv_train_reviews,y_train)\nprint(mnb_tfidf)", "MultinomialNB()\nMultinomialNB()\n" ], [ "#Predicting the model for bag of words\nmnb_bow_predict=mnb.predict(cv_test_reviews)\n#Predicting the model for tfidf features\nmnb_tfidf_predict=mnb.predict(tv_test_reviews)", "_____no_output_____" ], [ "mnb_bow_report = classification_report(y_test,mnb_bow_predict,target_names = ['0','1'])\nprint(mnb_bow_report)\nmnb_tfidf_report = classification_report(y_test,mnb_tfidf_predict,target_names = ['0','1'])\nprint(mnb_tfidf_report)", "_____no_output_____" ], [ "import keras\nfrom keras.layers import Dense,LSTM\nfrom keras.models import Sequential", "_____no_output_____" ], [ "model = Sequential()\nmodel.add(Dense(units = 75 , activation = 'relu' , input_dim = cv_train_reviews.shape[1]))\nmodel.add(Dense(units = 50 , activation = 'relu'))\nmodel.add(Dense(units = 25 , activation = 'relu'))\nmodel.add(Dense(units = 10 , activation = 'relu')) \nmodel.add(Dense(units = 1 , activation = 'sigmoid'))\nmodel.compile(optimizer = 'adam' , loss = 'binary_crossentropy' , metrics = ['accuracy'])", "_____no_output_____" ], [ "model.summary()", "Model: \"sequential_1\"\n_________________________________________________________________\nLayer (type) Output Shape Param # \n=================================================================\ndense (Dense) (None, 75) 89850 \n_________________________________________________________________\ndense_1 (Dense) (None, 50) 3800 \n_________________________________________________________________\ndense_2 (Dense) (None, 25) 1275 \n_________________________________________________________________\ndense_3 (Dense) (None, 10) 260 \n_________________________________________________________________\ndense_4 (Dense) (None, 1) 11 \n=================================================================\nTotal params: 95,196\nTrainable params: 95,196\nNon-trainable params: 0\n_________________________________________________________________\n" ], [ "model.fit(cv_train_reviews,y_train , epochs = 10)\n", "WARNING:tensorflow:Keras is training/fitting/evaluating on array-like data. Keras may not be optimized for this format, so if your input data format is supported by TensorFlow I/O (https://github.com/tensorflow/io) we recommend using that to load a Dataset instead.\nEpoch 1/10\n" ], [ "model.evaluate(cv_test_reviews,y_test)[1]\n", "_____no_output_____" ], [ "train_data = pd.read_csv('../input/used-cars-price-prediction/train-data.csv')\ntest_data = pd.read_csv('../input/used-cars-price-prediction/test-data.csv')", "_____no_output_____" ], [ "train_data.info()", "_____no_output_____" ], [ "train_data.head()", "_____no_output_____" ], [ "train_data.tail()", "_____no_output_____" ], [ "train_data = train_data.iloc[:,1:]\ntrain_data.head()", "_____no_output_____" ], [ "train_data.describe()\n", "_____no_output_____" ], [ "train_data.shape", "_____no_output_____" ], [ "train_data['Kilometers_Driven'].value_counts()", "_____no_output_____" ], [ "from sklearn.model_selection import train_test_split\nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 25)", "_____no_output_____" ], [ "from sklearn.linear_model import LinearRegression\nlinear_reg = LinearRegression()\nlinear_reg.fit(X_train, y_train)\ny_pred= linear_reg.predict(X_test)\nprint(\"Accuracy on Traing set: \",linear_reg.score(X_train,y_train))\nprint(\"Accuracy on Testing set: \",linear_reg.score(X_test,y_test))", "_____no_output_____" ], [ "from sklearn.ensemble import RandomForestRegressor\nrf_reg = RandomForestRegressor()\nrf_reg.fit(X_train, y_train)\ny_pred= rf_reg.predict(X_test)\nprint(\"Accuracy on Traing set: \",rf_reg.score(X_train,y_train))\nprint(\"Accuracy on Testing set: \",rf_reg.score(X_test,y_test))", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7c51ccf1dec150a49127b230d9c97d07219b4b

28,819

ipynb

Jupyter Notebook

webscraping-selenium/WebScrap_OLX.ipynb

cnornberg/data-engineering

c9a2d8dec03a6ae2904c239d190437a6f636cbd2

[ "MIT" ]

null

webscraping-selenium/WebScrap_OLX.ipynb

cnornberg/data-engineering

c9a2d8dec03a6ae2904c239d190437a6f636cbd2

[ "MIT" ]

null

webscraping-selenium/WebScrap_OLX.ipynb

cnornberg/data-engineering

c9a2d8dec03a6ae2904c239d190437a6f636cbd2

[ "MIT" ]

null

43.077728

287

0.505882

[ [ [ "# Webscraping utilizando BeautifulSoup e Selenium\n# Busca de produtos OLX\n\n\n", "_____no_output_____" ] ], [ [ "# !pip install bs4\n\nimport pandas as pd\nimport numpy as np\nimport requests\nfrom bs4 import BeautifulSoup\nimport re\nimport time\n\n#selenium\n#!pip install selenium\nfrom selenium import webdriver\nfrom selenium.webdriver.common.keys import Keys\nfrom selenium.webdriver.common.by import By\nimport time", "_____no_output_____" ], [ "#pg1: https://sp.olx.com.br/regiao-de-bauru-e-marilia?ot=1&q=honda\n#pg2: https://sp.olx.com.br/regiao-de-bauru-e-marilia?ot=1&o=2&q=honda\n# trocar pagina = &o=2\n\n# Last page: <a href=\"https://sp.olx.com.br/regiao-de-bauru-e-marilia?ot=1&o=12&q=honda\" data-lurker-detail=\"last_page\" class=\"sc-1bofr6e-0 iRQkdN\"><span class=\"sc-1bofr6e-1 iUNkan sc-ifAKCX bBbnjQ\" color=\"dark\" font-weight=\"400\">Última pagina</span></a>\n\n# numero de paginas o=12\n# data-lurker-detail=\"last_page\" \n# class=\"sc-1bofr6e-0 iRQkdN\">\n# \n", "Collecting bs4\n Downloading bs4-0.0.1.tar.gz (1.1 kB)\nCollecting beautifulsoup4\n Downloading beautifulsoup4-4.9.3-py3-none-any.whl (115 kB)\n\u001b[K |████████████████████████████████| 115 kB 2.9 MB/s eta 0:00:01\n\u001b[?25hCollecting soupsieve>1.2\n Downloading soupsieve-2.2.1-py3-none-any.whl (33 kB)\nBuilding wheels for collected packages: bs4\n Building wheel for bs4 (setup.py) ... \u001b[?25ldone\n\u001b[?25h Created wheel for bs4: filename=bs4-0.0.1-py3-none-any.whl size=1273 sha256=56b9db4c7aad64f32f9b7be46b87a1fbd14cf5cc55277a8da6c755e3f2f3f628\n Stored in directory: /home/cristian/.cache/pip/wheels/73/2b/cb/099980278a0c9a3e57ff1a89875ec07bfa0b6fcbebb9a8cad3\nSuccessfully built bs4\nInstalling collected packages: soupsieve, beautifulsoup4, bs4\nSuccessfully installed beautifulsoup4-4.9.3 bs4-0.0.1 soupsieve-2.2.1\n" ] ], [ [ "# Webscraping do site OLX - Busca de produtos\n", "_____no_output_____" ] ], [ [ "# Buscar:\n# Link da busca do produto\nurl = \"https://sp.olx.com.br/regiao-de-bauru-e-marilia?ot=1&q=honda\"\n\n# Necessário passar alguns parametros para que o request não seja bloqueado pelo site.\nPARAMS = {\n #definicoes do header do site\n \"autority\": \"sp.olx.com.br\",\n \"method\": \"GET\",\n \"path\": \"/regiao-de-bauru-e-marilia\",\n \"scheme\": \"https\",\n \"referrer\": \"https://sp.olx.com.br/\",\n \"sec-fetch-mode\": \"navigate\",\n \"sec-fecth-site\": \"same-origin\",\n \"sec-fech-user\": \"?1\",\n \"upgrade-insecure-requests\": \"1\",\n \"user-agent\": \"Mozilla/5.0\"\n}\n\n# Ler o conteudo da pagina e salvar numa variavel\npages = requests.get(url = url, headers = PARAMS)\n\n# Converter o conteudo da pagina no formato lxml, que possibilita buscar partes da pagina utilizando tags, classes, etc.\nconteudo = BeautifulSoup(pages.content, \"lxml\")\n\n#results = conteudo.find_all(\"li\", {\"class\" : \"nome_classe\"}) # resultado da busca - uma lista \n\n#for a in results: # percorrer todos os resultados da busca\n# try:\n# a.find_all(\"h2\")[0].contents[0]) # retorna o titulo do anuncio\n# a.find_all(\"p\", {\"class\" : \"nome_classe\"})[0].contents[0] # retorna a informação de um objeto/classe\n# except:\n# print(erro)\n\n#print(conteudo)\n", "_____no_output_____" ], [ "print(pages)", "<Response [200]>\n" ], [ "#print(conteudo.find_all(\"div\", {\"class\":\"fnmrjs-9 gqfQzY\"}))\n\n#print(len(conteudo.find_all(\"h2\")))", "100\n" ] ], [ [ "Este exemplo está retornando apenas os resultados da primeira página da busca.\n\n\nPara obter o resultado das outras páginas é necessario utilizar um loop, adicionando ao link da busca &o=2 (numero da pagina).", "_____no_output_____" ] ], [ [ "# Busca na página por uma lista, que nesse exmplo são os produtos anunciados e salva numa variavel itens\n# Para obter as tags como \"li\" e a \"class\" utilizamos o inspect do site\nitens = conteudo.find_all(\"li\", {\"class\" : \"sc-1fcmfeb-2 juiJqh\"})\n\n\n# Passar pela lista dos itens para obter detalhes dos anuncios\nfor a in itens:\n try:\n nomeItem = a.find_all(\"h2\")[0].contents[0] \n linkItem = a.find(\"a\")[\"href\"]\n precoItem = a.find_all(\"span\", {\"class\" : \"sc-ifAKCX eoKYee\"})[0].contents[0]\n precoItem = (precoItem.split()[1])\n precoItem = float(precoItem.replace(\".\", \"\"))\n precoOferta = precoItem * 0.8\n # <span color=\"dark\" class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" font-weight=\"400\">Hoje</span>\n data = a.find_all(\"span\", {\"class\" : \"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\"})\n dataItem = data[0].contents[0]\n horaItem = data[1].contents[0]\n horaItem = dataItem[1]\n #<span color=\"dark\" class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" font-weight=\"400\">16:31</span>\n horaItem = a.find_all(\"span\", {\"class\" : \"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\"})\n print(\"Link: \", linkItem)\n print(\"Preço: \",precoItem)\n print(\"Oferta: \", precoOferta)\n print(\"------\")\n print(nomeItem, precoItem)\n if dataItem == \"Hoje\":\n print(dataItem, horaItem)\n \n \n\n except:\n print(\"\")\n\n\n \n \n#print(linkItem)\n\n", "Link: https://sp.olx.com.br/regiao-de-bauru-e-marilia/ciclismo/skate-hondar-904571455\nPreço: 1.0\nOferta: 0.8\n------\nSkate Hondar 1.0\n\n\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/carros-vans-e-utilitarios/honda-civic-904458664\nPreço: 20000.0\nOferta: 16000.0\n------\nHonda Civic. 20000.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/carros-vans-e-utilitarios/honda-civic-904723723\nPreço: 36500.0\nOferta: 29200.0\n------\nHonda Civic 36500.0\nHoje [<span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">Hoje</span>, <span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">11:40</span>]\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-150-904363143\nPreço: 7000.0\nOferta: 5600.0\n------\nHonda 150 7000.0\n\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/pecas-e-acessorios/motos/escapamento-honda-904139228\nPreço: 150.0\nOferta: 120.0\n------\nEscapamento Honda 150.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/pecas-e-acessorios/motos/escapamento-honda-904124873\nPreço: 170.0\nOferta: 136.0\n------\nEscapamento honda 170.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/audio-tv-video-e-fotografia/radio-honda-civic-904836698\nPreço: 250.0\nOferta: 200.0\n------\nRÁDIO HONDA CIVIC 250.0\nHoje [<span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">Hoje</span>, <span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">15:01</span>]\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-bros-150-904916905\nPreço: 6500.0\nOferta: 5200.0\n------\nHonda Bros 150 6500.0\nHoje [<span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">Hoje</span>, <span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">17:35</span>]\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-biz-2014-904793488\nPreço: 9500.0\nOferta: 7600.0\n------\nHonda Biz 2014 9500.0\nHoje [<span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">Hoje</span>, <span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">13:54</span>]\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/carros-vans-e-utilitarios/honda-fit-2016-888341897\nPreço: 63000.0\nOferta: 50400.0\n------\nHonda Fit 2016 63000.0\nHoje [<span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">Hoje</span>, <span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">10:35</span>]\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/motocicleta-honda-903006773\nPreço: 15400.0\nOferta: 12320.0\n------\nMotocicleta Honda 15400.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-nx200-2001-904107733\nPreço: 6000.0\nOferta: 4800.0\n------\nHonda nx200 2001 6000.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/carros-vans-e-utilitarios/honda-city-2018-903805751\nPreço: 70000.0\nOferta: 56000.0\n------\nHonda City 2018 70000.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/carros-vans-e-utilitarios/honda-city-890024879\nPreço: 65500.0\nOferta: 52400.0\n------\nHonda city 65500.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-125-ml-903434972\nPreço: 6000.0\nOferta: 4800.0\n------\nHonda 125 ML 6000.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/moto-honda-902716026\nPreço: 6000.0\nOferta: 4800.0\n------\nMoto Honda 6000.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-falcon-400-903552430\nPreço: 13000.0\nOferta: 10400.0\n------\nHonda Falcon 400 13000.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-twister-250cc-884526403\nPreço: 8500.0\nOferta: 6800.0\n------\nHonda/twister 250cc 8500.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-xre-rally-2019-904964549\nPreço: 23000.0\nOferta: 18400.0\n------\nHonda XRE RALLY 2019 23000.0\nHoje [<span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">Hoje</span>, <span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">19:16</span>]\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-biz-2002-100cc-904946168\nPreço: 4300.0\nOferta: 3440.0\n------\nHonda biz 2002 100cc 4300.0\nHoje [<span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">Hoje</span>, <span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">18:42</span>]\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-biz-2002-100cc-904944927\nPreço: 4300.0\nOferta: 3440.0\n------\nHonda biz 2002 100cc 4300.0\nHoje [<span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">Hoje</span>, <span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">18:38</span>]\n\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/moto-vfr-1200cc-da-honda-904897033\nPreço: 0.0\nOferta: 0.0\n------\nmoto vfr 1200cc da honda 0.0\nHoje [<span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">Hoje</span>, <span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">17:00</span>]\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/carros-vans-e-utilitarios/honda-fit-exl-903280019\nPreço: 62380.0\nOferta: 49904.0\n------\nHonda Fit Exl 62380.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/carros-vans-e-utilitarios/honda-civic-902473972\nPreço: 35900.0\nOferta: 28720.0\n------\nHonda civic 35900.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-pop-110i-2016-897861041\nPreço: 7500.0\nOferta: 6000.0\n------\nHonda Pop 110i 2016 7500.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-biz-rosa-903158898\nPreço: 10100.0\nOferta: 8080.0\n------\nHonda Biz rosa 10100.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-fan-160cc-2016-885553233\nPreço: 11900.0\nOferta: 9520.0\n------\nHonda/Fan 160cc 2016 11900.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-titan-150-sport-904432589\nPreço: 8500.0\nOferta: 6800.0\n------\nHonda titan 150 Sport 8500.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-nxr-160-bros-904750994\nPreço: 5500.0\nOferta: 4400.0\n------\nHonda NXR 160 BROS 5500.0\nHoje [<span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">Hoje</span>, <span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">12:25</span>]\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-biz-100-2004-904914272\nPreço: 4500.0\nOferta: 3600.0\n------\nHONDA BIZ 100 2004 4500.0\nHoje [<span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">Hoje</span>, <span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">17:30</span>]\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-pcx-dlx-2021-904772708\nPreço: 18000.0\nOferta: 14400.0\n------\nHonda Pcx/DLX 2021 18000.0\nHoje [<span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">Hoje</span>, <span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">13:03</span>]\n\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-cb500-2003-2003-902927488\nPreço: 13000.0\nOferta: 10400.0\n------\nHonda Cb500 2003/2003 13000.0\nHoje [<span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">Hoje</span>, <span class=\"wlwg1t-1 fsgKJO sc-ifAKCX eLPYJb\" color=\"dark\" font-weight=\"400\">09:19</span>]\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/carros-vans-e-utilitarios/honda-civic-sport-cvt-904283654\nPreço: 96500.0\nOferta: 77200.0\n------\nHonda civic sport cvt 96500.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/carros-vans-e-utilitarios/honda-civic-lxs-903145417\nPreço: 55000.0\nOferta: 44000.0\n------\nHonda Civic Lxs 55000.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/carros-vans-e-utilitarios/honda-fit-2005-manual-904524895\nPreço: 18500.0\nOferta: 14800.0\n------\nHonda Fit 2005 manual 18500.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-pcx-150-898026682\nPreço: 16900.0\nOferta: 13520.0\n------\nHONDA/ PCX-150 16900.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/vendo-honda-biz-es-904306369\nPreço: 6900.0\nOferta: 5520.0\n------\nVendo Honda biz ES 6900.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-biz-125-903006672\nPreço: 9500.0\nOferta: 7600.0\n------\nHonda biz 125 9500.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-biz-2014-902995980\nPreço: 9500.0\nOferta: 7600.0\n------\nHonda biz 2014 9500.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/carros-vans-e-utilitarios/honda-crv-lx-902990732\nPreço: 50000.0\nOferta: 40000.0\n------\nHonda CRV LX 50000.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/moto-honda-falcon-902797445\nPreço: 10800.0\nOferta: 8640.0\n------\nMoto Honda Falcon 10800.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/carros-vans-e-utilitarios/honda-civic-exl-2017-903835559\nPreço: 97000.0\nOferta: 77600.0\n------\nHONDA CIVIC EXL 2017 97000.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-nx4-falcon-2008-903883209\nPreço: 13400.0\nOferta: 10720.0\n------\nHonda NX4 Falcon 2008 13400.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/carros-vans-e-utilitarios/honda-fit-ex-1-5-903883616\nPreço: 76900.0\nOferta: 61520.0\n------\nHonda Fit EX 1.5 76900.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/pecas-e-acessorios/motos/bloco-original-honda-160-903580953\nPreço: 50.0\nOferta: 40.0\n------\nBloco original Honda 160 50.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/pecas-e-acessorios/motos/susuki-honda-yamaha-shadow-903488586\nPreço: 300.0\nOferta: 240.0\n------\nSusuki honda yamaha shadow 300.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-cb-twister-2017-903422247\nPreço: 14900.0\nOferta: 11920.0\n------\nHonda cb Twister 2017 14900.0\nLink: https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-xre-300-p-903299041\nPreço: 20490.0\nOferta: 16392.0\n------\nHonda Xre 300 P 20490.0\n" ], [ "print(linkItem)", "https://sp.olx.com.br/regiao-de-bauru-e-marilia/autos-e-pecas/motos/honda-xre-300-p-903299041\n" ] ], [ [ "# Utilizar o Selenium para navegar automaticamente pelo site", "_____no_output_____" ] ], [ [ "#!pip install selenium\nfrom selenium import webdriver\nfrom selenium.webdriver.common.keys import Keys\nfrom selenium.webdriver.common.by import By\nimport time", "_____no_output_____" ], [ "# Necessário instalar um pacote/driver do sistema operacional\n# $ sudo apt install firefox-geckodriver\n\n# Inicializar o driver do Firefox:\nbrowser = webdriver.Firefox()\n\n# Abrir site de login:\nbrowser.get(\"https://conta.olx.com.br/acesso/\")\n\n# fazer login - Utilizamos o inspect para descobrir o type dos campos que precisam ser preenchidos\nusername = browser.find_element_by_xpath(\"//input[@type='email']\")\npassword = browser.find_element_by_xpath(\"//input[@type='password']\")\n\nusername.send_keys(\"usuario\")\npassword.send_keys(\"senha\")\n\n# Localizar o botão de submit e clicar no botão para realizar o login\nbrowser.find_element_by_xpath(\"//button[@type='text']\").click()\n\n# Aguardar alguns segundos para realizar o login com sucesso\ntime.sleep(5)\n\n# Abrir link da nova pagina apos fazer login\nbrowser.get(linkItem)\n\n# Clicar no botao Chat\n#<span color=\"white\" data-testid=\"largeFloatButton-title\" font-weight=\"400\" class=\"sc-ifAKCX eUXUWN\">Chat</span>\nbrowser.find_element_by_xpath(\"//span[@data-testid='largeFloatButton-title']\").click()\ntime.sleep(5)\n\n\n# inserir mensagem no campo do Chat:\n#<textarea placeholder=\"Digite uma mensagem...\" class=\"sc-hycgNl gACIxP sc-chAAoq fEgJLD\" style=\"height: 47px;\"></textarea>\nchat = browser.find_element_by_xpath(\"//textarea[@placeholder='Digite uma mensagem...']\")\nchat.send_keys(\"Aceita troca?\" + Keys.ENTER)\n\n# Fecha o navegador\nbrowser.quit()", "_____no_output_____" ], [ "", "_____no_output_____" ], [ "", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ] ]

ec7c5bde97361bf63a0fc58a4002d59b2540dacf

108,454

ipynb

Jupyter Notebook

analysis.ipynb

rkp8000/crossing_over

c89e300c115e4072cd52a733cda419b0d518c54d

[ "MIT" ]

null

analysis.ipynb

rkp8000/crossing_over

c89e300c115e4072cd52a733cda419b0d518c54d

[ "MIT" ]

null

analysis.ipynb

rkp8000/crossing_over

c89e300c115e4072cd52a733cda419b0d518c54d

[ "MIT" ]

null

293.118919

37,216

0.918712

[ [ [ "%matplotlib inline\nimport hashlib\nimport string\nimport matplotlib.pyplot as plt\nimport numpy as np\n\nfrom disp import set_font_size\n\n\nA = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRZTUVWXYZ '\nM = len(A)\nZ = 50\nN = 500", "_____no_output_____" ] ], [ [ "Key functions", "_____no_output_____" ] ], [ [ "def gen_seed(x_k, x_l):\n str_repr = ''.join(x_k.astype(str)) + ''.join(x_l.astype(str))\n return int(hashlib.sha256(str_repr.encode('utf-8')).hexdigest(), 16) % 10**9\n \n\ndef braid(x_k, x_l, q):\n np.random.seed(gen_seed(x_k, x_l))\n mask = np.random.rand(len(x_k)) < 1/(q+1)\n u = x_k.copy()\n u[mask] = x_l[mask]\n return u\n\n\ndef hamming(x_k, x_l):\n return np.mean(x_k != x_l)", "_____no_output_____" ] ], [ [ "Setup", "_____no_output_____" ] ], [ [ "np.random.seed(0)\nx_star = np.random.randint(0, Z, N)\nxs = [np.random.randint(0, Z, N) for i in range(M)]\n\ndef encode(s):\n y = x_star.copy()\n for t, a in enumerate(s):\n idx = A.index(a)\n y = braid(y, xs[idx], t+1)\n return y\n\n\ndef decode(y, l, return_d=False, force_decode=None):\n # get *set* of elements in sequence\n min_idxs = np.argsort([hamming(x, y) for x in xs])[:l]\n vs = [xs[idx] for idx in min_idxs]\n # reconstruct sequence\n y_star = x_star.copy()\n s_hat = ''\n d = [hamming(y_star, y)]\n for t in range(1, l+1):\n us = [braid(y_star, v, t) for v in vs]\n j = np.argmin([hamming(u, y) for u in us])\n \n if force_decode is not None and len(force_decode) >= t:\n next_sym = force_decode[t-1]\n next_x = xs[A.index(next_sym)]\n else:\n next_sym = A[min_idxs[j]]\n next_x = vs[j]\n \n s_hat += next_sym\n y_star = braid(y_star, next_x, t)\n d.append(hamming(y_star, y))\n \n if not all(y_star == y):\n print('Default reconstruction failed.')\n \n if not return_d:\n return s_hat\n else:\n return s_hat, np.array(d)", "_____no_output_____" ] ], [ [ "Demo", "_____no_output_____" ] ], [ [ "print('Hamming similarities between symbols:')\nsymbol_dists = []\nfor k in range(M-1):\n for l in range(k+1, M):\n symbol_dists.append(1 - hamming(xs[k], xs[l]))\nprint('Min = ', np.min(symbol_dists))\nprint('Max = ', np.max(symbol_dists))\nprint('Mean = ', np.mean(symbol_dists))\nprint('Std = ', np.std(symbol_dists))", "Hamming similarities between symbols:\nMin = 0.0020000000000000018\nMax = 0.04400000000000004\nMean = 0.02023076923076925\nStd = 0.006152548847290739\n" ], [ "s = 'random vectors for the win'\ny = encode(s)\nprint(y)\ns_star = decode(y, len(s))\nprint(s_star)", "[25 9 0 3 27 25 47 34 41 9 12 11 31 1 25 41 1 12 42 1 21 16 30 18\n 3 48 20 42 16 4 25 33 6 6 43 25 16 16 8 36 32 43 23 14 21 10 16 15\n 13 48 4 9 23 26 44 24 1 33 28 33 37 27 18 44 48 47 0 43 2 4 49 34\n 46 24 23 5 17 47 26 5 47 3 42 21 16 42 1 25 16 30 3 4 34 19 46 42\n 15 31 23 16 24 30 39 2 31 40 28 18 5 1 39 29 13 30 21 29 30 6 29 45\n 20 30 33 12 6 39 23 17 40 43 24 12 9 9 44 37 34 3 29 46 24 43 47 7\n 37 12 9 30 21 38 37 41 22 1 24 21 13 2 33 4 17 45 25 10 43 19 20 10\n 7 49 24 25 5 15 3 19 42 0 31 20 18 4 13 47 6 34 16 13 35 22 2 49\n 37 18 49 41 24 30 0 5 42 38 47 18 47 24 11 28 29 20 10 31 6 48 24 5\n 25 24 25 5 27 17 21 21 49 0 41 45 48 20 0 40 17 46 9 35 49 22 34 10\n 18 20 38 11 18 32 20 18 38 15 46 22 9 26 26 47 22 20 41 21 26 45 8 26\n 24 37 43 23 23 14 38 0 40 44 42 49 11 49 15 36 0 3 30 11 23 37 12 4\n 25 34 32 23 41 8 20 16 8 17 10 20 18 28 9 31 27 44 0 26 46 24 36 45\n 41 47 38 16 5 41 42 40 38 38 6 27 46 37 38 13 32 11 35 33 27 12 31 34\n 27 8 25 36 17 25 26 35 46 26 23 24 38 35 41 9 14 8 22 23 17 35 34 28\n 28 11 13 17 1 48 16 6 47 27 5 9 6 16 38 2 13 2 11 33 3 0 42 47\n 2 2 6 24 5 25 13 40 19 27 7 17 45 28 31 47 38 36 35 20 36 46 40 24\n 49 15 4 1 16 23 30 8 11 27 1 49 34 6 25 11 17 41 44 10 44 33 30 17\n 24 26 24 37 8 9 15 34 38 23 3 18 24 48 34 0 36 41 7 49 3 38 31 20\n 32 34 37 6 4 12 47 5 1 1 16 44 2 24 42 19 43 36 38 45 29 41 30 35\n 33 11 21 12 33 4 0 44 8 9 33 24 29 11 5 29 3 28 32 26]\nrandom vectors for the win\n" ], [ "test_seqs = [\n 'abcde',\n 'abced',\n 'aabcd',\n 'zyxwv',\n 'zyxvw',\n 'aaazz',\n 'abbbaaab',\n 'hi agostina'\n]\n\nfor s in test_seqs:\n s_hat = decode(encode(s), len(s))\n print(s, '--> y --> ', s_hat, '(', s == s_hat, ')')", "abcde --> y --> abcde ( True )\nabced --> y --> abced ( True )\naabcd --> y --> aabcd ( True )\nzyxwv --> y --> zyxwv ( True )\nzyxvw --> y --> zyxvw ( True )\naaazz --> y --> aaazz ( True )\nabbbaaab --> y --> abbbaaab ( True )\nhi agostina --> y --> hi agostina ( True )\n" ], [ "s = 'hyperdimensional computing via crossover'\ny = encode(s)\ns_hat, d = decode(y, len(s), return_d=True)\n\nt = np.arange(len(s)+1)\nfig, ax = plt.subplots(1, 1, figsize=(10, 5), tight_layout=True)\nax.plot(t, d, lw=2, c='k')\n\nax.plot(t, 1 - t/(len(t)-1), c='gray', ls='--')\n\nax.set_xlim(-1, len(t))\nax.set_ylim(-.05, 1.05)\n\nax.set_xticks(t)\nax.set_xticklabels('*' + s_hat)\n\nax.grid()\n\nset_font_size(ax, 16)", "Default reconstruction failed.\n" ], [ "s = 'hyperdimensional computing via crossover'\ny = encode(s)\ns_hat, d = decode(y, len(s), return_d=True, force_decode='h')\n\nt = np.arange(len(s)+1)\nfig, ax = plt.subplots(1, 1, figsize=(10, 5), tight_layout=True)\nax.plot(t, d, lw=2, c='k')\n\nax.plot(t, 1 - t/(len(t)-1), c='gray', ls='--')\n\nax.set_xlim(-1, len(t))\nax.set_ylim(-.05, 1.05)\n\nax.set_xticks(t)\nax.set_xticklabels('*' + s_hat)\n\nax.grid()\n\nset_font_size(ax, 16)", "_____no_output_____" ], [ "s = 'aaaaaaabbbbbbbccccccc'\ny = encode(s)\ns_hat, d = decode(y, len(s), return_d=True, force_decode='aaaaaaabbbbbbbccccccc')\n\nt = np.arange(len(s)+1)\nfig, ax = plt.subplots(1, 1, figsize=(10, 5), tight_layout=True)\nax.plot(t, d, lw=2, c='k')\n\nax.plot(t, 1 - t/(len(t)-1), c='gray', ls='--')\n\nax.set_xlim(-1, len(t))\nax.set_ylim(-.05, 1.05)\n\nax.set_xticks(t)\nax.set_xticklabels('*' + s_hat)\n\nax.grid()\n\nset_font_size(ax, 16)", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code" ] ]

ec7c5e207cae6b74e1e0f8065b688d6fad12a238

137,252

ipynb

Jupyter Notebook

assets/all_html/2019_11_21_Inmates_Import_v3.ipynb

dskw1/dskw1.github.io

ee85aaa7c99c4320cfac95e26063beaac3ae6fcb

[ "MIT" ]

null

assets/all_html/2019_11_21_Inmates_Import_v3.ipynb

dskw1/dskw1.github.io

ee85aaa7c99c4320cfac95e26063beaac3ae6fcb

[ "MIT" ]

1

2022-03-24T18:28:16.000Z

assets/all_html/2019_11_21_Inmates_Import_v3.ipynb

dskw1/dskw1.github.io

ee85aaa7c99c4320cfac95e26063beaac3ae6fcb

[ "MIT" ]

1

2021-09-01T16:54:38.000Z

31.087656

130

0.362436

[ [ [ "import os\ndef get_data_from_files(path):\n directory = os.listdir(path)\n results = []\n filenames = []\n for file in directory:\n f=open(path+file)\n filenames.append(file)\n results.append(f.read())\n f.close()\n return results, filenames\n\ninmates, filenames = get_data_from_files('FinalProject/inmates/')\n\nimport pandas as pd\nimport numpy as np\ndf = pd.DataFrame(inmates, filenames)", "_____no_output_____" ], [ "df.reset_index(inplace=True)", "_____no_output_____" ], [ "df['inmate_number'] = df.apply(lambda x: x['index'].split('_')[2], axis=1)\ndf['last_name'] = df.apply(lambda x: x['index'].split('_')[4].split('.')[0], axis=1)\ndf['first_name'] = df.apply(lambda x: x['index'].split('_')[3], axis=1)\ndf", "_____no_output_____" ], [ "import re\ndf['clean'] = df.apply(lambda x: re.sub(r'[\\W_]+', ' ', x[0].lower()), axis=1)", "_____no_output_____" ], [ "df", "_____no_output_____" ], [ "def get_occupation(summary):\n try:\n p = re.compile(r'(?<=occupation)(\\W.*?)(?=\\s)')\n r = p.search(summary).group().strip()\n if 'prior' in r or len(r) < 3:\n return 'none_listed'\n else:\n return r\n except:\n return 'none_listed'\n\noccupations = [get_occupation(summary) for summary in df['clean'].values]", "_____no_output_____" ], [ "occupations", "_____no_output_____" ], [ "df['occupation'] = occupations", "_____no_output_____" ], [ "df", "_____no_output_____" ], [ "def get_priors(summary):\n try:\n text = re.compile(r'(?<=record)(\\W.*?)(?=\\s)')\n result = text.search(summary).group().strip()\n return 'no' if 'none' in result else 'yes'\n except:\n return 'none_listed'\n\npriors = [get_priors(summary) for summary in df['clean'].values]\ndf['prior_record'] = priors\ndf", "_____no_output_____" ], [ "def get_edu(summary):\n try:\n text = re.compile(r'(?<=education)(.*?)(years|yrs|ged|prior)')\n result = text.search(summary).group().strip()\n number = re.compile(r'\\d+')\n number_result = number.search(result).group()\n# print(number_result)\n return str(number_result) + \" years\"\n# return 'no' if 'none' in result else 'yes'\n except:\n return 'none_listed'\n\nedu = [get_edu(summary) for summary in df['clean'].values]\ndf['education_level'] = edu\ndf.to_csv('V8_fromphotos.csv')", "_____no_output_____" ], [ "def get_vics(summary):\n try:\n text = re.compile(r'(male|men|man)')\n vics = text.findall(summary)\n print(len(vics))\n\n except:\n print('nope')\n \n \nvic_f = [get_vics(summary) for summary in df['clean'].values] ", "1\n1\n1\n3\n1\n0\n3\n0\n3\n3\n1\n2\n4\n4\n1\n2\n2\n2\n1\n2\n0\n1\n0\n2\n1\n1\n0\n2\n2\n2\n1\n3\n0\n1\n3\n0\n2\n0\n3\n3\n2\n1\n2\n2\n1\n0\n0\n1\n6\n2\n1\n2\n8\n3\n0\n0\n3\n1\n4\n1\n3\n2\n3\n6\n4\n0\n0\n1\n0\n0\n3\n1\n0\n1\n7\n3\n0\n0\n1\n3\n0\n2\n3\n0\n1\n3\n0\n1\n1\n3\n4\n8\n1\n0\n4\n4\n0\n0\n2\n2\n0\n0\n1\n0\n5\n1\n2\n1\n2\n2\n0\n2\n1\n1\n0\n2\n0\n1\n1\n0\n8\n5\n2\n2\n2\n3\n4\n1\n6\n1\n1\n1\n0\n2\n0\n4\n3\n1\n2\n0\n3\n0\n1\n5\n1\n1\n2\n2\n4\n0\n1\n2\n1\n1\n2\n3\n0\n3\n1\n0\n0\n2\n1\n1\n1\n0\n1\n1\n4\n1\n5\n1\n7\n6\n2\n1\n1\n1\n4\n0\n1\n1\n2\n0\n0\n3\n0\n5\n1\n2\n0\n2\n2\n0\n2\n3\n0\n2\n5\n1\n1\n1\n4\n0\n1\n2\n1\n1\n1\n0\n6\n5\n0\n1\n2\n1\n0\n1\n1\n1\n1\n1\n3\n0\n1\n2\n3\n0\n2\n2\n3\n0\n2\n5\n2\n2\n0\n2\n1\n1\n1\n2\n2\n3\n2\n3\n1\n3\n1\n2\n1\n1\n2\n7\n4\n0\n5\n1\n1\n2\n0\n1\n0\n10\n3\n2\n1\n2\n0\n1\n0\n1\n3\n0\n3\n3\n2\n1\n2\n3\n1\n0\n2\n3\n2\n1\n3\n3\n0\n2\n0\n2\n3\n0\n0\n0\n0\n1\n1\n0\n1\n2\n1\n1\n2\n3\n3\n0\n6\n2\n4\n2\n6\n2\n0\n1\n2\n2\n3\n3\n0\n3\n1\n1\n2\n1\n3\n4\n2\n1\n1\n3\n9\n3\n1\n1\n0\n1\n0\n0\n1\n2\n2\n0\n2\n2\n0\n2\n0\n1\n0\n1\n2\n1\n0\n1\n0\n8\n3\n1\n6\n2\n4\n2\n0\n2\n1\n2\n0\n0\n3\n2\n0\n0\n2\n0\n0\n1\n1\n2\n" ], [ "def get_vics(summary):\n try:\n# text = re.compile(r'(?<=race of victim s)\\W(black|white|hispanic|hite|asian)(.*?)(male|female)')\n text = re.compile(r'(?<=race of victim s)(.*?)(male|female)')\n result = text.search(summary).group().strip()\n result_s = result.split(' ')\n if len(result_s) > 3:\n return 'error'\n else:\n return result_s\n# if len(result_s) >\n# print(result)\n# number = re.compile(r'\\d+')\n# number_result = number.search(result).group()\n# # print(number_result)\n# return str(number_result) + \" years\"\n# # return 'no' if 'none' in result else 'yes'\n except:\n return ['none_listed']\n\nvic_deets = [get_vics(summary) for summary in df['clean'].values]\ndf['vic_deets'] = vic_deets\n# df['race_vic'] = df.apply()\n\n\n# if len(summary) == 3:\n# summary[0]\nmultiple_vics = [summary[0] if len(summary) == 3 else 'no' for summary in df['vic_deets'].values]\nfemale_vics = ['yes' if 'female' in summary else 'no' for summary in df['vic_deets'].values]\nmale_vics = ['yes' if 'male' in summary else 'no' for summary in df['vic_deets'].values]\n", "_____no_output_____" ], [ "df['multiple_vics'] = multiple_vics\ndf['vic_female'] = female_vics\ndf['vic_male'] = male_vics\ndf", "_____no_output_____" ], [ "race_vics = [summary[1] if len(summary) == 3 else summary[0] for summary in df['vic_deets'].values]", "_____no_output_____" ], [ "race_vics", "_____no_output_____" ], [ "df['race_vic'] = race_vics", "_____no_output_____" ], [ "df", "_____no_output_____" ], [ "def get_age_crime(summary):\n try:\n text = re.compile(r'(?<=age at time of offense )(\\d.*?)\\W')\n result = text.search(summary).group().strip()\n if len(result) < 2:\n return 'none_listed'\n else:\n return result\n except:\n return 'none_listed'\n\nage_crime = [get_age_crime(summary) for summary in df['clean'].values] ", "_____no_output_____" ], [ "age_crime", "_____no_output_____" ], [ "df['age_crime'] = age_crime\ndf", "_____no_output_____" ], [ "def get_weapon(summary):\n try:\n if 'knife' in summary:\n return 'knife'\n# weapon = 'knife'\n elif 'gun' in summary:\n return 'gun'\n# weapon = 'gun'\n elif 'cord ' in summary:\n# print(summary.split('cord')[1])\n return 'cord'\n elif 'blunt object':\n return 'blunt object'\n else:\n return 'other'\n except:\n return 'none_listed'\n\ndf['clean_summary'] = [summary.split('summary')[1] if 'summary' in summary else 'nope' for summary in df['clean'].values]\nweapon = [get_weapon(summary) for summary in df['clean_summary'].values] ", "_____no_output_____" ], [ "len(df[df['clean_summary'] == 'nope'])", "_____no_output_____" ], [ "weapon", "_____no_output_____" ], [ "df['weapon'] = weapon", "_____no_output_____" ], [ "df.columns", "_____no_output_____" ], [ "columns = ['inmate_number','last_name', 'first_name','education_level','age_crime',\n 'occupation','prior_record','multiple_vics','weapon','race_vic','vic_male','vic_female']", "_____no_output_____" ], [ "df1 = pd.DataFrame(df, columns=columns)\ndf1", "_____no_output_____" ], [ "df1.to_csv('V9_photo_inmates.csv')", "_____no_output_____" ], [ "len(df1)", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7c6442e554e5da1771e8056d400d4bb08e4a27

1,154

ipynb

Jupyter Notebook

sell_gap/Part7-Broker-Specific-Steps.ipynb

quantrocket-codeload/sell-gap

96eb93c1307e1a5eb3af0b0ef3b8caaeddd0fc2c

[ "Apache-2.0" ]

3

2020-10-21T14:36:50.000Z

2021-03-15T22:34:37.000Z

sell_gap/Part7-Broker-Specific-Steps.ipynb

quantrocket-codeload/sell-gap

96eb93c1307e1a5eb3af0b0ef3b8caaeddd0fc2c

[ "Apache-2.0" ]

null

sell_gap/Part7-Broker-Specific-Steps.ipynb

quantrocket-codeload/sell-gap

96eb93c1307e1a5eb3af0b0ef3b8caaeddd0fc2c

[ "Apache-2.0" ]

null

22.192308

111

0.569324

[ [ [ "<img alt=\"QuantRocket logo\" src=\"https://www.quantrocket.com/assets/img/notebook-header-logo.png\">\n\n<a href=\"https://www.quantrocket.com/disclaimer/\">Disclaimer</a>", "_____no_output_____" ], [ "# Broker-Specific Steps for Live Trading\n\nTo run the strategy in live trading, follow the setup steps for your broker.", "_____no_output_____" ], [ "* Part 7A: [Alpaca Steps](Part7A-Alpaca-Steps.ipynb)\n* Part 7B: [Interactive Brokers Steps](Part7B-Interactive-Brokers-Steps.ipynb)", "_____no_output_____" ] ] ]

[ "markdown" ]

[ [ "markdown", "markdown", "markdown" ] ]

ec7c64f30936499d18975a28b9f5b0094ff38655

45,918

ipynb

Jupyter Notebook

07_4_mnist_and_gradient_check.ipynb

takaiwai/deep-learning-notes

d55b0c8b33238791a08b5ff0b2ef85bff4496624

[ "MIT" ]

null

07_4_mnist_and_gradient_check.ipynb

takaiwai/deep-learning-notes

d55b0c8b33238791a08b5ff0b2ef85bff4496624

[ "MIT" ]

null

07_4_mnist_and_gradient_check.ipynb

takaiwai/deep-learning-notes

d55b0c8b33238791a08b5ff0b2ef85bff4496624

[ "MIT" ]

null

118.345361

19,294

0.849601

[ [ [ "import numpy as np\nimport matplotlib.pyplot as plt\nimport pickle", "_____no_output_____" ] ], [ [ "# 7-6. Applying for MNIST and Gradient check", "_____no_output_____" ] ], [ [ "log = pickle.load(open('./code/05/log.pkl', 'rb'))", "_____no_output_____" ], [ "log", "_____no_output_____" ], [ "plt.plot(log['loss_train_itr'], log['loss_train'], label='loss_train')\nplt.plot(log['loss_test_itr'], log['loss_test'], label='loss_test')\nplt.legend()\nplt.show()", "_____no_output_____" ], [ "plt.plot(log['accuracy_train_itr'], log['accuracy_train'], label='accuracy_train')\nplt.plot(log['accuracy_test_itr'], log['accuracy_test'], label='accuracy_test')\nplt.legend()\nplt.show()", "_____no_output_____" ] ], [ [ "## Gradient Checking\n\n[Coursera](https://www.coursera.org/learn/deep-neural-network/lecture/htA0l/gradient-checking)\n\n$$\ndifference = \\frac{\\lVert grad - numericalgrad \\rVert_2}{\\lVert grad \\rVert_2 + \\lVert numericalgrad \\rVert_2}\n$$\n\nThreshold: $10^{-7}$", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown" ]

[ [ "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ] ]

ec7c6ededeac6aadabbe136cd2b0b3df8b05cba5

2,442

ipynb

Jupyter Notebook

src/Part1/Quiz1.ipynb

Drogon1573/PyChallenge-Tips

bc6e6ffd45ed9098af50c5258458452b0098efe1

[ "MIT" ]

2

2020-01-12T10:32:03.000Z

2021-10-21T08:25:36.000Z

src/Part1/Quiz1.ipynb

Drogon1573/PyChallenge-Tips

bc6e6ffd45ed9098af50c5258458452b0098efe1

[ "MIT" ]

null

src/Part1/Quiz1.ipynb

Drogon1573/PyChallenge-Tips

bc6e6ffd45ed9098af50c5258458452b0098efe1

[ "MIT" ]

null

28.729412

213

0.586405

[ [ [ "# What about making trans?\n\n[![GitHub license](https://img.shields.io/github/license/Dragon1573/PyChallenge-Tips?color=important&label=Licence&style=flat-square)](https://github.com/Dragon1573/PyChallenge-Tips/blob/master/LICENSE)\n[![Jump to Level 1](https://img.shields.io/badge/Jump%20to-Level%201-blue)](http://www.pythonchallenge.com/pc/def/map.html)\n\n<img src=\"../../resources/imgs/Quiz1.png\" />\n\n*****\n\n&emsp;&emsp;根据图片及金色文字提示，可以发现映射关系符合位移量为2的凯撒密码。借助函数方法对字符进行转换即可。", "_____no_output_____" ] ], [ [ "# 导入小写字母表\nfrom string import ascii_lowercase as letters", "_____no_output_____" ], [ "s = \"g fmnc wms bgblr rpylqjyrc gr zw fylb. rfyrq ufyr amknsrcpq ypc\" + \\\n \" dmp. bmgle gr gl zw fylb gq glcddgagclr ylb rfyr'q ufw rfgq rcvr\" + \\\n \" gq qm jmle. sqgle qrpgle.kyicrpylq() gq pcamkkclbcb. \" + \\\n \"lmu ynnjw ml rfc spj.\"\n# 将字母表整体后移2位，拼接为新的字母表\ndecodeStr = letters[2:] + letters[:2]\ntransTable = str.maketrans(letters, decodeStr)\nprint(s.translate(transTable))", "i hope you didnt translate it by hand. thats what computers are for. doing it in by hand is inefficient and that's why this text is so long. using string.maketrans() is recommended. now apply on the url.\n" ] ], [ [ "> &emsp;&emsp;我希望你没有人工地进行翻译，那是计算机做的事。人工翻译是非常低效的，这也是本文段又臭又长地原因，建议使用`string.maketrans()`。现在，对URL链接做相同地处理。\n\n&emsp;&emsp;将`map`逐字母后移2位，得到`ocr`。下一关链接即为 <http://www.pythonchallenge.com/pc/def/ocr.html> 。", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown" ]

[ [ "markdown" ], [ "code", "code" ], [ "markdown" ] ]

ec7c745b2929ad49d1bc5c2ec5669721472301d6

20,679

ipynb

Jupyter Notebook

examples/asr/notebooks/7_VAD_Offline_Online_Microphone_Demo.ipynb

shangw-nvidia/NeMo

c1dac403184b5d97c1c7b5ca60951265ec4a9fdb

[ "Apache-2.0" ]

1

2021-01-14T13:15:40.000Z

examples/asr/notebooks/7_VAD_Offline_Online_Microphone_Demo.ipynb

shangw-nvidia/NeMo

c1dac403184b5d97c1c7b5ca60951265ec4a9fdb

[ "Apache-2.0" ]

null

examples/asr/notebooks/7_VAD_Offline_Online_Microphone_Demo.ipynb

shangw-nvidia/NeMo

c1dac403184b5d97c1c7b5ca60951265ec4a9fdb

[ "Apache-2.0" ]

null

32.361502

334

0.537937

[ [ [ "This notebook demonstrates voice activity detection from a microphone's stream (online) and a given wav file (offline) in NeMo.", "_____no_output_____" ], [ "The notebook requires PyAudio library to get a signal from an audio device.\nFor Ubuntu, please run the following commands to install it:\n```\nsudo apt-get install -y portaudio19-dev\npip install pyaudio\n```", "_____no_output_____" ] ], [ [ "import os\nimport nemo\nimport nemo.collections.asr as nemo_asr\nimport numpy as np\nimport pyaudio as pa\nimport time\n\nimport librosa\nimport IPython.display as ipd\nimport matplotlib.pyplot as plt\n%matplotlib inline", "_____no_output_____" ] ], [ [ "# Model Architecture and Weights\n\nThe model architecture is defined in a YAML file available in the config directory. MatchboxNet 3x1x64 has been trained on the [Google Speech Commands v2 dataset](https://arxiv.org/abs/1804.03209) and [freesound](https://freesound.org), and these weights are available on NGC. They will automatically be downloaded if not found.", "_____no_output_____" ] ], [ [ "MODEL_YAML = '../configs/quartznet_vad_3x1.yaml'", "_____no_output_____" ], [ "# Download the checkpoint files\n\nbase_checkpoint_path = './checkpoints/matchboxnet_3x1x1/'\nCHECKPOINT_ENCODER = os.path.join(base_checkpoint_path, 'JasperEncoder-STEP-90800.pt')\nCHECKPOINT_DECODER = os.path.join(base_checkpoint_path, 'JasperDecoderForClassification-STEP-90800.pt')\n\nif not os.path.exists(base_checkpoint_path):\n os.makedirs(base_checkpoint_path)\n \nif not os.path.exists(CHECKPOINT_ENCODER):\n !wget https://api.ngc.nvidia.com/v2/models/nvidia/vad_matchboxnet_3x1x1/versions/1/files/JasperEncoder-STEP-90800.pt -P {base_checkpoint_path};\nif not os.path.exists(CHECKPOINT_DECODER):\n !wget https://api.ngc.nvidia.com/v2/models/nvidia/vad_matchboxnet_3x1x1/versions/1/files/JasperDecoderForClassification-STEP-90800.pt -P {base_checkpoint_path};\n ", "_____no_output_____" ] ], [ [ "# Construct the Neural Modules and the eval graph", "_____no_output_____" ] ], [ [ "from ruamel.yaml import YAML\nyaml = YAML(typ=\"safe\")\nwith open(MODEL_YAML) as f:\n model_definition = yaml.load(f)", "_____no_output_____" ], [ "neural_factory = nemo.core.NeuralModuleFactory(\n placement=nemo.core.DeviceType.GPU,\n backend=nemo.core.Backend.PyTorch)", "_____no_output_____" ] ], [ [ "## Define a Neural Module to iterate over audio\n\nHere we define a custom Neural Module which acts as an iterator over a stream of audio that is supplied to it. ", "_____no_output_____" ] ], [ [ "from nemo.backends.pytorch.nm import DataLayerNM\nfrom nemo.core.neural_types import NeuralType, AudioSignal, LengthsType\nimport torch\n\n# simple data layer to pass audio signal\nclass AudioDataLayer(DataLayerNM):\n @property\n def output_ports(self):\n return {\n 'audio_signal': NeuralType(('B', 'T'), AudioSignal(freq=self._sample_rate)),\n 'a_sig_length': NeuralType(tuple('B'), LengthsType()),\n }\n\n def __init__(self, sample_rate):\n super().__init__()\n self._sample_rate = sample_rate\n self.output = True\n \n def __iter__(self):\n return self\n \n def __next__(self):\n if not self.output:\n raise StopIteration\n self.output = False\n return torch.as_tensor(self.signal, dtype=torch.float32), \\\n torch.as_tensor(self.signal_shape, dtype=torch.int64)\n \n def set_signal(self, signal):\n self.signal = np.reshape(signal.astype(np.float32)/32768., [1, -1])\n self.signal_shape = np.expand_dims(self.signal.size, 0).astype(np.int64)\n self.output = True\n\n def __len__(self):\n return 1\n\n @property\n def dataset(self):\n return None\n\n @property\n def data_iterator(self):\n return self", "_____no_output_____" ] ], [ [ "## Instantiate the Neural Modules\n\nWe now instantiate the neural modules and the encoder and decoder, set the weights of these models with the downloaded pretrained weights and construct the DAG to evaluate MatchboxNet on audio streams", "_____no_output_____" ] ], [ [ "# Instantiate necessary neural modules\ndata_layer = AudioDataLayer(sample_rate=model_definition['sample_rate'])\n\ndata_preprocessor = nemo_asr.AudioToMFCCPreprocessor(\n **model_definition['AudioToMFCCPreprocessor'])\n\njasper_encoder = nemo_asr.JasperEncoder(\n **model_definition['JasperEncoder'])\n\njasper_decoder = nemo_asr.JasperDecoderForClassification(\n feat_in=model_definition['JasperEncoder']['jasper'][-1]['filters'],\n num_classes=len(model_definition['labels']))\n\n# load pre-trained model\njasper_encoder.restore_from(CHECKPOINT_ENCODER)\njasper_decoder.restore_from(CHECKPOINT_DECODER)\n\n# Define inference DAG\naudio_signal, audio_signal_len = data_layer()\nprocessed_signal, processed_signal_len = data_preprocessor(\n input_signal=audio_signal,\n length=audio_signal_len)\nencoded, encoded_len = jasper_encoder(audio_signal=processed_signal,\n length=processed_signal_len)\nlog_probs = jasper_decoder(encoder_output=encoded)\n\n# inference method for audio signal (single instance)\ndef infer_signal(self, signal):\n data_layer.set_signal(signal)\n tensors = self.infer([log_probs], verbose=False)\n logits = tensors[0][0]\n return logits\n\nneural_factory.infer_signal = infer_signal.__get__(neural_factory)", "_____no_output_____" ] ], [ [ "# FrameASR: Helper class for streaming inference\nHere we adopt FrameASR for streaming inference for voice activatity detection", "_____no_output_____" ] ], [ [ "# class for streaming frame-based ASR\n# 1) use reset() method to reset FrameASR's state\n# 2) call transcribe(frame) to do ASR on\n# contiguous signal's frames\nclass FrameASR:\n \n def __init__(self, neural_factory, model_definition,\n frame_len=2, frame_overlap=2.5, \n offset=10):\n '''\n Args:\n frame_len: frame's duration, seconds\n frame_overlap: duration of overlaps before and after current frame, seconds\n offset: number of symbols to drop for smooth streaming\n '''\n self.vocab = list(model_definition['labels'])\n self.vocab.append('_')\n \n self.sr = model_definition['sample_rate']\n self.frame_len = frame_len\n self.n_frame_len = int(frame_len * self.sr)\n self.frame_overlap = frame_overlap\n self.n_frame_overlap = int(frame_overlap * self.sr)\n timestep_duration = model_definition['AudioToMFCCPreprocessor']['window_stride']\n for block in model_definition['JasperEncoder']['jasper']:\n timestep_duration *= block['stride'][0] ** block['repeat']\n self.buffer = np.zeros(shape=2*self.n_frame_overlap + self.n_frame_len,\n dtype=np.float32)\n self.offset = offset\n self.reset()\n \n def _decode(self, frame, offset=0):\n assert len(frame)==self.n_frame_len\n self.buffer[:-self.n_frame_len] = self.buffer[self.n_frame_len:]\n self.buffer[-self.n_frame_len:] = frame\n logits = neural_factory.infer_signal(self.buffer).to('cpu').numpy()[0]\n decoded = self._greedy_decoder(\n logits, \n self.vocab\n )\n return decoded[:len(decoded)-offset]\n \n def transcribe(self, frame=None):\n if frame is None:\n frame = np.zeros(shape=self.n_frame_len, dtype=np.float32)\n if len(frame) < self.n_frame_len:\n frame = np.pad(frame, [0, self.n_frame_len - len(frame)], 'constant')\n unmerged = self._decode(frame, self.offset)\n return unmerged\n \n def reset(self):\n '''\n Reset frame_history and decoder's state\n '''\n self.buffer=np.zeros(shape=self.buffer.shape, dtype=np.float32)\n self.prev_char = ''\n\n @staticmethod\n def _greedy_decoder(logits, vocab):\n s = ''\n s = []\n if logits.shape[0]:\n probs = torch.softmax(torch.as_tensor(logits), dim=-1)\n probas, preds = torch.max(probs, dim=-1)\n s = [preds.item(), str(vocab[preds]), probs[0].item(), probs[1].item(), str(logits)]\n return s", "_____no_output_____" ] ], [ [ "## What classes can this model recognize?\n\nBefore we begin inference on the actual audio stream, lets look at what are the classes this model was trained to recognize", "_____no_output_____" ] ], [ [ "labels = model_definition['labels']\nprint(labels)", "_____no_output_____" ] ], [ [ "# Listening to audio stream and perform inference using FrameASR", "_____no_output_____" ], [ "## Offline Inference", "_____no_output_____" ], [ "You can experiment with differents **STEP** and **WINDOW_SIZE** for streaming VAD inference.", "_____no_output_____" ] ], [ [ "STEP_LIST = [0.01, 0.01, 0.01]\nWINDOW_SIZE_LIST = [0.25, 0.20, 0.15]", "_____no_output_____" ], [ "import wave\n\ndef offline_inference(wave_file, STEP = 0.025, WINDOW_SIZE = 0.5):\n \n FRAME_LEN = STEP # infer every STEP seconds \n CHANNELS = 1 # number of audio channels (expect mono signal)\n RATE = 16000 # sample rate, Hz\n \n CHUNK_SIZE = int(FRAME_LEN*RATE)\n asr = FrameASR(neural_factory, model_definition,\n frame_len=FRAME_LEN, frame_overlap = (WINDOW_SIZE-FRAME_LEN)/2,\n offset=0)\n\n wf = wave.open(wave_file, 'rb')\n p = pa.PyAudio()\n\n empty_counter = 0\n\n preds = []\n proba_b = []\n proba_s = []\n \n def callback(in_data, frame_count, time_info, status):\n data = wf.readframes(frame_count)\n global empty_counter\n signal = np.frombuffer(data, dtype=np.int16)\n result = asr.transcribe(signal)\n\n preds.append(result[0])\n proba_b.append(result[2])\n proba_s.append(result[3])\n if len(result):\n print(result,end='\\n')\n empty_counter = 3\n elif empty_counter > 0:\n empty_counter -= 1\n if empty_counter == 0:\n print(' ',end='')\n\n return (data, pa.paContinue)\n\n stream = p.open(format=p.get_format_from_width(wf.getsampwidth()),\n channels=CHANNELS,\n rate=RATE,\n output = True,\n stream_callback=callback,\n frames_per_buffer=CHUNK_SIZE) # Specifies the number of frames per buffer.\n \n stream.start_stream()\n\n while stream.is_active():\n time.sleep(0.1)\n\n stream.stop_stream()\n stream.close()\n p.terminate()\n\n asr.reset()\n return preds, proba_b, proba_s", "_____no_output_____" ] ], [ [ "### Here we show an example of offline streaming inference\nYou can use your file or download the provided toy dataset. ", "_____no_output_____" ] ], [ [ "toy_data = './vad'\nif not os.path.exists(toy_data):\n !wget -c \"https://github.com/NVIDIA/NeMo/blob/master/tests/data/vad.tar.xz?raw=true\" -O vad.tar.xz \n !tar -xvf vad.tar.xz", "_____no_output_____" ], [ "wave_file = './vad/welcome_noisy.wav'\nCHANNELS = 1\nRATE = 16000\naudio, sample_rate = librosa.load(wave_file, sr=RATE)", "_____no_output_____" ], [ "results = []\nfor STEP, WINDOW_SIZE in zip(STEP_LIST, WINDOW_SIZE_LIST):\n print(f'====== STEP is {STEP}s, WINDOW_SIZE is {WINDOW_SIZE}s ====== ')\n preds, proba_b, proba_s = offline_inference(wave_file, STEP, WINDOW_SIZE)\n results.append([STEP, WINDOW_SIZE, preds, proba_b, proba_s])", "_____no_output_____" ], [ "import matplotlib.pyplot as plt\nfrom pylab import *\nimport numpy as np\nimport librosa.display\nplt.figure(figsize=[16,10])\nplt.title('Audio, Preictions and Probas')\nplt.rcParams.update({'font.size': 10, 'font.family': 'sans-serif'})\nsubplots_adjust(hspace=2.00)\n\n\nFRAME_LEN = STEP_LIST[0]\nlen_pred = len(results[0][2]) \n\nnum = len(results)\nfor i,v in enumerate(range(num + 1)):\n v = v + 1\n if v > len(results):\n\n ax = plt.subplot(num + 2, 1, v)\n S = librosa.feature.melspectrogram(y=audio, sr=sample_rate, n_mels=128,\n fmax=8000)\n S_dB = librosa.power_to_db(S, ref=np.max)\n librosa.display.specshow(S_dB, x_axis='time', y_axis='mel', \n sr=sample_rate, fmax=8000)\n ax.set_title('Mel-frequency spectrogram')\n ax.grid()\n\n ax = plt.subplot(num + 2, 1, v + 1)\n ax.plot(np.arange(audio.size) / sample_rate, audio, 'b')\n ax.set_xlim([-0.01, len_pred * FRAME_LEN])\n ax.set_ylabel('Signal')\n ax.set_xlabel('Time, seconds')\n ax.set_title(f'File: {str(wave_file)}')\n ax.set_ylim([-0.5, 0.5])\n ax.grid()\n else:\n ax = plt.subplot(num + 2, 1, v)\n ax.plot(results[i][2], 'r', label='pred')\n ax.plot(results[i][3], 'g--', label='prob for background')\n ax.plot(results[i][4], 'b--', label='prob for speech')\n ax.set_xlim([0, len_pred])\n ax.set_title(f'step {results[i][0]}s, buffer size {results[i][1]}s')\n ax.set_ylabel('Preds and Probas')\n ax.set_xlabel('Segments')\n ax.grid()\n legend = ax.legend(loc='lower left', shadow=True)\nplt.show()\n", "_____no_output_____" ], [ "import librosa\nipd.Audio(audio, rate=sample_rate)", "_____no_output_____" ] ], [ [ "## Online inference through microphone", "_____no_output_____" ] ], [ [ "STEP = 0.01 \nWINDOW_SIZE = 0.20\nCHANNELS = 1 \nRATE = 16000\n\nCHUNK_SIZE = int(STEP * RATE)\nasr = FrameASR(neural_factory, model_definition,\n frame_len=STEP, frame_overlap=(WINDOW_SIZE - FRAME_LEN) / 2, \n offset=0)", "_____no_output_____" ], [ "p = pa.PyAudio()\nprint('Available audio input devices:')\nfor i in range(p.get_device_count()):\n dev = p.get_device_info_by_index(i)\n if dev.get('maxInputChannels'):\n print(i, dev.get('name'))\nprint('Please type input device ID:')\ndev_idx = int(input())\n\nempty_counter = 0\n\ndef callback(in_data, frame_count, time_info, status):\n global empty_counter\n signal = np.frombuffer(in_data, dtype=np.int16)\n text = asr.transcribe(signal)\n if len(text):\n print(text,end='\\n')\n empty_counter = 3\n elif empty_counter > 0:\n empty_counter -= 1\n if empty_counter == 0:\n print(' ',end='')\n return (in_data, pa.paContinue)\n\nstream = p.open(format=pa.paInt16,\n channels=CHANNELS,\n rate=RATE,\n input=True,\n input_device_index=dev_idx,\n stream_callback=callback,\n frames_per_buffer=CHUNK_SIZE)\n\nprint('Listening...')\n\nstream.start_stream()\n\nwhile stream.is_active():\n time.sleep(0.1)", "_____no_output_____" ], [ "stream.stop_stream()\nstream.close()\np.terminate()", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ] ]

ec7c876786b14f6f7f2096e82278a9ac55872121

22,166

ipynb

Jupyter Notebook

module3-cross-validation/LS_DS_223_assignment.ipynb

zack-murray/DS-Unit-2-Kaggle-Challenge

ffe6d051446df992b7e82480c72bb1677a527d6c

[ "MIT" ]

null

module3-cross-validation/LS_DS_223_assignment.ipynb

zack-murray/DS-Unit-2-Kaggle-Challenge

ffe6d051446df992b7e82480c72bb1677a527d6c

[ "MIT" ]

null

module3-cross-validation/LS_DS_223_assignment.ipynb

zack-murray/DS-Unit-2-Kaggle-Challenge

ffe6d051446df992b7e82480c72bb1677a527d6c

[ "MIT" ]

null

44.870445

452

0.561581

[ [ [ "<a href=\"https://colab.research.google.com/github/zack-murray/DS-Unit-2-Kaggle-Challenge/blob/master/module3-cross-validation/LS_DS_223_assignment.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>", "_____no_output_____" ], [ "Lambda School Data Science\n\n*Unit 2, Sprint 2, Module 3*\n\n---", "_____no_output_____" ], [ "# Cross-Validation\n\n\n## Assignment\n- [ ] [Review requirements for your portfolio project](https://lambdaschool.github.io/ds/unit2), then submit your dataset.\n- [ ] Continue to participate in our Kaggle challenge. \n- [ ] Use scikit-learn for hyperparameter optimization with RandomizedSearchCV.\n- [ ] Submit your predictions to our Kaggle competition. (Go to our Kaggle InClass competition webpage. Use the blue **Submit Predictions** button to upload your CSV file. Or you can use the Kaggle API to submit your predictions.)\n- [ ] Commit your notebook to your fork of the GitHub repo.\n\n\nYou won't be able to just copy from the lesson notebook to this assignment.\n\n- Because the lesson was ***regression***, but the assignment is ***classification.***\n- Because the lesson used [TargetEncoder](https://contrib.scikit-learn.org/categorical-encoding/targetencoder.html), which doesn't work as-is for _multi-class_ classification.\n\nSo you will have to adapt the example, which is good real-world practice.\n\n1. Use a model for classification, such as [RandomForestClassifier](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestClassifier.html)\n2. Use hyperparameters that match the classifier, such as `randomforestclassifier__ ...`\n3. Use a metric for classification, such as [`scoring='accuracy'`](https://scikit-learn.org/stable/modules/model_evaluation.html#common-cases-predefined-values)\n4. If you’re doing a multi-class classification problem — such as whether a waterpump is functional, functional needs repair, or nonfunctional — then use a categorical encoding that works for multi-class classification, such as [OrdinalEncoder](https://contrib.scikit-learn.org/categorical-encoding/ordinal.html) (not [TargetEncoder](https://contrib.scikit-learn.org/categorical-encoding/targetencoder.html))\n\n\n\n## Stretch Goals\n\n### Reading\n- Jake VanderPlas, [Python Data Science Handbook, Chapter 5.3](https://jakevdp.github.io/PythonDataScienceHandbook/05.03-hyperparameters-and-model-validation.html), Hyperparameters and Model Validation\n- Jake VanderPlas, [Statistics for Hackers](https://speakerdeck.com/jakevdp/statistics-for-hackers?slide=107)\n- Ron Zacharski, [A Programmer's Guide to Data Mining, Chapter 5](http://guidetodatamining.com/chapter5/), 10-fold cross validation\n- Sebastian Raschka, [A Basic Pipeline and Grid Search Setup](https://github.com/rasbt/python-machine-learning-book/blob/master/code/bonus/svm_iris_pipeline_and_gridsearch.ipynb)\n- Peter Worcester, [A Comparison of Grid Search and Randomized Search Using Scikit Learn](https://blog.usejournal.com/a-comparison-of-grid-search-and-randomized-search-using-scikit-learn-29823179bc85)\n\n### Doing\n- Add your own stretch goals!\n- Try other [categorical encodings](https://contrib.scikit-learn.org/categorical-encoding/). See the previous assignment notebook for details.\n- In additon to `RandomizedSearchCV`, scikit-learn has [`GridSearchCV`](https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.GridSearchCV.html). Another library called scikit-optimize has [`BayesSearchCV`](https://scikit-optimize.github.io/notebooks/sklearn-gridsearchcv-replacement.html). Experiment with these alternatives.\n- _[Introduction to Machine Learning with Python](http://shop.oreilly.com/product/0636920030515.do)_ discusses options for \"Grid-Searching Which Model To Use\" in Chapter 6:\n\n> You can even go further in combining GridSearchCV and Pipeline: it is also possible to search over the actual steps being performed in the pipeline (say whether to use StandardScaler or MinMaxScaler). This leads to an even bigger search space and should be considered carefully. Trying all possible solutions is usually not a viable machine learning strategy. However, here is an example comparing a RandomForestClassifier and an SVC ...\n\nThe example is shown in [the accompanying notebook](https://github.com/amueller/introduction_to_ml_with_python/blob/master/06-algorithm-chains-and-pipelines.ipynb), code cells 35-37. Could you apply this concept to your own pipelines?\n", "_____no_output_____" ], [ "### BONUS: Stacking!\n\nHere's some code you can use to \"stack\" multiple submissions, which is another form of ensembling:\n\n```python\nimport pandas as pd\n\n# Filenames of your submissions you want to ensemble\nfiles = ['submission-01.csv', 'submission-02.csv', 'submission-03.csv']\n\ntarget = 'status_group'\nsubmissions = (pd.read_csv(file)[[target]] for file in files)\nensemble = pd.concat(submissions, axis='columns')\nmajority_vote = ensemble.mode(axis='columns')[0]\n\nsample_submission = pd.read_csv('sample_submission.csv')\nsubmission = sample_submission.copy()\nsubmission[target] = majority_vote\nsubmission.to_csv('my-ultimate-ensemble-submission.csv', index=False)\n```", "_____no_output_____" ] ], [ [ "%%capture\nimport sys\n\n# If you're on Colab:\nif 'google.colab' in sys.modules:\n DATA_PATH = 'https://raw.githubusercontent.com/LambdaSchool/DS-Unit-2-Kaggle-Challenge/master/data/'\n !pip install category_encoders==2.*\n\n# If you're working locally:\nelse:\n DATA_PATH = '../data/'", "_____no_output_____" ], [ "pip install category_encoders", "Collecting category_encoders\n Downloading category_encoders-2.1.0-py2.py3-none-any.whl (100 kB)\nRequirement already satisfied: patsy>=0.4.1 in e:\\anaconda\\lib\\site-packages (from category_encoders) (0.5.1)\nRequirement already satisfied: scikit-learn>=0.20.0 in e:\\anaconda\\lib\\site-packages (from category_encoders) (0.22.1)\nRequirement already satisfied: numpy>=1.11.3 in e:\\anaconda\\lib\\site-packages (from category_encoders) (1.18.1)\nRequirement already satisfied: pandas>=0.21.1 in e:\\anaconda\\lib\\site-packages (from category_encoders) (1.0.1)\nRequirement already satisfied: scipy>=0.19.0 in e:\\anaconda\\lib\\site-packages (from category_encoders) (1.4.1)\nRequirement already satisfied: statsmodels>=0.6.1 in e:\\anaconda\\lib\\site-packages (from category_encoders) (0.11.0)\nRequirement already satisfied: six in e:\\anaconda\\lib\\site-packages (from patsy>=0.4.1->category_encoders) (1.14.0)\nRequirement already satisfied: joblib>=0.11 in e:\\anaconda\\lib\\site-packages (from scikit-learn>=0.20.0->category_encoders) (0.14.1)\nRequirement already satisfied: python-dateutil>=2.6.1 in e:\\anaconda\\lib\\site-packages (from pandas>=0.21.1->category_encoders) (2.8.1)\nRequirement already satisfied: pytz>=2017.2 in e:\\anaconda\\lib\\site-packages (from pandas>=0.21.1->category_encoders) (2019.3)\nInstalling collected packages: category-encoders\nSuccessfully installed category-encoders-2.1.0\nNote: you may need to restart the kernel to use updated packages.\n" ], [ "import pandas as pd\nfrom sklearn.model_selection import train_test_split\n\n# Merge train_features.csv & train_labels.csv\ntrain = pd.merge(pd.read_csv(r'C:\\Users\\Z Dubs\\Waterpumps\\train_features.csv'), \n pd.read_csv(r'C:\\Users\\Z Dubs\\Waterpumps\\train_labels.csv'))\n\n# Read test_features.csv & sample_submission.csv\ntest = pd.read_csv(r'C:\\Users\\Z Dubs\\Waterpumps\\test_features.csv')\nsample_submission = pd.read_csv(r'C:\\Users\\Z Dubs\\Waterpumps\\sample_submission.csv')\n\ntrain.shape, test.shape", "_____no_output_____" ], [ "# Split training set into train & val\ntrain, val = train_test_split(train, train_size=.75, stratify=train['status_group'], \n random_state=22)\n\ntrain.shape, val.shape, test.shape", "_____no_output_____" ], [ "import numpy as np\n# Define a function to wrangle train, validate, and test sets in the same way\n\ndef wrangle(X):\n \"\"\"Wrangle train, validate, and test sets in the same way\"\"\"\n \n # Prevent SettingWithCopyWarning\n X = X.copy()\n \n # About 3% of the time, latitude has small values near zero,\n # outside Tanzania, so we'll treat these values like zero.\n X['latitude'] = X['latitude'].replace(-2e-08, 0)\n \n # When columns have zeros and shouldn't, they are like null values.\n # So we will replace the zeros with nulls, and impute missing values later.\n cols_with_zeros = ['longitude', 'latitude', 'construction_year', \n 'gps_height', 'population']\n for col in cols_with_zeros:\n X[col] = X[col].replace(0, np.nan)\n X[col+'_MISSING'] = X[col].isnull()\n \n # quantity & quantity_group are duplicates, so drop one\n X = X.drop(columns='quantity_group')\n # drop more columns with striking similarities\n X = X.drop(columns=['extraction_type_group', 'extraction_type_class', \n 'payment_type', 'quality_group', 'source_type', 'source_class',\n 'region_code', 'district_code', 'waterpoint_type_group'])\n # drop ambiguous columns that look to add little value\n X = X.drop(columns=['num_private', 'lga', 'ward', 'recorded_by', 'id'])\n # return the wrangled dataframe\n return X\n\ntrain = wrangle(train)\nval = wrangle(val)\ntest = wrangle(test)", "_____no_output_____" ], [ "# The status_group column is the target\ntarget = 'status_group'\n\n# Get a dataframe with all train columns except the target & id\ntrain_features = train.drop(columns=[target])\n\n# Get a list of the numeric features\nnumeric_features = train_features.select_dtypes(include='number').columns.tolist()\n\n# Get a series with the cardinality of the nonnumeric features\ncardinality = train_features.select_dtypes(exclude='number').nunique()\n\ntrain_features.select_dtypes(exclude='number').nunique()\n# Get a list of all categorical features with cardinality <= 50\ncategorical_features = cardinality[cardinality <= 50].index.tolist()\n\n# Combine the lists to define our features\nfeatures = numeric_features + categorical_features\nprint(features)", "['amount_tsh', 'gps_height', 'longitude', 'latitude', 'population', 'construction_year', 'basin', 'region', 'public_meeting', 'scheme_management', 'permit', 'extraction_type', 'management', 'management_group', 'payment', 'water_quality', 'quantity', 'source', 'waterpoint_type', 'longitude_MISSING', 'latitude_MISSING', 'construction_year_MISSING', 'gps_height_MISSING', 'population_MISSING']\n" ], [ "# Arrange data into X features matrix and y target vector \nX_train = train[features]\ny_train = train[target]\nX_val = val[features]\ny_val = val[target]\nX_test = test[features]", "_____no_output_____" ], [ "import category_encoders as ce\nfrom sklearn.impute import SimpleImputer\nfrom sklearn.ensemble import RandomForestClassifier\nfrom sklearn.pipeline import make_pipeline\nfrom sklearn.model_selection import GridSearchCV, RandomizedSearchCV\nfrom scipy.stats import randint, uniform\n\npipeline = make_pipeline(\n ce.OrdinalEncoder(),\n SimpleImputer(),\n RandomForestClassifier(random_state=7))\n\nparam_distributions = {\n 'simpleimputer__strategy': ['mean', 'median'], \n 'randomforestclassifier__n_estimators': randint(50, 500), \n 'randomforestclassifier__max_depth': [10, 20, 30, 40, None], \n 'randomforestclassifier__max_features': uniform(0, 1), \n 'randomforestclassifier__min_samples_split': [1, 2, 3, 4],\n 'randomforestclassifier__min_samples_leaf': [1, 2, 3, 4],\n}\n\n# If you're on Colab, decrease n_iter & cv parameters\nsearch = RandomizedSearchCV(\n pipeline, \n param_distributions=param_distributions, \n n_iter=100, \n cv=5, \n scoring='accuracy', \n verbose=10, \n return_train_score=True, \n n_jobs=-1\n)\n\nsearch.fit(X_train, y_train);", "Fitting 5 folds for each of 100 candidates, totalling 500 fits\n" ], [ "print('Best hyperparameters', search.best_params_)\nprint('Cross-validation Accuracy', -search.best_score_)", "Best hyperparameters {'randomforestclassifier__max_depth': 40, 'randomforestclassifier__max_features': 0.17864439372232666, 'randomforestclassifier__min_samples_leaf': 2, 'randomforestclassifier__min_samples_split': 3, 'randomforestclassifier__n_estimators': 226, 'simpleimputer__strategy': 'median'}\nCross-validation Accuracy -0.8024691358024691\n" ] ] ]

[ "markdown", "code" ]

[ [ "markdown", "markdown", "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7c899d7b5a15181203ce550196e04892362aca

89,234

ipynb

Jupyter Notebook

Documentation/check/RDcheck.ipynb

dkaramit/MiMeS

a3c97a4877f181b54e880d7b144271c5659291b5

[ "MIT" ]

2

2022-01-27T20:10:19.000Z

2022-01-29T04:26:16.000Z

Documentation/check/RDcheck.ipynb

dkaramit/MiMeS

a3c97a4877f181b54e880d7b144271c5659291b5

[ "MIT" ]

null

Documentation/check/RDcheck.ipynb

dkaramit/MiMeS

a3c97a4877f181b54e880d7b144271c5659291b5

[ "MIT" ]

null

79.17835

39,823

0.695665

[ [ [ "import os\nos.chdir('..')\n\nimport numpy as np\n# import FT as FT\n\n\nimport matplotlib\n#matplotlib.use('WebAgg')\n#matplotlib.use('Qt4Cairo')\n#matplotlib.use('Qt5Cairo')\nmatplotlib.use('nbAgg')\nimport matplotlib.pyplot as plt\n\nplt.rcParams['font.size']=16\nplt.rcParams['font.family']='dejavu sans'\n\nplt.rcParams['mathtext.fontset']='stix'\nplt.rcParams['mathtext.rm']='custom'\nplt.rcParams['mathtext.it']='stix:italic'\nplt.rcParams['mathtext.bf']='stix:bold'\n#-------------------------#\nos.chdir('check')\n", "_____no_output_____" ], [ "#load the module\nfrom sys import path as sysPath\nfrom os import path as osPath\nsysPath.append(osPath.join(osPath.dirname('./'), '../../src'))\n\nfrom interfacePy.FT import FT #easy tick formatting", "_____no_output_____" ], [ "Lattice_theta=np.loadtxt('./1606.07494/fa_vs_theta.dat')#data from 1606.07494\n\nAxionCosmoRev=np.loadtxt('./0910.1066/theta_vs_fa.dat')#data from 0910.1066 \n\nGonVis=np.loadtxt('./0912.0015/RDline.dat')#data from 0912.0015 \n\nMiMeS=np.loadtxt('./MiMeS/RDAxion.dat')\n\n", "_____no_output_____" ], [ "if True:\n fig=plt.figure(figsize=(9,4))\n fig.subplots_adjust(bottom=0.15, left=0.15, top = 0.95, right=0.9,wspace=0.0,hspace=0.0)\n sub = fig.add_subplot(1,1,1)\n \n sub.plot(MiMeS[:,1],MiMeS[:,0],linestyle='-',linewidth=2.5,alpha=1,c='xkcd:red',label=r\"MiMeS\")\n \n sub.plot(Lattice_theta[:,0],Lattice_theta[:,1],linestyle=(1,(2,5)),linewidth=2,alpha=1,c='xkcd:black',label=r\"1606.07494\")\n\n sub.plot(AxionCosmoRev[:,1],AxionCosmoRev[:,0],linewidth=2,alpha=1,linestyle=(1,(4,10)),c='xkcd:blue',label=r\"0910.1066\")\n \n sub.plot(GonVis[:,0],GonVis[:,1]*np.pi,linewidth=2,alpha=1,linestyle=(1,(4,10)),c='xkcd:gray',label=r\"0912.0015\")\n \n \n sub.set_xlabel(r'$f_\\alpha ~[{\\rm GeV}]$')\n sub.xaxis.set_label_coords(0.5, -0.15) \n sub.set_ylabel(r'$ \\theta_{\\rm ini}$')\n sub.yaxis.set_label_coords(-0.1,0.5) \n sub.set_yscale('log')\n sub.set_xscale('log')\n \n sub.set_xlim(1e10,1e20)\n sub.set_ylim(1e-5,10)\n\n sub.legend(bbox_to_anchor=(0.02, 0.02),borderaxespad=0., \n borderpad=0,ncol=1,loc='lower left',fontsize=10,framealpha=0)\n #set major ticks\n _M_xticks=[ 10.**i for i in range(9,21) ]\n _M_yticks=[ 10.**i for i in range(-5,2) ]\n\n #set major ticks that will not have a label\n _M_xticks_exception=[]\n _M_yticks_exception=[]\n\n _m_xticks=[]\n _m_yticks=[] \n ft=FT(_M_xticks,_M_yticks,\n _M_xticks_exception,_M_yticks_exception,\n _m_xticks,_m_yticks,\n xmin=1e10,xmax=1e20,ymin=1e-5,ymax=10,xscale='log',yscale='log')\n\n ft.format_ticks(plt,sub) \n\n \n fig.savefig('RD_fa_vs_theta_i_check.pdf',bbox_inches='tight')\n\n fig.show()", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code" ] ]

ec7c9eee4f419e1cfcf5290a5c9c4c8feaf91142

24,540

ipynb

Jupyter Notebook

docs_src/tutorial.itemlist.ipynb

VictorXunS/fastai

ba401da9e93f4bee3595754abe1b7c14175c8c27

[ "Apache-2.0" ]

1

2019-04-08T09:52:28.000Z

docs_src/tutorial.itemlist.ipynb

gianfa/fastai

017c8b11c9996849337aad348c8609842792b5c3

[ "Apache-2.0" ]

null

docs_src/tutorial.itemlist.ipynb

gianfa/fastai

017c8b11c9996849337aad348c8609842792b5c3

[ "Apache-2.0" ]

1

2020-05-19T12:56:20.000Z

44.456522

578

0.627262

[ [ [ "## Customizing datasets in fastai", "_____no_output_____" ] ], [ [ "from fastai.gen_doc.nbdoc import *\nfrom fastai.vision import *", "_____no_output_____" ] ], [ [ "In this tutorial, we'll see how to create custom subclasses of [`ItemBase`](/core.html#ItemBase) or [`ItemList`](/data_block.html#ItemList) while retaining everything the fastai library has to offer. To allow basic functions to work consistently across various applications, the fastai library delegates several tasks to one of those specific objects, and we'll see here which methods you have to implement to be able to have everything work properly. But first let's take a step back to see where you'll use your end result.", "_____no_output_____" ], [ "## Links with the data block API", "_____no_output_____" ], [ "The data block API works by allowing you to pick a class that is responsible to get your items and another class that is charged with getting your targets. Combined together, they create a pytorch [`Dataset`](https://pytorch.org/docs/stable/data.html#torch.utils.data.Dataset) that is then wrapped inside a [`DataLoader`](https://pytorch.org/docs/stable/data.html#torch.utils.data.DataLoader). The training set, validation set and maybe test set are then all put in a [`DataBunch`](/basic_data.html#DataBunch).\n\nThe data block API allows you to mix and match what class your inputs have, what class your targets have, how to do the split between train and validation set, then how to create the [`DataBunch`](/basic_data.html#DataBunch), but if you have a very specific kind of input/target, the fastai classes might no be sufficient to you. This tutorial is there to explain what is needed to create a new class of items and what methods are important to implement or override.\n\nIt goes in two phases: first we focus on what you need to create a custom [`ItemBase`](/core.html#ItemBase) class (which is the type of your inputs/targets) then on how to create your custom [`ItemList`](/data_block.html#ItemList) (which is basically a set of [`ItemBase`](/core.html#ItemBase)) while highlighting which methods are called by the library.", "_____no_output_____" ], [ "## Creating a custom [`ItemBase`](/core.html#ItemBase) subclass", "_____no_output_____" ], [ "The fastai library contains three basic types of [`ItemBase`](/core.html#ItemBase) that you might want to subclass:\n- [`Image`](/vision.image.html#Image) for vision applications\n- [`Text`](/text.data.html#Text) for text applications\n- [`TabularLine`](/tabular.data.html#TabularLine) for tabular applications\n\nWhether you decide to create your own item class or to subclass one of the above, here is what you need to implement:", "_____no_output_____" ], [ "### Basic attributes", "_____no_output_____" ], [ "Those are the more important attributes your custom [`ItemBase`](/core.html#ItemBase) needs as they're used everywhere in the fastai library:\n- `ItemBase.data` is the thing that is passed to pytorch when you want to create a [`DataLoader`](https://pytorch.org/docs/stable/data.html#torch.utils.data.DataLoader). This is what needs to be fed to your model. Note that it might be different from the representation of your item since you might want something that is more understandable.\n- `__str__` representation: if applicable, this is what will be displayed when the fastai library has to show your item.\n\nIf we take the example of a [`MultiCategory`](/core.html#MultiCategory) object `o` for instance:\n- `o.data` is a tensor where the tags are one-hot encoded\n- `str(o)` returns the tags separated by ;\n\nIf you want to code the way data augmentation should be applied to your custom `Item`, you should write an `apply_tfms` method. This is what will be called if you apply a [`transform`](/vision.transform.html#vision.transform) block in the data block API.", "_____no_output_____" ], [ "### Example: ImageTuple", "_____no_output_____" ], [ "For cycleGANs, we need to create a custom type of items since we feed the model tuples of images. Let's look at how to code this. The basis is to code the `obj` and [`data`](/vision.data.html#vision.data) attributes. We do that in the init. The object is the tuple of images and the data their underlying tensors normalized between -1 and 1.", "_____no_output_____" ] ], [ [ "class ImageTuple(ItemBase):\n def __init__(self, img1, img2):\n self.img1,self.img2 = img1,img2\n self.obj,self.data = (img1,img2),[-1+2*img1.data,-1+2*img2.data]", "_____no_output_____" ] ], [ [ "Then we want to apply data augmentation to our tuple of images. That's done by writing and `apply_tfms` method as we saw before. Here we just pass that call to the two underlying images then update the data.", "_____no_output_____" ] ], [ [ " def apply_tfms(self, tfms, **kwargs):\n self.img1 = self.img1.apply_tfms(tfms, **kwargs)\n self.img2 = self.img2.apply_tfms(tfms, **kwargs)\n self.data = [-1+2*self.img1.data,-1+2*self.img2.data]\n return self", "_____no_output_____" ] ], [ [ "We define a last method to stack the two images next ot each other, which we will use later for a customized `show_batch`/ `show_results` behavior.", "_____no_output_____" ] ], [ [ " def to_one(self): return Image(0.5+torch.cat(self.data,2)/2)", "_____no_output_____" ] ], [ [ "This is all your need to create your custom [`ItemBase`](/core.html#ItemBase). You won't be able to use it until you have put it inside your custom [`ItemList`](/data_block.html#ItemList) though, so you should continue reading the next section.", "_____no_output_____" ], [ "## Creating a custom [`ItemList`](/data_block.html#ItemList) subclass", "_____no_output_____" ], [ "This is the main class that allows you to group your inputs or your targets in the data block API. You can then use any of the splitting or labelling methods before creating a [`DataBunch`](/basic_data.html#DataBunch). To make sure everything is properly working, here is what you need to know.", "_____no_output_____" ], [ "### Class variables", "_____no_output_____" ], [ "Whether you're directly subclassing [`ItemList`](/data_block.html#ItemList) or one of the particular fastai ones, make sure to know the content of the following three variables as you may need to adjust them:\n- `_bunch` contains the name of the class that will be used to create a [`DataBunch`](/basic_data.html#DataBunch) \n- `_processor` contains a class (or a list of classes) of [`PreProcessor`](/data_block.html#PreProcessor) that will then be used as the default to create processor for this [`ItemList`](/data_block.html#ItemList)\n- `_label_cls` contains the class that will be used to create the labels by default\n\n`_label_cls` is the first to be used in the data block API, in the labelling function. If this variable is set to `None`, the label class will be set to [`CategoryList`](/data_block.html#CategoryList), [`MultiCategoryList`](/data_block.html#MultiCategoryList) or [`FloatList`](/data_block.html#FloatList) depending on the type of the first item. The default can be overridden by passing a `label_cls` in the kwargs of the labelling function.\n\n`_processor` is the second to be used. The processors are called at the end of the labelling to apply some kind of function on your items. The default processor of the inputs can be overriden by passing a `processor` in the kwargs when creating the [`ItemList`](/data_block.html#ItemList), the default processor of the targets can be overridden by passing a `processor` in the kwargs of the labelling function. \n\nProcessors are useful for pre-processing some data, but you also need to put in their state any variable you want to save for the call of `data.export()` before creating a [`Learner`](/basic_train.html#Learner) object for inference: the state of the [`ItemList`](/data_block.html#ItemList) isn't saved there, only their processors. For instance `SegmentationProcessor`'s only reason to exist is to save the dataset classes, and during the process call, it doesn't do anything apart from setting the `classes` and `c` attributes to its dataset.\n``` python\nclass SegmentationProcessor(PreProcessor):\n def __init__(self, ds:ItemList): self.classes = ds.classes\n def process(self, ds:ItemList): ds.classes,ds.c = self.classes,len(self.classes)\n```\n\n`_bunch` is the last class variable usd in the data block. When you type the final `databunch()`, the data block API calls the `_bunch.create` method with the `_bunch` of the inputs. ", "_____no_output_____" ], [ "### Keeping \\_\\_init\\_\\_ arguments", "_____no_output_____" ], [ "If you pass additional arguments in your `__init__` call that you save in the state of your [`ItemList`](/data_block.html#ItemList), we have to make sure they are also passed along in the `new` method as this one is used to create your training and validation set when splitting. To do that, you just have to add their names in the `copy_new` argument of your custom [`ItemList`](/data_block.html#ItemList), preferably during the `__init__`. Here we will need two collections of filenames (for the two type of images) so we make sure the second one is copied like this:\n\n```python\ndef __init__(self, items, itemsB=None, **kwargs):\n super().__init__(items, **kwargs)\n self.itemsB = itemsB\n self.copy_new.append('itemsB')\n```\n\nBe sure to keep the kwargs as is, as they contain all the additional stuff you can pass to an [`ItemList`](/data_block.html#ItemList).", "_____no_output_____" ], [ "### Important methods", "_____no_output_____" ], [ "#### - get", "_____no_output_____" ], [ "The most important method you have to implement is `get`: this one will enable your custom [`ItemList`](/data_block.html#ItemList) to generate an [`ItemBase`](/core.html#ItemBase) from the thing stored in its `items` array. For instance an [`ImageItemList`](/vision.data.html#ImageItemList) has the following `get` method:\n``` python\ndef get(self, i):\n fn = super().get(i)\n res = self.open(fn)\n self.sizes[i] = res.size\n return res\n```\nThe first line basically looks at `self.items[i]` (which is a filename). The second line opens it since the `open`method is just\n``` python\ndef open(self, fn): return open_image(fn)\n```\nThe third line is there for [`ImagePoints`](/vision.image.html#ImagePoints) or [`ImageBBox`](/vision.image.html#ImageBBox) targets that require the size of the input [`Image`](/vision.image.html#Image) to be created. Note that if you are building a custom target class and you need the size of an image, you should call `self.x.size[i]`. ", "_____no_output_____" ] ], [ [ "jekyll_note(\"\"\"If you just want to customize the way an `Image` is opened, subclass `Image` and just change the\n`open` method.\"\"\")", "_____no_output_____" ] ], [ [ "#### - reconstruct", "_____no_output_____" ], [ "This is the method that is called in `data.show_batch()`, `learn.predict()` or `learn.show_results()` to transform a pytorch tensor back in an [`ItemBase`](/core.html#ItemBase). In a way, it does the opposite of calling `ItemBase.data`. It should take a tensor `t` and return the same king of thing as the `get` method.\n\nIn some situations ([`ImagePoints`](/vision.image.html#ImagePoints), [`ImageBBox`](/vision.image.html#ImageBBox) for instance) you need to have a look at the corresponding input to rebuild your item. In this case, you should have a second argument called `x` (don't change that name). For instance, here is the `reconstruct` method of [`PointsItemList`](/vision.data.html#PointsItemList):\n```python\ndef reconstruct(self, t, x): return ImagePoints(FlowField(x.size, t), scale=False)\n```", "_____no_output_____" ], [ "#### - analyze_pred", "_____no_output_____" ], [ "This is the method that is called in `learn.predict()` or `learn.show_results()` to transform predictions in an output tensor suitable for `reconstruct`. For instance we may need to take the maximum argument (for [`Category`](/core.html#Category)) or the predictions greater than a certain threshold (for [`MultiCategory`](/core.html#MultiCategory)). It should take a tensor, along with optional kwargs and return a tensor.\n\nFor instance, here is the `analyze_pred` method of [`MultiCategoryList`](/data_block.html#MultiCategoryList):\n```python\ndef analyze_pred(self, pred, thresh:float=0.5): return (pred >= thresh).float()\n```\n`thresh` can then be passed as kwarg during the calls to `learn.predict()` or `learn.show_results()`.", "_____no_output_____" ], [ "### Advanced show methods", "_____no_output_____" ], [ "If you want to use methods such a `data.show_batch()` or `learn.show_results()` with a brand new kind of [`ItemBase`](/core.html#ItemBase) you will need to implement two other methods. In both cases, the generic function will grab the tensors of inputs, targets and predictions (if applicable), reconstruct the corresponding [`ItemBase`](/core.html#ItemBase) (as seen before) but it will delegate to the [`ItemList`](/data_block.html#ItemList) the way to display the results.\n\n``` python\ndef show_xys(self, xs, ys, **kwargs)->None:\n\ndef show_xyzs(self, xs, ys, zs, **kwargs)->None:\n```\nIn both cases `xs` and `ys` represent the inputs and the targets, in the second case `zs` represent the predictions. They are lists of the same length that depend on the `rows` argument you passed. The kwargs are passed from `data.show_batch()` / `learn.show_results()`. As an example, here is the source code of those methods in [`ImageItemList`](/vision.data.html#ImageItemList):\n\n``` python\ndef show_xys(self, xs, ys, figsize:Tuple[int,int]=(9,10), **kwargs):\n \"Show the `xs` and `ys` on a figure of `figsize`. `kwargs` are passed to the show method.\"\n rows = int(math.sqrt(len(xs)))\n fig, axs = plt.subplots(rows,rows,figsize=figsize)\n for i, ax in enumerate(axs.flatten() if rows > 1 else [axs]):\n xs[i].show(ax=ax, y=ys[i], **kwargs)\n plt.tight_layout()\n \ndef show_xyzs(self, xs, ys, zs, figsize:Tuple[int,int]=None, **kwargs):\n \"\"\"Show `xs` (inputs), `ys` (targets) and `zs` (predictions) on a figure of `figsize`. \n `kwargs` are passed to the show method.\"\"\"\n figsize = ifnone(figsize, (6,3*len(xs)))\n fig,axs = plt.subplots(len(xs), 2, figsize=figsize)\n fig.suptitle('Ground truth / Predictions', weight='bold', size=14)\n for i,(x,y,z) in enumerate(zip(xs,ys,zs)):\n x.show(ax=axs[i,0], y=y, **kwargs)\n x.show(ax=axs[i,1], y=z, **kwargs)\n``` \n\nLinked to this method is the class variable `_show_square` of an [`ItemList`](/data_block.html#ItemList). It defaults to `False` but if it's `True`, the `show_batch` method will send `rows * rows` `xs` and `ys` to `show_xys` (so that it shows a square of inputs/targets), like here for images.", "_____no_output_____" ], [ "### Example: ImageTupleList", "_____no_output_____" ], [ "Continuing our custom item example, we create a custom [`ItemList`](/data_block.html#ItemList) class that will wrap those `ImageTuple`s properly. The first thing is to write a custom `__init__` method (since we need a list of filenames here) which means we also have to change the `new` method.", "_____no_output_____" ] ], [ [ "class ImageTupleList(ImageItemList):\n def __init__(self, items, itemsB=None, **kwargs):\n super().__init__(items, **kwargs)\n self.itemsB = itemsB\n self.copy_new.append('itemsB')", "_____no_output_____" ] ], [ [ "We then specify how to get one item. Here we pass the image in the first list of items, and pick one randomly in the second list.", "_____no_output_____" ] ], [ [ " def get(self, i):\n img1 = super().get(i)\n fn = self.itemsB[random.randint(0, len(self.itemsB)-1)]\n return ImageTuple(img1, open_image(fn))", "_____no_output_____" ] ], [ [ "We also add a custom factory method to directly create an `ImageTupleList` from two folders.", "_____no_output_____" ] ], [ [ " @classmethod\n def from_folders(cls, path, folderA, folderB, **kwargs):\n itemsB = ImageItemList.from_folder(path/folderB).items\n res = super().from_folder(path/folderA, itemsB=itemsB, **kwargs)\n res.path = path\n return res", "_____no_output_____" ] ], [ [ "Finally, we have to specify how to reconstruct the `ImageTuple` from tensors if we want `show_batch` to work. We recreate the images and denormalize.", "_____no_output_____" ] ], [ [ " def reconstruct(self, t:Tensor): \n return ImageTuple(Image(t[0]/2+0.5),Image(t[1]/2+0.5))", "_____no_output_____" ] ], [ [ "There is no need to write a `analyze_preds` method since the default behavior (returning the output tensor) is what we need here. However `show_results` won't work properly unless the target (which we don't really care about here) has the right `reconstruct` method: the fastai library uses the `reconstruct` method of the target on the outputs. That's why we create another custom [`ItemList`](/data_block.html#ItemList) with just that `reconstruct` method. The first line is to reconstruct our dummy targets, and the second one is the same as in `ImageTupleList`.", "_____no_output_____" ] ], [ [ "class TargetTupleList(ItemList):\n def reconstruct(self, t:Tensor): \n if len(t.size()) == 0: return t\n return ImageTuple(Image(t[0]/2+0.5),Image(t[1]/2+0.5))", "_____no_output_____" ] ], [ [ "To make sure our `ImageTupleList` uses that for labelling, we pass it in `_label_cls` and this is what the result looks like.", "_____no_output_____" ] ], [ [ "class ImageTupleList(ImageItemList):\n _label_cls=TargetTupleList\n def __init__(self, items, itemsB=None, **kwargs):\n super().__init__(items, **kwargs)\n self.itemsB = itemsB\n self.copy_new.append('itemsB')\n \n def get(self, i):\n img1 = super().get(i)\n fn = self.itemsB[random.randint(0, len(self.itemsB)-1)]\n return ImageTuple(img1, open_image(fn))\n \n def reconstruct(self, t:Tensor): \n return ImageTuple(Image(t[0]/2+0.5),Image(t[1]/2+0.5))\n \n @classmethod\n def from_folders(cls, path, folderA, folderB, **kwargs):\n itemsB = ImageItemList.from_folder(path/folderB).items\n res = super().from_folder(path/folderA, itemsB=itemsB, **kwargs)\n res.path = path\n return res", "_____no_output_____" ] ], [ [ "Lastly, we want to customize the behavior of `show_batch` and `show_results`. Remember the `to_one` method just puts the two images next to each other.", "_____no_output_____" ] ], [ [ " def show_xys(self, xs, ys, figsize:Tuple[int,int]=(12,6), **kwargs):\n \"Show the `xs` and `ys` on a figure of `figsize`. `kwargs` are passed to the show method.\"\n rows = int(math.sqrt(len(xs)))\n fig, axs = plt.subplots(rows,rows,figsize=figsize)\n for i, ax in enumerate(axs.flatten() if rows > 1 else [axs]):\n xs[i].to_one().show(ax=ax, **kwargs)\n plt.tight_layout()\n\n def show_xyzs(self, xs, ys, zs, figsize:Tuple[int,int]=None, **kwargs):\n \"\"\"Show `xs` (inputs), `ys` (targets) and `zs` (predictions) on a figure of `figsize`.\n `kwargs` are passed to the show method.\"\"\"\n figsize = ifnone(figsize, (12,3*len(xs)))\n fig,axs = plt.subplots(len(xs), 2, figsize=figsize)\n fig.suptitle('Ground truth / Predictions', weight='bold', size=14)\n for i,(x,z) in enumerate(zip(xs,zs)):\n x.to_one().show(ax=axs[i,0], **kwargs)\n z.to_one().show(ax=axs[i,1], **kwargs)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

ec7caa95e7fc0734887291da8aed749d714d3e82

45,275

ipynb

Jupyter Notebook

Mnist_1.ipynb

everestso/Fall21Spring22

0a1039f59f43086a96168211d7bdc7cae93cf3bd

[ "Apache-2.0" ]

null

Mnist_1.ipynb

everestso/Fall21Spring22

0a1039f59f43086a96168211d7bdc7cae93cf3bd

[ "Apache-2.0" ]

null

Mnist_1.ipynb

everestso/Fall21Spring22

0a1039f59f43086a96168211d7bdc7cae93cf3bd

[ "Apache-2.0" ]

1

2021-02-09T20:46:41.000Z

85.424528

5,586

0.7564

[ [ [ "<a href=\"https://colab.research.google.com/github/everestso/Fall2021/blob/main/Mnist_1.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>", "_____no_output_____" ], [ "# Digit Recognizer", "_____no_output_____" ] ], [ [ "import numpy as np\nimport matplotlib.pyplot as plt\n\nimport pandas as pd\n\nfrom sklearn.decomposition import PCA\nfrom sklearn import datasets\n\nfrom sklearn.linear_model import LogisticRegression\nfrom sklearn.metrics import accuracy_score", "_____no_output_____" ], [ "mnist = pd.read_csv('MnistTrain.csv')\ndata = mnist.values\nprint(data.shape)\nprint (data[1,1:])", "(42000, 785)\n[ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 18 30 137 137\n 192 86 72 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 13 86 250 254 254 254 254 217 246 151 32 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 16 179 254 254 254\n 254 254 254 254 254 254 231 54 15 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 72 254 254 254 254 254 254 254 254 254 254 254 254\n 104 0 0 0 0 0 0 0 0 0 0 0 0 0 61 191 254 254\n 254 254 254 109 83 199 254 254 254 254 243 85 0 0 0 0 0 0\n 0 0 0 0 0 0 172 254 254 254 202 147 147 45 0 11 29 200\n 254 254 254 171 0 0 0 0 0 0 0 0 0 0 0 1 174 254\n 254 89 67 0 0 0 0 0 0 128 252 254 254 212 76 0 0 0\n 0 0 0 0 0 0 0 47 254 254 254 29 0 0 0 0 0 0\n 0 0 83 254 254 254 153 0 0 0 0 0 0 0 0 0 0 80\n 254 254 240 24 0 0 0 0 0 0 0 0 25 240 254 254 153 0\n 0 0 0 0 0 0 0 0 0 64 254 254 186 7 0 0 0 0\n 0 0 0 0 0 166 254 254 224 12 0 0 0 0 0 0 0 0\n 14 232 254 254 254 29 0 0 0 0 0 0 0 0 0 75 254 254\n 254 17 0 0 0 0 0 0 0 0 18 254 254 254 254 29 0 0\n 0 0 0 0 0 0 0 48 254 254 254 17 0 0 0 0 0 0\n 0 0 2 163 254 254 254 29 0 0 0 0 0 0 0 0 0 48\n 254 254 254 17 0 0 0 0 0 0 0 0 0 94 254 254 254 200\n 12 0 0 0 0 0 0 0 16 209 254 254 150 1 0 0 0 0\n 0 0 0 0 0 15 206 254 254 254 202 66 0 0 0 0 0 21\n 161 254 254 245 31 0 0 0 0 0 0 0 0 0 0 0 60 212\n 254 254 254 194 48 48 34 41 48 209 254 254 254 171 0 0 0 0\n 0 0 0 0 0 0 0 0 0 86 243 254 254 254 254 254 233 243\n 254 254 254 254 254 86 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 114 254 254 254 254 254 254 254 254 254 254 239 86 11 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 13 182 254 254 254 254\n 254 254 254 254 243 70 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 8 76 146 254 255 254 255 146 19 15 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n 0 0 0 0 0 0 0 0 0 0]\n" ], [ "# print one number\nmyNumber = data[1,1:]\nprint (myNumber[:5])\nmyNumber=myNumber.reshape((28,28))\nplt.imshow(myNumber, cmap = plt.cm.binary,\n interpolation=\"nearest\")\nplt.axis(\"off\")\nplt.show()\nmnist['label'].hist()\nplt.show()\nprint(mnist.columns)\nplt.hist(data[:,0])\nplt.show()", "[0 0 0 0 0]\n" ], [ "mnist_test = pd.read_csv('MnistTest.csv')\nmnist_test.info()\nmnist_test.index.name='ImageId'\n\nmnist_test['label']=1\n\nmnist_test.index+=1\nmnist_test['label'].to_csv('MnistOne.csv', index=True, header=True)", "<class 'pandas.core.frame.DataFrame'>\nRangeIndex: 28000 entries, 0 to 27999\nColumns: 784 entries, pixel0 to pixel783\ndtypes: int64(784)\nmemory usage: 167.5 MB\n" ], [ "#rnd_pca = PCA(n_components=154)\n#rnd_pca = PCA(n_components=169)\nrnd_pca = PCA(n_components=4)\n\nrnd_pca.fit(data[:,1:])\nX_reduced = rnd_pca.transform(data[:,1:])\nprint (X_reduced.shape)", "(42000, 4)\n" ], [ "#eigennumbers = rnd_pca.components_.reshape((154,28,28))\n#eigennumbers = rnd_pca.components_.reshape((169,28,28))\neigennumbers = rnd_pca.components_.reshape((4,28,28))\nmyNumber = eigennumbers[0]\nprint (len(myNumber))\nmyNumber=myNumber.reshape((28,28))\nplt.imshow(myNumber, cmap = plt.cm.binary,\n interpolation=\"nearest\")\nplt.axis(\"off\")\nplt.show()\nmyNumber = eigennumbers[1]\nmyNumber=myNumber.reshape((28,28))\nplt.imshow(myNumber, cmap = plt.cm.binary,\n interpolation=\"nearest\")\nplt.axis(\"off\")\nplt.show()\nmyNumber = eigennumbers[2]\nmyNumber=myNumber.reshape((28,28))\nplt.imshow(myNumber, cmap = plt.cm.binary,\n interpolation=\"nearest\")\nplt.axis(\"off\")\nplt.show()", "28\n" ], [ "import math\nmyNumber = X_reduced[0,:]\nmyNumber=myNumber.reshape((2,2))\nplt.imshow(myNumber, cmap = plt.cm.binary,\n interpolation=\"nearest\")\nplt.show()", "_____no_output_____" ], [ "#clf1 = LogisticRegression(max_iter=10000, solver='saga',)\n\nclf1.fit(data[:,1:], data[:, 0] )", "_____no_output_____" ], [ "clf1 = LogisticRegression(solver='saga',\n max_iter=10000)\nclf1.fit(X_reduced, data[:, 0] )", "_____no_output_____" ], [ "predict = clf1.predict(X_reduced)\nprint(\"Accuracy = \", accuracy_score(predict, data[:, 0]))", "Accuracy = 0.5607619047619048\n" ], [ "mnistTest = pd.read_csv('MnistTest.csv')\ntest = mnistTest.values\nXTest_reduced = rnd_pca.transform(test)\nprint(XTest_reduced.shape)\npredict = clf1.predict(XTest_reduced)", "(28000, 4)\n" ], [ "mnistTest.info()\nmnistTest.index.name='ImageId'\n\nmnistTest['label']=predict\n\nmnistTest.index+=1\nmnistTest['label'].to_csv('Mnist2x2.csv', index=True, header=True)", "<class 'pandas.core.frame.DataFrame'>\nRangeIndex: 28000 entries, 0 to 27999\nColumns: 784 entries, pixel0 to pixel783\ndtypes: int64(784)\nmemory usage: 167.5 MB\n" ] ], [ [ "", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown" ]

[ [ "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ] ]

ec7cbfd1ef1726968130ddc0117394605a3fb45d

95,937

ipynb

Jupyter Notebook

doc/pub/eigvalues/ipynb/eigvalues.ipynb

xxzpl/ComputationalPhysics

794c53c54a3a404e933626b0b4c807f3c92e61f0

[ "CC0-1.0" ]

1

2020-06-17T16:16:59.000Z

doc/pub/eigvalues/ipynb/eigvalues.ipynb

cosmologist10/ComputationalPhysics

c6642becb1036e2faaf4f1da78a31785b2033fe7

[ "CC0-1.0" ]

null

doc/pub/eigvalues/ipynb/eigvalues.ipynb

cosmologist10/ComputationalPhysics

c6642becb1036e2faaf4f1da78a31785b2033fe7

[ "CC0-1.0" ]

null

30.572658

516

0.482077

[ [ [ "empty" ] ] ]

[ "empty" ]

[ [ "empty" ] ]

ec7cc4199b94c6b3067e4302a5f487fad8ade59f

918,353

ipynb

Jupyter Notebook

9_Prognose_with_Python.ipynb

kafasin/big_data_science

b989cea4e6f53eb66adfb9add7bab71006775a27

[ "MIT" ]

1

2020-08-16T19:35:48.000Z

9_Prognose_with_Python.ipynb

kafasin/big_data_science

b989cea4e6f53eb66adfb9add7bab71006775a27

[ "MIT" ]

null

9_Prognose_with_Python.ipynb

kafasin/big_data_science

b989cea4e6f53eb66adfb9add7bab71006775a27

[ "MIT" ]

null

238.657225

115,400

0.899207

[ [ [ "## Forecasting", "_____no_output_____" ] ], [ [ "# Libraries\n\nimport numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport seaborn as sns\n\n%matplotlib inline\n\n# Databank \nimport mysql.connector\nfrom sqlalchemy import create_engine, exc\n\n\nimport itertools ### Modul, welches bestimmte Schleifen ermöglicht\nimport warnings ###Modul, welches uns erlaubt, bestimmte Warnungen ein- und auszublenden\n\n#Sonstige Module und Funktionen\nfrom time import gmtime, strftime #Für die aktuelle Zeit\n", "_____no_output_____" ], [ "#Einstellen der Darstellungsoptionen\n\n#Pandas - Mehr Zeilen und Spalten anzeigen\npd.set_option('display.max_rows', 500)\npd.set_option('display.max_columns', 500)\n\n#Seaborn\nsns.set_style(\"darkgrid\")\nplt.matplotlib.style.use('default')\n\nmy_colors = [\"windows blue\", \"saffron\", \"hot pink\", \"algae green\", \"dusty purple\", \"greyish\", \"petrol\", \"denim blue\", \"lime\"]\nsns.set_palette(sns.xkcd_palette(my_colors))\ncolors = sns.xkcd_palette(my_colors)\n\n#Warnings\nwarnings.filterwarnings(\"ignore\")\n\n#Connection-String für die Datenbank\nstr_db_typ = 'mysql+mysqlconnector://'\nstr_db_user = 'root'\nstr_db_password = ''\nstr_db_adr = '@localhost'\nstr_db_schema = '' ", "_____no_output_____" ], [ "### Sammlung aller selbstgeschriebenen Funktionen\n\n#Mein Connection-String\ndef my_con_str(db_typ='mysql+mysqlconnector', db_user='root', db_passwort='', db_adr = 'localhost', db_schema = ''):\n '''Funktion, die einen DB-Connection-String zusammensetzt'''\n try:\n if len(db_schema) == 0:\n schema_sep = ''\n print('Achtung: Es wurde keine Datenbank angeben.')\n else:\n schema_sep ='/' \n con_str = '{0}://{1}{2}@{3}{4}{5}'.format(db_typ, db_user, db_passwort, db_adr, schema_sep, db_schema)\n print('Connection: {}'.format(con_str))\n return con_str\n except:\n print('Etwas ist schief gelaufen. Bitte alle Parameter überprüfen.')\n con_str = ''\n return con_str\n \n\n#Anlegen einer neuen Datenbank, falls diese noch nicht da ist.\ndef my_create_schema(dbname, con):\n engine = create_engine(con)\n engine.execute('CREATE DATABASE IF NOT EXISTS {}'.format(dbname))\n return\n\n\n#Meine Datenzusammenfassung\ndef my_df_summary(data):\n '''Eigene Funktion für die Summary'''\n try:\n dat = data.copy()\n df = pd.DataFrame([dat.min(), dat.max(), dat.mean(), dat.std(), dat.isna().sum(), dat.nunique(), dat.dtypes],\n index=['Minimum', 'Maximum', 'Mittelwert', 'Stand. Abw.','#NA', '#Uniques', 'dtypes']) \n return df\n except:\n print('Es konnte keine Summary erstellt werden.')\n return data \n\n", "_____no_output_____" ] ], [ [ "### Datenbank einrichten", "_____no_output_____" ] ], [ [ "con = my_con_str()", "Achtung: Es wurde keine Datenbank angeben.\nConnection: mysql+mysqlconnector://root@localhost\n" ], [ "my_create_schema('timeseries',con)", "_____no_output_____" ], [ "con = my_con_str(db_schema='timeseries')", "Connection: mysql+mysqlconnector://root@localhost/timeseries\n" ] ], [ [ "### Datensatz", "_____no_output_____" ] ], [ [ "#Einlesen und anschauen des zu bearbeitenden Datensatzes\ndf = pd.read_excel(r'NYSE_sample.xlsx')", "_____no_output_____" ], [ "df.head()", "_____no_output_____" ], [ "df = df.iloc[:,0:6]", "_____no_output_____" ], [ "df.info()", "<class 'pandas.core.frame.DataFrame'>\nRangeIndex: 3020 entries, 0 to 3019\nData columns (total 6 columns):\n # Column Non-Null Count Dtype \n--- ------ -------------- ----- \n 0 Date 3020 non-null object \n 1 AABA 3019 non-null float64\n 2 AAPL 3019 non-null float64\n 3 AMZN 3019 non-null float64\n 4 AXP 3020 non-null float64\n 5 BA 3020 non-null float64\ndtypes: float64(5), object(1)\nmemory usage: 141.7+ KB\n" ], [ "df.describe()", "_____no_output_____" ], [ "my_df_summary(df)", "_____no_output_____" ], [ "df.plot()", "_____no_output_____" ], [ "#Oder grafisch\n%matplotlib inline\nplt.rcParams['figure.figsize'] = [15, 6]\nfig = df.plot(kind='line')\nplt.title('Plot aller numerischen Werte', size=14)\nplt.ylabel('Preis in USD')\nplt.legend(fontsize=12)\nplt.show()", "_____no_output_____" ] ], [ [ "### 1. Schritt: In welcher Frequenz liegen meine Zeitreihendaten vor?", "_____no_output_____" ] ], [ [ "df_ = df.copy()\n \ns_Dates = pd.to_datetime(df_.iloc[:,0], format='%Y-%m-%d', errors='ignore')\n \n#Mit \"Nummer\" des Wochentages\ndf_NuDay = pd.DataFrame(s_Dates.dt.dayofweek.value_counts(dropna=False))\ndf_NuDay = df_NuDay.reset_index()\ndf_NuDay.columns=['#Tag', 'Anzahl']\n \n#Mit \"Namen\" des Wochentages\ndf_NaDay = pd.DataFrame(s_Dates.dt.day_name().value_counts(dropna=False))\ndf_NaDay = df_NaDay.reset_index()\ndf_NaDay.columns=['Tag', 'Anzahl']", "_____no_output_____" ], [ "df_NaDay", "_____no_output_____" ], [ "#Oder grafisch\n%matplotlib inline\nplt.rcParams['figure.figsize'] = [15, 4]\nfig=sns.barplot(x='Tag', y='Anzahl', data=df_NaDay)\nplt.title(\"Wie häufig sind die einzelnen Wochentage enthalten?\", size=14)\nplt.xlabel(\"\")\nplt.ylabel(\"Häufigkeiten\")\nplt.show()", "_____no_output_____" ] ], [ [ "### Aufbereiten der fehlenden Werte", "_____no_output_____" ] ], [ [ "#Das muss Zeitreihe für Zeitreihe individuell gemacht werden!\n#Schritt 0: Auswahl einer Zeitreihe\ndf_ = df.iloc[:,0:2].copy() #Hier wähle ich den Index (Spalte 0) und die erste Datenspalte (Spalte 1) aus.\n\n# df.iloc[:,[0,4]].copy() #So würde ich bspw. die 4. Datenspalte (Spalte 5) auswählen.", "_____no_output_____" ], [ "df_.head()", "_____no_output_____" ], [ "#Cleansing\n#Schritt 1:\nl_colnames = df_.columns.to_list()\nl_colnames[0] = 'Date'\ndf_.columns = l_colnames\n\ndf_['Date'] = pd.to_datetime(df_['Date'], format='%Y-%m-%d', errors='ignore')\ndaterange = pd.date_range(start=min(df_['Date']), end=max(df_['Date']), freq='B')\ndf_ts = pd.DataFrame(daterange)\ndf_ts.columns = ['Date']\ndf_ts = df_ts.merge(df_, how='left', on='Date')\n\n#Schritt 2\nprint('{} fehlende Werte werden durch den zuletzt gültigen Wert ersetzt.'.format(df_ts.iloc[:,1].isna().sum()))\ndf_ts = df_ts.fillna(method='ffill')\ndf_ts = df_ts.set_index('Date', drop=True)", "110 fehlende Werte werden durch den zuletzt gültigen Wert ersetzt.\n" ], [ "#Plotten der vollständigen Daten\n%matplotlib inline\nplt.rcParams['figure.figsize'] = [15, 6]\nfig = df_ts.plot(kind='line')\nplt.title('Vollständige Zeitreihe', size=18)\nplt.legend(fontsize=12)\nplt.ylabel('Preis in USD', size=12)\nplt.show()", "_____no_output_____" ] ], [ [ "### Ausreißer erkennen", "_____no_output_____" ] ], [ [ "#Ausreißer Anzeigen - Für Zeitreihen NICHT entfernen.\n#Ausreißer erkennt man in Zeitreihen an dem Verhalten der prozentualen Veränderungen zum jeweils vorherigen Wert.\n%matplotlib inline\ndf_ = df.iloc[:,0:2].copy()\ndf_.iloc[:,1] = np.log(df_.iloc[:,1]) #Log-Differenzen sind die prozentualen Veränderungen\ndf_diff = df_.iloc[:,1].diff()\n\nplt.rcParams['figure.figsize'] = [15, 2]\n\nfig = sns.boxplot(data=df_diff, orient='h')\n\nplt.title('Häufigkeiten der prozentualen Wertveränderungen', size=14)\nplt.xlabel('')\nplt.show()", "_____no_output_____" ], [ "i_iqr_faktor = 2\ndf_ = df.iloc[:,0:2].copy()\ndf_['Date'] = pd.to_datetime(df_['Date'], format='%Y-%m-%d', errors='ignore')\ndf_.iloc[:,1:] = np.log(df_.iloc[:,1:])\ndf_diff = df_.diff()\n\n\nq25 = df_diff.iloc[:,1].quantile(0.25)\nq75 = df_diff.iloc[:,1].quantile(0.75)\n\niqr = q75-q25\n\ngrenze_unten = q25 - (i_iqr_faktor*iqr)\ngrenze_oben = q75 + (i_iqr_faktor*iqr)\ndf_[((df_diff.iloc[:,1] < grenze_unten) | (df_diff.iloc[:,1] > grenze_oben))]\n\n\n\n\ndf_dates = df_[((df_diff.iloc[:,1] < grenze_unten) | (df_diff.iloc[:,1] > grenze_oben))]\ndf_dates = df_dates.reset_index()\ndf_dates = df_dates.iloc[:,0:2]\n \nprint('Bei den eingegebenen Daten und des IQR-Faktors sind Tage, an denen die Wertveränderung < {0:.2f} oder > {1:.2f} war, auffällig.'.format(grenze_unten, grenze_oben))\nprint('Dies tritt an {} Tagen auf: '.format(len(df_dates)))\ndf_dates", "Bei den eingegebenen Daten und des IQR-Faktors sind Tage, an denen die Wertveränderung < -0.05 oder > 0.06 war, auffällig.\nDies tritt an 89 Tagen auf: \n" ], [ "#Sind einzelne Jahre (Monate) besonders auffällig?\n%matplotlib inline\ndf_dates['Jahr'] = df_dates['Date'].map(lambda x: x.strftime('%Y'))\nplt.rcParams['figure.figsize'] = [15, 4]\ndf_dates.groupby('Jahr').size().plot(kind = 'bar')\nplt.xlabel('Periode')\nplt.ylabel('Anzahl')\nplt.show()", "_____no_output_____" ], [ "#Zu untersuchende Daten auswählen\ndf_ts_clean = df_ts[df_ts.index > '2013-12-31']\nprint('Insgesamt liegen {} zusammenhängende ähnliche Beobachtungen vor.'.format(len(df_ts_clean)))", "Insgesamt liegen 1043 zusammenhängende ähnliche Beobachtungen vor.\n" ] ], [ [ "### Erzeugen eines Trainings- und eines Testdatensatzes", "_____no_output_____" ] ], [ [ "#Zerteilung in Trainings- und Testdaten - Im Normalfall beginnt man mit dem Verhältnis 80:20\n#Weil wir Tagesdaten haben und mit \"klassischen\" Zeitreihenanalyseverfahren schwerlich mehr als 60 Perioden mit hoher\n#Genauigkeit schätzen kann, teilen wir 95:5.\n\ni_split = int(0.95*len(df_ts_clean))\n\ndf_train, df_test = df_ts_clean.iloc[:i_split,:], df_ts_clean.iloc[i_split:,:]\n\nprint('Train und Test sind zusammen {} Einträge lang.'.format(len(df_train)+len(df_test)))\nprint('D.h., alle Forecasts müssen {} Perioden lang sein.'.format(len(df_test)))", "Train und Test sind zusammen 1043 Einträge lang.\nD.h., alle Forecasts müssen 53 Perioden lang sein.\n" ], [ "#Plot\n\nplt.rcParams['figure.figsize'] = [15, 6]\n\nplt.plot(df_train.index, df_train.values, label='Trainingsdaten')\nplt.plot(df_test.index, df_test.values, label='Testdaten', color=colors[1])\n\nplt.axvline(x = df_ts_clean.index[i_split], linewidth=2, color='grey', ls='--')\nplt.legend(loc=2, fontsize=10)\nplt.title('Aktienkurs (Close) von {} an der NYSE'.format(df_train.columns[0]), fontsize=14)\nplt.xlabel('Zeit', fontsize=10)\nplt.ylabel('Preis in USD', fontsize=10)\nplt.show()", "_____no_output_____" ] ], [ [ "## Prognose-Verfahren", "_____no_output_____" ], [ "## Naiver Forecast - Einfaches Fortschreiben des letzten Wertes <br>\n\n<font size=\"4\">\n\\begin{align}\n \\hat{y}_{t+1} &= y_{t} \n\\end{align} <br>\n</font>", "_____no_output_____" ] ], [ [ "vals = np.asarray(df_train.values)\ny_hat = df_test.copy()\ny_hat['naiv'] = vals[-1][0]", "_____no_output_____" ], [ "%matplotlib inline\nplt.rcParams['figure.figsize'] = [15, 6]\n\nplt.plot(df_train.index, df_train.values, label='Traininsdaten')\nplt.plot(df_test.index, df_test.values, label='Testdaten')\nplt.plot(y_hat.index, y_hat['naiv'], label='Naiver Forecast')\nplt.legend(loc='best')\nplt.title(\"Naiver Forecast\")\nplt.show()", "_____no_output_____" ] ], [ [ "### Wie lässt sich die Güte eines Forecasts bewerten?", "_____no_output_____" ], [ "### Root Mean Squared Error - RMSE\n<font size=\"4\">\nTipp: https://en.wikipedia.org/wiki/Root-mean-square_deviation <br>\n\n\n\\begin{align}\n\\text{RMSE} \\; &= \\sqrt{\\frac{\\sum_{t=1}^T (\\hat{y}_t - y_t)^2}{T}} \n\\end{align} <br>\n\nSeltener schaut man auch einfach auch den durchschnittlichen Fehler.<br>\n\n\\begin{align*}\n\\text{ME} \\; &= \\frac{\\sum_{t=1}^T (\\hat{y}_t - y_t)}{T}\n\\end{align*} <br>\n\n</font>", "_____no_output_____" ] ], [ [ "#Importieren der Fehler-Schätzstatistiken aus sklearn\nfrom sklearn.metrics import mean_squared_error\nfrom math import sqrt #Importieren einer Wurzel-Funktion aus math", "_____no_output_____" ], [ "rmse = sqrt(mean_squared_error(df_test, y_hat.naiv))\nme = (df_test.iloc[:,0] - y_hat['naiv']).sum() / len(df_test)\nprint('Für den naiven Forecast ergeben sich ein ME: {0:.4f} und ein RMSE: {1:.4f}.'.format(me,rmse))", "Für den naiven Forecast ergeben sich ein ME: 3.2947 und ein RMSE: 3.7385.\n" ] ], [ [ "<font size=4>\nAuch kann man die Güte es Forecasts an bestimmten Eigenschaften der Residuen (Schätzfehler) ablesen.<br>\n\n\\begin{align*}\n\\text{Residuen} \\; &= \\hat{y}_t - y_t \\quad = \\epsilon_t\n\\end{align*}\n\n</font>", "_____no_output_____" ] ], [ [ "residuen = (df_test.iloc[:,0] - y_hat['naiv'])\nprint('Die Residuen haben für diesen Forecast folgene Standardabweichung: {0:.4f}.'.format(residuen.std()))\nstdres = residuen.std()", "Die Residuen haben für diesen Forecast folgene Standardabweichung: 1.7835.\n" ], [ "%matplotlib inline\nplt.rcParams['figure.figsize'] = [15, 12]\nplt.subplot(2,1,1)\nresiduen.hist(bins=50, density=True)\nplt.subplot(2,1,2)\nplt.plot(df_test.index, residuen.values, label='Residuen', linewidth=2)\nplt.legend(loc=2)\nplt.title(\"Residuen\")\nplt.show()", "_____no_output_____" ] ], [ [ "### Wie kann man Verfahren denn Vergleichen?", "_____no_output_____" ] ], [ [ "#Anlegen einer Tabelle, um später die Güte verschiedener Verfahren miteinander vergleichen zu können.\n#ACHTUNG: Mit dieser Zeile wird ein leerer DataFrame erzeugt.\ndf_Fehler = pd.DataFrame(columns=['Methode', 'ME', 'RMSE', 'StdRes'])", "_____no_output_____" ], [ "#Einfügen der Güte-Maße\ndf_Fehler = df_Fehler.append({'Methode': 'Naives Fortschreiben', 'ME': me, 'RMSE': rmse, 'StdRes': stdres},\n ignore_index=True)", "_____no_output_____" ], [ "df_Fehler", "_____no_output_____" ] ], [ [ "## Moving Average (Gleitender Mittelwert)", "_____no_output_____" ] ], [ [ "#Mit rolling und mean() kann man den gleitenden Mittelwert ganz einfach erzeugen. \n#Mit einer Schleife und append schreibt man die Werte fort.\n\nn = 60 #Bspw. den gleitenden Durchschnitt über alle Handelstage des letzten Quartals.\ndf_train.rolling(n).mean().iloc[-1][0]", "_____no_output_____" ], [ "n = 60\ndf_mav = df_train.copy()\nfor i in range(len(df_test)):\n df_mav = df_mav.append({df_mav.columns[0] : df_mav.rolling(n).mean().iloc[-1][0]}, ignore_index=True)", "_____no_output_____" ], [ "y_hat = df_test.copy()\ny_hat_mav = df_mav.iloc[-len(df_test):].copy()\ny_hat['mav'] = y_hat_mav.values", "_____no_output_____" ], [ "%matplotlib inline\nplt.rcParams['figure.figsize'] = [15, 6]\n\nplt.plot(df_train.index, df_train.values, label='Trainingsdaten')\nplt.plot(df_test.index, df_test.values, label='Testdaten')\nplt.plot(df_test.index, y_hat['mav'].values, label='Moving Average Forecast')\n\nplt.title('Gleitender Durchschnitt')\nplt.legend(loc='best')\nplt.show()", "_____no_output_____" ], [ "rmse = sqrt(mean_squared_error(df_test, y_hat.mav))\nme = (df_test.iloc[:,0] - y_hat['mav']).sum() / len(df_test)\nprint('Für den naiven Forecast ergeben sich ein ME: {0:.4f} und ein RMSE: {1:.4f}.'.format(me,rmse))\n\nresiduen = (df_test.iloc[:,0] - y_hat['mav'])\nprint('Die Residuen haben für diesen Forecast folgene Standardabweichung: {0:.4f}.'.format(residuen.std()))\nstdres = residuen.std()", "Für den naiven Forecast ergeben sich ein ME: 4.6000 und ein RMSE: 4.8247.\nDie Residuen haben für diesen Forecast folgene Standardabweichung: 1.4691.\n" ], [ "df_Fehler = df_Fehler.append({'Methode': 'Moving Average', 'ME': me, 'RMSE': rmse, 'StdRes': stdres},\n ignore_index=True)", "_____no_output_____" ], [ "df_Fehler", "_____no_output_____" ], [ "%matplotlib inline\nplt.rcParams['figure.figsize'] = [15, 12]\nplt.suptitle('Moving Average')\nplt.subplot(2,1,1)\nresiduen.hist(bins=50, density=True)\nplt.subplot(2,1,2)\nplt.plot(df_test.index, residuen.values, label='Residuen', linewidth=2)\nplt.legend(loc=2)\nplt.title(\"Residuen - Moving Average\")\nplt.show()", "_____no_output_____" ] ], [ [ "<font size = 4>\nTipp: Wer sich intensiver mit der Mathematik hinter den meisten bekannten und etablierten Verfahren zur Zeitreihenen-Analyse beschäftigen möchte, kann sich u.a. hier einlesen: <br>\n\nhttps://www.stat.berkeley.edu/~arturof/Teaching/STAT248/lab10_part1.html\n</font>", "_____no_output_____" ], [ "### Exponentielle Glättung", "_____no_output_____" ], [ "<h3>Einfache (naive) exponentielle Glättung </h3><br>\n<font size=4>\nFür die einfache exponentielle Glättung gilt: <br><br>\n\\begin{align}\n \\hat{y}_{t+1} &= \\hat{y}_{t} + \\alpha ( y_t - \\hat{y}_{t}) \n\\end{align} <br>\n\nWeil $\\alpha$ zwischen 0 und 1 liegen muss, kann man die Gleichung umschreiben: <br><br>\n\\begin{align}\n \\hat{y}_{t+1} &= \\alpha y_t + (1-\\alpha) \\hat{y}_{t} \n\\end{align} <br>\n\nUnd das lässt sich schreiben als: <br><br>\n\\begin{align}\n \\hat{y}_{t+1} &= \\alpha y_{t} + \\alpha(1-\\alpha) y_{t-1} + \\alpha(1-\\alpha)^2 y_{t-2} + \\cdots + \\alpha(1-\\alpha)^{t-1} y_{1} + \\alpha(1-\\alpha)^{t} \\hat{y}_{1} \n\\end{align} <br>\n</font>", "_____no_output_____" ] ], [ [ "import statsmodels.api as sm\nfrom statsmodels.tsa.api import ExponentialSmoothing, SimpleExpSmoothing, Holt", "_____no_output_____" ], [ "#Forecasts mit verschiedenen Glättungsparametern erzeugen\nfit1 = SimpleExpSmoothing(df_train).fit(smoothing_level=0.25,optimized=False)\nfcast1 = fit1.forecast(len(df_test)).rename(r'$\\alpha=0.25$')\n\n\nfit2 = SimpleExpSmoothing(df_train).fit(smoothing_level=0.50,optimized=False)\nfcast2 = fit2.forecast(len(df_test)).rename(r'$\\alpha=0.50$')\n\n\nfit3 = SimpleExpSmoothing(df_train).fit()\nfcast3 = fit3.forecast(len(df_test)).rename(r'$\\alpha=%s$'%round(fit3.model.params['smoothing_level'],1))", "_____no_output_____" ], [ "%matplotlib inline\nplt.rcParams['figure.figsize'] = [15, 6]\n\nfcast1.plot(legend=True, color=colors[0], ls='--')\nfit1.fittedvalues.plot(color=colors[0])\n\nfcast2.plot(legend=True, color=colors[1], ls='--')\nfit2.fittedvalues.plot(color=colors[1])\n\nfcast3.plot(legend=True, color=colors[2], ls='--')\nfit3.fittedvalues.plot(color=colors[2])\n\nplt.title('Anpassung der einfachen exponentiellen Glättung auf {}'.format(df_train.columns[0]), fontsize=14)\nplt.xlabel('Zeit', fontsize=10)\nplt.ylabel('Preis in USD', fontsize=10)\n\nplt.show()", "_____no_output_____" ], [ "fit3.params", "_____no_output_____" ], [ "rmse = sqrt(mean_squared_error(df_test, fcast1.values))\nme = (df_test.iloc[:,0] - fcast1.values).sum() / len(df_test)\nresiduen = (df_test.iloc[:,0] - fcast1.values)\nstdres = residuen.std()\nprint('Forecast 1 - RMSE: {0:.4f}, ME: {1:.4f}, StdRes: {2:.4f}'.format(rmse, me, stdres))", "Forecast 1 - RMSE: 2.9550, ME: 2.3688, StdRes: 1.7835\n" ], [ "df_Fehler = df_Fehler.append({'Methode': 'EinfExpGlätt 0.25', 'ME': me, 'RMSE': rmse, 'StdRes': stdres},\n ignore_index=True)", "_____no_output_____" ], [ "rmse = sqrt(mean_squared_error(df_test, fcast2.values))\nme = (df_test.iloc[:,0] - fcast2.values).sum() / len(df_test)\nresiduen = (df_test.iloc[:,0] - fcast2.values)\nstdres = residuen.std()\nprint('Forecast 2 - RMSE: {0:.4f}, ME: {1:.4f}, StdRes: {2:.4f}'.format(rmse, me, stdres))", "Forecast 2 - RMSE: 3.2135, ME: 2.6843, StdRes: 1.7835\n" ], [ "df_Fehler = df_Fehler.append({'Methode': 'EinfExpGlätt 0.5', 'ME': me, 'RMSE': rmse, 'StdRes': stdres},\n ignore_index=True)", "_____no_output_____" ], [ "rmse = sqrt(mean_squared_error(df_test, fcast3.values))\nme = (df_test.iloc[:,0] - fcast3.values).sum() / len(df_test)\nresiduen = (df_test.iloc[:,0] - fcast3.values)\nstdres = residuen.std()\nprint('Forecast 3 - RMSE: {0:.4f}, ME: {1:.4f}, StdRes: {2:.4f}'.format(rmse, me, stdres))", "Forecast 3 - RMSE: 3.7143, ME: 3.2672, StdRes: 1.7835\n" ], [ "df_Fehler = df_Fehler.append({'Methode': 'EinfExpGlätt 1', 'ME': me, 'RMSE': rmse, 'StdRes': stdres},\n ignore_index=True)", "_____no_output_____" ], [ "df_Fehler", "_____no_output_____" ] ], [ [ "<h3>Exponentielle Glättung mit 'Level' und 'Trend' (Holt-Verfahren)</h3><br>\n<font size=4>\nDas Holt-Verfahren nennt man auch zweifache exponentielle Glättung:<br><br>\n\n\\begin{align*}\n\\text{Level: } \\; \\quad \\ell_t &= \\alpha y_t + (1-\\alpha) (\\ell_{t-1} + b_{t-1}) \\\\ \\\\\n\\text{Growth: } \\; \\quad b_t &= \\beta^* (\\ell_t - \\ell_{t-1}) + (1-\\beta^*) b_{t-1} \\\\ \\\\\n\\text{Forecast: } \\; \\hat{y}_{t+h|t} &= \\ell_t + b_t h \\\\ \\\\\n\\end{align*}\n\nManchmal wird der Trend auch gedämpft. <br><br>\n\n\\begin{align*}\n\\text{Level: } \\; \\quad \\ell_t &= \\alpha y_t + (1-\\alpha) (\\ell_{t-1} + \\phi b_{t-1}) \\\\ \\\\\n\\text{Growth: } \\; \\quad b_t &= \\beta^* (\\ell_t - \\ell_{t-1}) + (1-\\beta^*) b_{t-1} \\\\ \\\\\n\\text{Forecast: } \\; \\hat{y}_{t+h|t} &= \\ell_t + (\\phi + \\phi^2 + \\cdots + \\phi^h) b_t h \\\\ \\\\\n\\end{align*}\n</font>", "_____no_output_____" ] ], [ [ "#Schätzen der zweifachen exponentiellen Glättung\nfit4 = Holt(df_train.iloc[:,0]).fit(smoothing_level=0.8, smoothing_slope=0.2, optimized=False)\nfcast4 = fit4.forecast(len(df_test)).rename(\"Holt's linear trend\")\n\nfit5 = Holt(df_train.iloc[:,0], exponential=True).fit(smoothing_level=0.8, smoothing_slope=0.2, optimized=False)\nfcast5 = fit5.forecast(len(df_test)).rename(\"Exponential trend\")\n\nfit6 = Holt(df_train.iloc[:,0], damped=True).fit()\nfcast6 = fit6.forecast(len(df_test)).rename(\"Additive damped trend\")", "_____no_output_____" ], [ "fit6.params", "_____no_output_____" ], [ "#Plotten der Ergebnisse\n\n%matplotlib inline\nplt.rcParams['figure.figsize'] = [15, 6]\n\nfcast4.plot(legend=True, color=colors[3], ls='--')\nfit4.fittedvalues.plot(color=colors[3])\n\nfcast5.plot(legend=True, color=colors[4], ls='--')\nfit5.fittedvalues.plot(color=colors[4])\n\nfcast6.plot(legend=True, color=colors[5], ls='--')\nfit6.fittedvalues.plot(color=colors[5])\n\nplt.title('Anpassung der zweifachen exponentiellen Glättung auf {}'.format(df_train.columns[0]), fontsize=14)\nplt.xlabel('Zeit', fontsize=10)\nplt.ylabel('Preis in USD', fontsize=10)\n\nplt.show()", "_____no_output_____" ], [ "rmse = sqrt(mean_squared_error(df_test, fcast4.values))\nme = (df_test.iloc[:,0] - fcast4.values).sum() / len(df_test)\nresiduen = (df_test.iloc[:,0] - fcast4.values)\nstdres = residuen.std()\nprint('Forecast 4 - RMSE: {0:.4f}, ME: {1:.4f}, StdRes: {2:.4f}'.format(rmse, me, stdres))\n\ndf_Fehler = df_Fehler.append({'Methode': 'Holt LT', 'ME': me, 'RMSE': rmse, 'StdRes': stdres},\n ignore_index=True)\n\nrmse = sqrt(mean_squared_error(df_test, fcast5.values))\nme = (df_test.iloc[:,0] - fcast5.values).sum() / len(df_test)\nresiduen = (df_test.iloc[:,0] - fcast5.values)\nstdres = residuen.std()\nprint('Forecast 5 - RMSE: {0:.4f}, ME: {1:.4f}, StdRes: {2:.4f}'.format(rmse, me, stdres))\n\ndf_Fehler = df_Fehler.append({'Methode': 'Holt ET', 'ME': me, 'RMSE': rmse, 'StdRes': stdres},\n ignore_index=True)\n\nrmse = sqrt(mean_squared_error(df_test, fcast6.values))\nme = (df_test.iloc[:,0] - fcast6.values).sum() / len(df_test)\nresiduen = (df_test.iloc[:,0] - fcast6.values)\nstdres = residuen.std()\nprint('Forecast 6 - RMSE: {0:.4f}, ME: {1:.4f}, StdRes: {2:.4f}'.format(rmse, me, stdres))\n\ndf_Fehler = df_Fehler.append({'Methode': 'Holt add, damped trend', 'ME': me, 'RMSE': rmse, 'StdRes': stdres},\n ignore_index=True)", "Forecast 4 - RMSE: 9.0572, ME: 8.1390, StdRes: 4.0116\nForecast 5 - RMSE: 8.6140, ME: 7.7632, StdRes: 3.7684\nForecast 6 - RMSE: 3.6790, ME: 3.2271, StdRes: 1.7836\n" ], [ "df_Fehler", "_____no_output_____" ] ], [ [ "<h3>Exponentielle Glättung mit 'Level', 'Trend' und 'Saisoneffekten' (Holt-Winters)</h3><br>\n<font size=4>\nDas Holt-Winters-Verfahren nennt man (spezielle) dreifache exponentielle Glättung:<br><br>\n\n\\begin{align*}\n\\text{Level: } \\; \\quad \\ell_t &= \\alpha \\frac{y_t}{s_{t-m}} + (1-\\alpha) (\\ell_{t-1} + b_{t-1}) \\\\\\\\\n\\text{Growth: } \\; \\quad b_t &= \\beta^* (\\ell_t - \\ell_{t-1}) + (1-\\beta^*) b_{t-1} \\\\\\\\\n\\text{Seasonal: } \\; \\quad s_t &= \\gamma \\frac{y_t}{\\ell_{t-1} + b_{t-1}} + (1-\\gamma) s_{t-m} \\\\\\\\\n\\text{Forecast: }\\; \\hat{y}_{t+h|t} &= (\\ell_t + b_t h ) s_{t-m+h_m^+} \\\\\n\\end{align*}\n</font>", "_____no_output_____" ] ], [ [ "#Schätzen von Holt-Winters\nfit7 = ExponentialSmoothing(df_train.iloc[:,0], seasonal_periods=5, trend='add', seasonal='add').fit()\nfcast7 = fit7.forecast(len(df_test)).rename(\"Holt-Winters Additive\")\n\nfit8 = ExponentialSmoothing(df_train.iloc[:,0], seasonal_periods=5, trend='add', seasonal='mul').fit()\nfcast8 = fit8.forecast(len(df_test)).rename(\"Holt-Winters Multiplikativ\")\n", "_____no_output_____" ], [ "#Plotten der Ergebnisse\n\n%matplotlib inline\nplt.rcParams['figure.figsize'] = [15, 6]\n\nfcast7.plot(legend=True, color=colors[1], ls='--')\nfit7.fittedvalues.plot(color=colors[2])\n\nfcast8.plot(legend=True, color=colors[3], ls='--')\nfit8.fittedvalues.plot(color=colors[4])\n\nplt.title('Anpassung der dreifachen exponentiellen Glättung auf {}'.format(df_train.columns[0]), fontsize=14)\nplt.xlabel('Zeit', fontsize=10)\nplt.ylabel('Preis in USD', fontsize=10)\n\nplt.show()", "_____no_output_____" ], [ "rmse = sqrt(mean_squared_error(df_test, fcast7.values))\nme = (df_test.iloc[:,0] - fcast7.values).sum() / len(df_test)\nresiduen = (df_test.iloc[:,0] - fcast7.values)\nstdres = residuen.std()\nprint('Forecast 7 - RMSE: {0:.4f}, ME: {1:.4f}, StdRes: {2:.4f}'.format(rmse, me, stdres))\n\ndf_Fehler = df_Fehler.append({'Methode': 'Holt Winters add', 'ME': me, 'RMSE': rmse, 'StdRes': stdres},\n ignore_index=True)\n\nrmse = sqrt(mean_squared_error(df_test, fcast8.values))\nme = (df_test.iloc[:,0] - fcast8.values).sum() / len(df_test)\nresiduen = (df_test.iloc[:,0] - fcast8.values)\nstdres = residuen.std()\nprint('Forecast 8 - RMSE: {0:.4f}, ME: {1:.4f}, StdRes: {2:.4f}'.format(rmse, me, stdres))\n\ndf_Fehler = df_Fehler.append({'Methode': 'Holt Winters mult', 'ME': me, 'RMSE': rmse, 'StdRes': stdres},\n ignore_index=True)\n", "Forecast 7 - RMSE: 2.3152, ME: 1.6727, StdRes: 1.6160\nForecast 8 - RMSE: 2.3310, ME: 1.6947, StdRes: 1.6158\n" ], [ "df_Fehler", "_____no_output_____" ] ], [ [ "## Faktor-Dekomposition\n### Wie bekomme ich heraus, ob und welche Saisonalität in den Daten vorliegt?", "_____no_output_____" ] ], [ [ "%matplotlib inline\nplt.rcParams['figure.figsize'] = [15, 8]\nsm.tsa.seasonal_decompose(df_train).plot()\nplt.show()", "_____no_output_____" ] ], [ [ "## Autoregressive Prozesse AR, ARMA und ARIMA", "_____no_output_____" ], [ "<font size=4>\nTipp: Eine Einführung in ARIMA-Modelle findet sich hier:<br>\n \nhttps://towardsdatascience.com/unboxing-arima-models-1dc09d2746f8\n\n</font>", "_____no_output_____" ], [ "<h3>Autoregressive Prognose-Modelle</h3><br>\n<font size=4>\nBei ARIMA-Modellen wird eine Zeitreihe durch einen verzögerten Term der zu erklärenden Variable (AR) und einen Moving-Average-Term (MA) erklärt. Unter bestimmten Umständen muss das Modell insgesamt als integrierte Gleichung (I; ein- oder mehrfach differenziert) optimiert werden.<br><br>\nDafür gilt es, die optimale Anzahl an Verzögerungen (p), das optimale Momentum (q) und, ggf., das optimale Differenz-Niveau (d) gefunden werden.<br>\n\n\\begin{align*}\n\\text{Forecast:} \\; \\quad \\hat{y}_{t} &= \\mu + \\phi_{1}y{t-1}+\\cdots+\\phi_{p}y_{t-p}-\\theta_{1}\\epsilon_{t-1}-\\cdots-\\theta_{q}\\epsilon_{t-q} \\\\\\\\\n\\text{Für d = 0:} \\; \\quad y_t &= Y_t \\\\\\\\\n\\text{Für d = 1:} \\; \\quad y_t &= Y_t - Y_{t-1} \\\\\\\\\n\\text{Für d = 2:} \\; \\quad y_t &= (Y_t - Y_{t-1})-(Y_{t-1} - Y_{t-2})\n\\end{align*}\n</font>", "_____no_output_____" ] ], [ [ "from statsmodels.tsa.arima_model import ARIMA\nwarnings.filterwarnings(\"ignore\") # specify to ignore warning messages", "_____no_output_____" ], [ "#Eigene Auto-Arima-Funktion\n#Erzeugen einer Liste von allen Parametern, die getestet werden sollen.\np = range(0, 6)\nd = range(0, 3)\nq = range(0, 5)\npdq = list(itertools.product(p, d, q))\npdq", "_____no_output_____" ], [ "ts = df_train.copy()", "_____no_output_____" ], [ "AIC = []\nARIMA_model = []\ni = 0\n\nfor param in pdq:\n try:\n mod = ARIMA(ts, order=param)\n results = mod.fit()\n print() \n print(i, \": \",'ARIMA{} - AIC:{}'.format(param, results.aic), end='\\r')\n AIC.append(results.aic)\n ARIMA_model.append([param])\n except:\n continue\n i = i + 1", "\n0 : ARIMA(0, 0, 0) - AIC:7013.454578932967\n1 : ARIMA(0, 0, 1) - AIC:5771.249334363882\n2 : ARIMA(0, 1, 0) - AIC:2151.2132474557666\n3 : ARIMA(0, 1, 1) - AIC:2152.659153522236\n4 : ARIMA(0, 1, 2) - AIC:2154.6282701671657\n5 : ARIMA(0, 1, 3) - AIC:2154.2940952408876\n6 : ARIMA(0, 1, 4) - AIC:2156.2916346441207\n7 : ARIMA(0, 2, 0) - AIC:2855.776649047264\n8 : ARIMA(0, 2, 1) - AIC:2155.557809569962\n9 : ARIMA(0, 2, 2) - AIC:2156.873683836948\n10 : ARIMA(0, 2, 3) - AIC:2158.8664849005554\n11 : ARIMA(0, 2, 4) - AIC:2158.185413456307\n12 : ARIMA(1, 0, 0) - AIC:2161.715398545889\n13 : ARIMA(1, 1, 0) - AIC:2152.652069831722\n14 : ARIMA(1, 1, 1) - AIC:2154.05849222667\n15 : ARIMA(1, 1, 2) - AIC:2156.00943161475\n16 : ARIMA(1, 1, 3) - AIC:2156.283781964605\n17 : ARIMA(1, 1, 4) - AIC:2158.1436481289193\n18 : ARIMA(1, 2, 0) - AIC:2554.3582284510103\n19 : ARIMA(1, 2, 1) - AIC:2156.868439790889\n20 : ARIMA(1, 2, 2) - AIC:2158.318046812955\n21 : ARIMA(1, 2, 4) - AIC:2159.039955302013\n22 : ARIMA(2, 0, 0) - AIC:2163.274614068158\n23 : ARIMA(2, 1, 0) - AIC:2154.6011925086386\n24 : ARIMA(2, 1, 1) - AIC:2156.015511628101\n25 : ARIMA(2, 1, 2) - AIC:2156.9449701306753\n26 : ARIMA(2, 1, 3) - AIC:2157.324687034196\n27 : ARIMA(2, 1, 4) - AIC:2159.308318725509\n28 : ARIMA(2, 2, 0) - AIC:2460.2116081875092\n29 : ARIMA(2, 2, 1) - AIC:2158.8482822691212\n30 : ARIMA(2, 2, 4) - AIC:2156.52216639959\n31 : ARIMA(3, 0, 0) - AIC:2165.1799595355556\n32 : ARIMA(3, 1, 0) - AIC:2154.555067840616\n33 : ARIMA(3, 1, 1) - AIC:2156.5490428831517\n34 : ARIMA(3, 1, 2) - AIC:2157.348646689716\n35 : ARIMA(3, 1, 3) - AIC:2153.1014330217354\n36 : ARIMA(3, 1, 4) - AIC:2154.1306360821527\n37 : ARIMA(3, 2, 0) - AIC:2387.4182292152723\n38 : ARIMA(3, 2, 1) - AIC:2158.5533428297467\n39 : ARIMA(3, 2, 2) - AIC:2158.862946838637\n40 : ARIMA(3, 2, 3) - AIC:2159.023606585547\n41 : ARIMA(3, 2, 4) - AIC:2163.844070173267\n42 : ARIMA(4, 0, 0) - AIC:2165.367570202199\n43 : ARIMA(4, 1, 0) - AIC:2156.5530637506677\n44 : ARIMA(4, 1, 1) - AIC:2157.5206766635756\n45 : ARIMA(4, 1, 2) - AIC:2159.348635885285\n46 : ARIMA(4, 1, 3) - AIC:2154.070742005638\n47 : ARIMA(4, 1, 4) - AIC:2147.873205912918\n48 : ARIMA(4, 2, 0) - AIC:2351.2852861599386\n49 : ARIMA(4, 2, 1) - AIC:2160.5327629534695\n50 : ARIMA(4, 2, 2) - AIC:2160.8313532391257\n51 : ARIMA(4, 2, 3) - AIC:2162.708666584628\n52 : ARIMA(4, 2, 4) - AIC:2157.925120733335\n53 : ARIMA(5, 0, 0) - AIC:2167.3500849145366\n54 : ARIMA(5, 1, 0) - AIC:2158.134017272393\n55 : ARIMA(5, 1, 1) - AIC:2158.0045903186533\n56 : ARIMA(5, 1, 2) - AIC:2157.36639751832\n57 : ARIMA(5, 1, 3) - AIC:2155.7195524462245\n58 : ARIMA(5, 1, 4) - AIC:2152.439435093944\n59 : ARIMA(5, 2, 0) - AIC:2337.485641686427\n60 : ARIMA(5, 2, 1) - AIC:2161.9573835202627\r" ], [ "print('Das AIC nimmt mit {} für das Modell ARIMA{} den kleinsten Wert an.'.format(min(AIC), ARIMA_model[AIC.index(min(AIC))][0]))", "Das AIC nimmt mit 2147.873205912918 für das Modell ARIMA(4, 1, 4) den kleinsten Wert an.\n" ], [ "mod = ARIMA(ts,order=ARIMA_model[AIC.index(min(AIC))][0])\nfit9 = mod.fit()", "_____no_output_____" ], [ "#Achtung, wenn d != 0, dann werden Differenzen vorhergesagt.\nfcast9 = fit9.predict(start=len(ts), end=len(ts)+len(df_test)-1, dynamic=False)\nfcast9.head()", "_____no_output_____" ], [ "y_hat_ = fcast9.copy()\ny_hat_[0] = y_hat_[0] + ts.iloc[-1][0]\ny_hat_ = np.cumsum(y_hat_.values)\nfcast9[:] = y_hat_", "_____no_output_____" ], [ "#Plotten der Ergebnisse\n\n%matplotlib inline\nplt.rcParams['figure.figsize'] = [15, 6]\n\nplt.plot(df_train.index, df_train.values, label='Trainingsdaten')\nplt.plot(fcast9.index, fcast9.values, label='ARIMA', color=colors[2])\n\nplt.legend(loc=2)\nplt.title('Anpassung von autoregressiven Modellen auf {}'.format(df_train.columns[0]), fontsize=14)\nplt.xlabel('Zeit', fontsize=10)\nplt.ylabel('Preis in USD', fontsize=10)\n\nplt.show()", "_____no_output_____" ], [ "rmse = sqrt(mean_squared_error(df_test, fcast9.values))\nme = (df_test.iloc[:,0] - fcast9.values).sum() / len(df_test)\nresiduen = (df_test.iloc[:,0] - fcast9.values)\nstdres = residuen.std()\nprint('Forecast 7 - RMSE: {0:.4f}, ME: {1:.4f}, StdRes: {2:.4f}'.format(rmse, me, stdres))\n\ndf_Fehler = df_Fehler.append({'Methode': 'ARIMA(4,1,4)', 'ME': me, 'RMSE': rmse, 'StdRes': stdres},\n ignore_index=True)", "Forecast 7 - RMSE: 2.6475, ME: 2.0076, StdRes: 1.7424\n" ], [ "df_Fehler", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7cc8c1fd5ecc6d20178a279e72bb988ea8523e

68,828

ipynb

Jupyter Notebook

climate_starter.ipynb

extasy44/sqlalchemy-challenge

6940b3fc35dac3b81800dd9f1621546c0c1c8091

[ "ADSL" ]

null

climate_starter.ipynb

extasy44/sqlalchemy-challenge

6940b3fc35dac3b81800dd9f1621546c0c1c8091

[ "ADSL" ]

null

climate_starter.ipynb

extasy44/sqlalchemy-challenge

6940b3fc35dac3b81800dd9f1621546c0c1c8091

[ "ADSL" ]

null

141.330595

31,212

0.886776

[ [ [ "%matplotlib inline\nfrom matplotlib import style\nstyle.use('fivethirtyeight')\nimport matplotlib.pyplot as plt", "_____no_output_____" ], [ "import numpy as np\nimport pandas as pd\nimport datetime as dt", "_____no_output_____" ] ], [ [ "# Reflect Tables into SQLAlchemy ORM", "_____no_output_____" ] ], [ [ "# Python SQL toolkit and Object Relational Mapper\nimport sqlalchemy\nfrom sqlalchemy.ext.automap import automap_base\nfrom sqlalchemy.orm import Session\nfrom sqlalchemy import create_engine, func", "_____no_output_____" ], [ "# create engine to hawaii.sqlite\nengine = create_engine(\"sqlite:///Resources/hawaii.sqlite\")", "_____no_output_____" ], [ "# reflect an existing database into a new model\nBase = automap_base()\n# reflect the tables\nBase.prepare(engine, reflect=True)", "_____no_output_____" ], [ "# View all of the classes that automap found\nBase.classes.keys()", "_____no_output_____" ], [ "# Save references to each table\nMeasurement = Base.classes.measurement\nStation = Base.classes.station", "_____no_output_____" ], [ "# Create our session (link) from Python to the DB\nsession = Session(engine)", "_____no_output_____" ] ], [ [ "# Exploratory Precipitation Analysis", "_____no_output_____" ] ], [ [ "# Find the most recent date in the data set.\nrecent_date = session.query(Measurement.date).\\\norder_by(Measurement.date.desc()).first().date\n\nrecent_date", "_____no_output_____" ], [ "# Design a query to retrieve the last 12 months of precipitation data and plot the results. \n# Starting from the most recent data point in the database. \n\n# Calculate the date one year from the last date in data set.\nyear_ago = dt.datetime.strptime(recent_date, '%Y-%m-%d') - dt.timedelta(days=365)\nprint(year_ago)\n\n# Perform a query to retrieve the data and precipitation scores\nprcp_res = session.query(Measurement.date, Measurement.prcp).\\\n filter(Measurement.date >= year_ago).all()\n\n# Save the query results as a Pandas DataFrame and set the index to the date column\nprcp_df = pd.DataFrame(prcp_res, columns=['Date', 'Precipitation'])\nprcp_df = prcp_df[prcp_df.Precipitation != 'NaN']\nprcp_df['Date']= pd.to_datetime(prcp_df['Date'])\n\n# Sort the dataframe by date\nprcp_df_idx=prcp_df.set_index('Date')\n\n# Use Pandas Plotting with Matplotlib to plot the data\nprcp_df_idx.plot(kind=\"line\")\nplt.title(\"Precipitation in 2016-09 ~ 2017-09\", size=15)\nplt.legend(bbox_to_anchor=(0.52,0.86), fontsize=\"16\")\nplt.xlabel(\"Date\", size=10)\nplt.ylabel(\"Precipitation (Inches)\", size=12)\nplt.show()\n\n", "2016-08-23 00:00:00\n" ], [ "# Use Pandas to calcualte the summary statistics for the precipitation data\nprcp_df.describe()", "_____no_output_____" ] ], [ [ "## Exploratory Station Analysis", "_____no_output_____" ] ], [ [ "# Design a query to calculate the total number stations in the dataset\ntotal_station = session.query(Station.id).count()\ntotal_station", "_____no_output_____" ], [ "# Design a query to find the most active stations (i.e. what stations have the most rows?)\n# List the stations and the counts in descending order.\n\nactive_stations = session.query(Measurement.station, func.count(Measurement.station)).\\\n group_by(Measurement.station).\\\n order_by(func.count(Measurement.station).desc()).all()\nactive_stations", "_____no_output_____" ], [ "# Using the most active station id from the previous query, calculate the lowest, highest, and average temperature.\nmost_active_station = active_stations[0][0]\n\nsession.query(func.min(Measurement.tobs), func.avg(Measurement.tobs), func.max(Measurement.tobs)).\\\n filter(Measurement.station == most_active_station).all()", "_____no_output_____" ], [ "# Using the most active station id\n# Query the last 12 months of temperature observation data for this station and plot the results as a histogram\ntemp_result = session.query(Measurement.station, Measurement.tobs).\\\n filter(Measurement.station == most_active_station).\\\n filter(Measurement.date >= year_ago).all()\ntobs_df = pd.DataFrame(temp_result)\ntobs_df.head()\n\ntobs = tobs_df['tobs']\nplt.hist(tobs, bins=12, alpha=0.5, histtype='stepfilled', color='steelblue', edgecolor='none')\nplt.xlabel('Tobs')\nplt.ylabel('Frequency')\nplt.title(\"Temperture Observation Data [Station: USC00519281]\")\nplt.show()", "_____no_output_____" ] ], [ [ "# Close session", "_____no_output_____" ] ], [ [ "# Close Session\nsession.close()", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code" ] ]

ec7cd074ae83a4cb87d9c3185176ec3a6dbff4a6

19,232

ipynb

Jupyter Notebook

nesh/02_customize_your_cluster.ipynb

ExaESM-WP4/Dask-jobqueue-configs

2bbada956455b1701424fcadb4c2ca106c4208d7

[ "MIT" ]

1

2020-08-24T11:52:31.000Z

nesh/02_customize_your_cluster.ipynb

ExaESM-WP4/dask-jobqueue-configs

2bbada956455b1701424fcadb4c2ca106c4208d7

[ "MIT" ]

7

2020-01-24T18:00:18.000Z

2020-02-14T16:47:10.000Z

nesh/02_customize_your_cluster.ipynb

ExaESM-WP4/Dask-jobqueue-configs

2bbada956455b1701424fcadb4c2ca106c4208d7

[ "MIT" ]

null

51.013263

1,324

0.515599

[ [ [ "# How to customize your Dask jobqueue cluster\n\ncovers the following aspects, i.e. how to\n* choose a NEC Linux cluster queue\n* adjust Dask jobqueue worker resources\n* scale Dask clusters adaptively", "_____no_output_____" ], [ "## Choose a NEC Linux cluster queue\n\nThe [NEC Linux cluster system](https://www.rz.uni-kiel.de/de/angebote/hiperf/nec-linux-cluster) has a theoretical number of 198 batch nodes with 24-32 CPUs and 128-384 GB memory available per execution host. Many use cases for Dask jobqueue workers require only rather short lifetimes and workers would survive with the walltime limits set for the `clexpress` batch queue. However, the very limited node/execution host number might lead to very unpredictable Dask worker job execution starting times ([note the spiky workload tendency](https://nbviewer.jupyter.org/github/ExaESM-WP4/nesh-monitoring/blob/v2020.01.24.1/analysis.ipynb)), which is not good for a default configuration. Therefore, the batch class `clmedium` has been chosen as default. A [relatively stable number of 200-400 idle CPUs](https://nbviewer.jupyter.org/github/ExaESM-WP4/nesh-monitoring/blob/v2020.01.24.1/analysis.ipynb) (at least during the considered log period!) is observed and (not too big!) Dask jobqueue clusters should always immediately start/connect. However, as available resources generally fluctuate with present user behaviour (which might change!), below you find an example on how to get an overview on currently available CPU resources and how to set up your Dask jobqueue cluster in a different than the default batch class.", "_____no_output_____" ], [ "### Get details on available batch class resources\n\nTher [recommended way](https://www.rz.uni-kiel.de/de/angebote/hiperf/nec-linux-cluster) of getting an overview on available batch queue resources is via the `qcl` command.", "_____no_output_____" ] ], [ [ "!qcl", " Batch class Walltime[h] Max.cores/node Max.RAM[gb] Total[*] Used[*] Avail[*] Run.jobs/user\n ----------- ----------- -------------- ----------- -------- ------- -------- -------------\n clexpress 2 32 192 2 2 0 -\n clmedium 48 32 192 116 116 0 20\n cllong 100 32 192 49 49 0 10\n clbigmem 200 32 384 8 8 0 2\n clfo2 200 24 128 17 17 0 8\n feque 1 32 64 1 1 0 -\n ----------- ----------- -------------- ----------- -------- ------- -------- -------------\n Sum 193 193 0\n ----------- ----------- -------------- ----------- -------- ------- -------- -------------\n [*] = number of neshcl### nodes\n" ] ], [ [ "However, the NEC Linux cluster is a shared host/node system and `qcl` doesn't provide useful details on actually allocated resources. The following `qstat` command lists free resources of the currently active nodes/execution hosts grouped alphabetically by batch queue class, filtered for nodes that have at least one unoccupied CPU. It gives an overview on theoretically available resources and might help with choosing a batch queue for your Dask jobqueue cluster. Please note, that each batch class has its own running job number limit that constrains the total number of Dask workers and hence your Dask cluster size.", "_____no_output_____" ] ], [ [ "%%bash\n\n# provide header\nqstat -E -F fcpu,fmem1,ucpu,umem1,ehost,quenm | head -n2\n\n# generate list of currently available resources with memory units in GB\nqstat -E -F fcpu,fmem1,ucpu,umem1,ehost,quenm \\\n | awk '{ if ($1>0 && substr($5,1,6)==\"neshcl\") { printf \"%9d%11.3f%10d%11.3f %-15s %-15s\\n\",$1,$2/2**8,$3,$4/2**8,$5,$6 } }' \\\n | sort -k6\n\n# # uncomment to generate list with raw output\n# qstat -E -F fcpu,fmem1,ucpu,umem1,ehost,quenm \\\n# | awk '{ if ($1>0 && substr($5,1,6)==\"neshcl\") { print } }' \\\n# | sort -k6\n", "Free_CPUs Free_Mem1 Used_CPUs Used_Mem1 ExecutionHost QueueName \n--------- ---------- --------- ---------- --------------- ---------------\n 16 334.586 16 48.125 neshcl219 clbigmem \n 30 363.465 2 19.246 neshcl224 clbigmem \n 30 373.582 2 9.129 neshcl225 clbigmem \n 31 376.113 1 6.598 neshcl220 clbigmem \n 31 377.371 1 5.340 neshcl221 clbigmem \n 3 110.418 21 17.477 neshcl216 clfo2 \n 4 101.266 20 26.629 neshcl209 clfo2 \n 4 101.629 20 26.266 neshcl211 clfo2 \n 4 101.742 20 26.152 neshcl201 clfo2 \n 4 102.043 20 25.852 neshcl207 clfo2 \n 4 102.625 20 25.270 neshcl217 clfo2 \n 4 105.160 20 22.734 neshcl215 clfo2 \n 4 105.688 20 22.207 neshcl208 clfo2 \n 4 105.863 20 22.031 neshcl203 clfo2 \n 4 109.965 20 17.930 neshcl200 clfo2 \n 4 110.223 20 17.672 neshcl213 clfo2 \n 10 165.035 22 25.676 neshcl364 cllong \n 10 165.879 22 24.832 neshcl348 cllong \n 10 169.770 22 20.941 neshcl352 cllong \n 12 161.445 20 29.266 neshcl367 cllong \n 12 165.141 20 25.570 neshcl378 cllong \n 12 165.211 20 25.500 neshcl392 cllong \n 12 168.848 20 21.863 neshcl371 cllong \n 13 156.832 19 33.879 neshcl380 cllong \n 13 173.480 19 17.230 neshcl360 cllong \n 14 142.973 18 47.738 neshcl376 cllong \n 14 177.219 18 13.492 neshcl386 cllong \n 25 143.242 7 47.469 neshcl390 cllong \n 27 165.004 5 25.707 neshcl368 cllong \n 27 186.824 5 3.887 neshcl356 cllong \n 28 153.117 4 37.594 neshcl354 cllong \n 28 181.199 4 9.512 neshcl387 cllong \n 30 154.660 2 36.051 neshcl363 cllong \n 30 170.984 2 19.727 neshcl396 cllong \n 30 173.418 2 17.293 neshcl350 cllong \n 31 156.836 1 33.875 neshcl366 cllong \n 4 160.504 28 30.207 neshcl351 cllong \n 6 169.426 26 21.285 neshcl395 cllong \n 6 178.137 26 12.574 neshcl374 cllong \n 7 182.180 25 8.531 neshcl393 cllong \n 10 170.605 22 20.105 neshcl252 clmedium \n 12 162.438 20 28.273 neshcl300 clmedium \n 12 163.422 20 27.289 neshcl267 clmedium \n 12 164.902 20 25.809 neshcl268 clmedium \n 12 165.273 20 25.438 neshcl260 clmedium \n 12 165.453 20 25.258 neshcl303 clmedium \n 12 167.855 20 22.855 neshcl280 clmedium \n 12 168.238 20 22.473 neshcl293 clmedium \n 12 168.281 20 22.430 neshcl257 clmedium \n 12 170.238 20 20.473 neshcl230 clmedium \n 12 170.602 20 20.109 neshcl253 clmedium \n 12 170.672 20 20.039 neshcl332 clmedium \n 12 170.727 20 19.984 neshcl305 clmedium \n 12 171.816 20 18.895 neshcl277 clmedium \n 14 169.398 18 21.312 neshcl310 clmedium \n 22 169.422 10 21.289 neshcl279 clmedium \n 32 173.406 0 17.305 neshcl231 clmedium \n 32 188.000 0 2.711 neshcl313 clmedium \n 4 4.082 28 186.629 neshcl341 clmedium \n 5 122.023 27 68.688 neshcl301 clmedium \n 6 3.613 26 187.098 neshcl287 clmedium \n 6 4.855 26 185.855 neshcl299 clmedium \n 6 98.504 26 92.207 neshcl284 clmedium \n 7 100.797 25 89.914 neshcl292 clmedium \n 7 3.625 25 187.086 neshcl250 clmedium \n 7 75.750 25 114.961 neshcl294 clmedium \n 8 139.859 24 50.852 neshcl334 clmedium \n 8 3.680 24 187.031 neshcl262 clmedium \n 32 184.520 0 6.191 neshcl343 cltestque \n 32 179.602 0 11.109 neshcl342 cltestque,clinteractive\n" ] ], [ [ "### Start Dask jobqueue cluster in a different batch class\nManually choosing a batch queue for your Dask jobqueue cluster is as simple as using the default configuration and calling the `PBSCluster` method with the `queue` keyword argument alone. This will override the selection of the queue, but will follow all other aspects of the default config.", "_____no_output_____" ] ], [ [ "import os; os.environ['DASK_CONFIG']='.'\nimport dask.config; dask.config.get('jobqueue')", "_____no_output_____" ], [ "import dask_jobqueue\ncustom_queue_cluster = dask_jobqueue.PBSCluster(\n config_name='nesh-jobqueue-config',\n queue='clbigmem'\n)", "_____no_output_____" ], [ "print(custom_queue_cluster.job_script())", "#!/bin/bash\n\n#PBS -N dask-worker\n#PBS -q clbigmem\n#PBS -l elapstim_req=00:45:00,cpunum_job=4,memsz_job=24gb\n#PBS -o dask_jobqueue_logs/dask-worker.o%s\n#PBS -e dask_jobqueue_logs/dask-worker.e%s\nJOB_ID=${PBS_JOBID%%.*}\n\n/sfs/fs6/home-geomar/smomw260/miniconda3/envs/dask-minimal-20191218/bin/python -m distributed.cli.dask_worker tcp://192.168.31.10:32866 --nthreads 4 --memory-limit 24.00GB --name name --nanny --death-timeout 60 --local-directory /scratch --interface ib0\n\n" ] ], [ [ "## Adjust Dask jobqueue worker resources\nGetting details on currently available batch class resources might also help in guiding your choice on single Dask worker resources. The default values are chosen to allocate one eighth (a subjective choice!) of the average `clmedium` batch class host resources. However, depending on your dataset characteristics, the calculation/operation properties and the currently occupied Linux cluster resources, other values might be more useful.\n\nPlease note, that available [resources at the NEC Linux cluster system are mostly CPU limited, with plenty of memory available](https://nbviewer.jupyter.org/github/ExaESM-WP4/nesh-monitoring/blob/v2020.01.24.1/analysis.ipynb). Below you therefore find an example of how to decrease single Dask worker resources in terms of CPU allocation. Such Dask workers might occasionally better squeeze into the available NEC Linux cluster resources. In terms of total available Dask cluster resources, you can compensate the smaller single Dask worker size by increasing the total number of Dask jobqueue workers, i.e. using a larger value for the `jobs` keyword in `cluster.scale(jobs=4)`.\n\nHowever, this approach might both increase or decrease total execution time of your calculation, depending on the task types associated with your operation and the actual load on the node/execution host. If you play around with jobqueue worker resources (and total worker numbers) never forget that your total Dask cluster size is not only limited by the available resources and the batch class running job number limit, but also by the resources demanded by other already queued jobs (which influence the starting times of your Dask jobqueue workers!).", "_____no_output_____" ] ], [ [ "import os; os.environ['DASK_CONFIG']='.'\nimport dask.config; dask.config.get('jobqueue')", "_____no_output_____" ], [ "import dask_jobqueue\ncustom_worker_cluster = dask_jobqueue.PBSCluster(\n config_name='nesh-jobqueue-config',\n # Define smaller Linux cluster resources per Dask worker.\n resource_spec='elapstim_req=00:45:00,cpunum_job=2,memsz_job=24gb',\n # Adjust Dask worker resource limits.\n cores=2\n)", "_____no_output_____" ], [ "print(custom_worker_cluster.job_script())", "#!/bin/bash\n\n#PBS -N dask-worker\n#PBS -q clmedium\n#PBS -l elapstim_req=00:45:00,cpunum_job=2,memsz_job=12gb\n#PBS -o dask_jobqueue_logs/dask-worker.o%s\n#PBS -e dask_jobqueue_logs/dask-worker.e%s\nJOB_ID=${PBS_JOBID%%.*}\n\n/sfs/fs6/home-geomar/smomw260/miniconda3/envs/dask-minimal-20191218/bin/python -m distributed.cli.dask_worker tcp://192.168.31.10:38568 --nthreads 2 --memory-limit 12.00GB --name name --nanny --death-timeout 60 --local-directory /scratch --interface ib0\n\n" ] ], [ [ "## Scale Dask clusters adaptively\nYou can operate a Dask jobqueue cluster using fixed scaling with the `cluster.scale(jobs=2)` method, or by using adaptive scaling with the `cluster.adapt(minimum=2, maximum=10)` method. For more examples on advanced scaling methods see for example [this Dask jobqueue workshop materials notebook](https://nbviewer.jupyter.org/github/willirath/dask_jobqueue_workshop_materials/blob/v1.1.0/notebooks/03_tuning_adaptive_clusters.ipynb).", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ] ]

ec7cd0c0e4ec75716eeacea8d62376b363b3efc0

4,120

ipynb

Jupyter Notebook

comparing-models.ipynb

Cheukting/legend_data

e348ce0188bbcf7c651c0c80f62f5110ba1ea886

[ "CC-BY-3.0" ]

3

2020-06-24T09:40:57.000Z

2022-01-10T23:59:08.000Z

comparing-models.ipynb

Cheukting/legend_data

e348ce0188bbcf7c651c0c80f62f5110ba1ea886

[ "CC-BY-3.0" ]

5

2021-06-08T22:10:08.000Z

2022-03-12T00:44:06.000Z

comparing-models.ipynb

Cheukting/legend_data

e348ce0188bbcf7c651c0c80f62f5110ba1ea886

[ "CC-BY-3.0" ]

1

2020-10-11T15:28:37.000Z

27.651007

84

0.580097

[ [ [ "# Roc Curve", "_____no_output_____" ] ], [ [ "# Author: Tim Head <[email protected]>\n#\n# License: BSD 3 clause\n\nimport numpy as np\nnp.random.seed(10)\n\nimport matplotlib.pyplot as plt\n\nfrom sklearn.datasets import make_classification\nfrom sklearn.linear_model import LogisticRegression\nfrom sklearn.ensemble import (RandomTreesEmbedding, RandomForestClassifier,\n GradientBoostingClassifier)\nfrom sklearn.preprocessing import OneHotEncoder\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import roc_curve\nfrom sklearn.pipeline import make_pipeline\n\nn_estimator = 2\nX, y = make_classification(n_samples=80000)\nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5)\n\n# Supervised transformation based on random forests\nrf = RandomForestClassifier(max_depth=3, n_estimators=n_estimator)\nrf.fit(X_train, y_train)\n\n# Supervised transformation based on gradient boosted trees\ngrd = GradientBoostingClassifier(n_estimators=n_estimator)\ngrd.fit(X_train, y_train)\n\n# The gradient boosted model by itself\ny_pred_grd = grd.predict_proba(X_test)[:, 1]\nfpr_grd, tpr_grd, _ = roc_curve(y_test, y_pred_grd)\n\n# The random forest model by itself\ny_pred_rf = rf.predict_proba(X_test)[:, 1]\nfpr_rf, tpr_rf, _ = roc_curve(y_test, y_pred_rf)\n\nplt.figure(1)\nplt.plot([0, 1], [0, 1], 'k--')\nplt.plot(fpr_rf, tpr_rf, label='RF')\nplt.plot(fpr_grd, tpr_grd, label='GBT')\nplt.xlabel('False positive rate')\nplt.ylabel('True positive rate')\nplt.title('ROC curve')\nplt.legend(loc='best')\nplt.show()\n\nplt.figure(2)\nplt.xlim(0, 0.2)\nplt.ylim(0.8, 1)\nplt.plot([0, 1], [0, 1], 'k--')\nplt.plot(fpr_rf, tpr_rf, label='RF')\nplt.plot(fpr_grd, tpr_grd, label='GBT')\nplt.xlabel('False positive rate')\nplt.ylabel('True positive rate')\nplt.title('ROC curve (zoomed in at top left)')\nplt.legend(loc='best')\nplt.show()", "_____no_output_____" ] ], [ [ "# Accuracy Score", "_____no_output_____" ] ], [ [ "from sklearn.metrics import accuracy_score\ny_pred_rf = rf.predict(X_test)\ny_pred_grd = grd.predict(X_test)\nrf_acc = accuracy_score(y_test, y_pred_rf)\ngrd_acc = accuracy_score(y_test, y_pred_grd)\nprint(f\"Accuracy for RF: {rf_acc}\")\nprint(f\"Accuracy for Grd: {grd_acc}\")\n", "_____no_output_____" ] ], [ [ "# Precision Score", "_____no_output_____" ] ], [ [ "from sklearn.metrics import precision_score\nrf_pre = precision_score(y_test, y_pred_rf)\ngrd_pre= precision_score(y_test, y_pred_grd)\nprint(f\"Precision for RF: {rf_pre}\")\nprint(f\"Precision for Grd: {grd_pre}\")", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

ec7cdc829fc1bb702e6d2399e012aa726a8f1764

17,429

ipynb

Jupyter Notebook

distributed/notebooks/01_intro_spark.ipynb

ssulca/dipdata

1872b21cb9adc11017a70fe9a44b5edc87f10f7e

[ "MIT" ]

1

2020-11-20T22:59:31.000Z

distributed/notebooks/01_intro_spark.ipynb

ser0090/dipdata

1872b21cb9adc11017a70fe9a44b5edc87f10f7e

[ "MIT" ]

null

distributed/notebooks/01_intro_spark.ipynb

ser0090/dipdata

1872b21cb9adc11017a70fe9a44b5edc87f10f7e

[ "MIT" ]

1

2021-07-20T14:02:21.000Z

29.641156

234

0.540421

[ [ [ "\n<center>\n <h1><a href=\"http://diplodatos.famaf.unc.edu.ar/\">Diplomatura en Ciencia de Datos, Aprendizaje Automático y sus Aplicaciones</a></h1>\n <h2>Curso <a href=\"https://sites.google.com/view/eleccion-optativas-diplodatos/programaci%C3%B3n-distribu%C3%ADda-sobre-grandes-vol%C3%BAmenes-de-datos\">Programación Distribuida sobre Grandes Volúmenes de Datos</a></h2>\n</center>\n\n<br>\n\n<h3 style=\"text-align:center;\"> Damián Barsotti </h3>\n\n<h3 style=\"text-align:center;\">\n <a href=\"http://www.famaf.unc.edu.ar\">\n Facultad de Matemática Astronomía Física y Computación\n </a>\n<br/>\n <a href=\"http://www.unc.edu.ar\">\n Universidad Nacional de Córdoba\n </a>\n<br/>\n <center>\n <a href=\"http://www.famaf.unc.edu.ar\">\n <img src=\"https://www.famaf.unc.edu.ar/static/assets/logosFooterBottom.svg\" alt=\"Drawing\" style=\"width:50%;\"/>\n </a>\n </center>\n</h3>\n\n<p style=\"font-size:15px;\">\n <br />\n This work is licensed under a\n <a rel=\"license\" href=\"http://creativecommons.org/licenses/by-nc-sa/4.0/\">Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License</a>.\n <a rel=\"license\" href=\"http://creativecommons.org/licenses/by-nc-sa/4.0/\">\n <img alt=\"Creative Commons License\" style=\"border-width:0;vertical-align:middle;float:right\" src=\"https://i.creativecommons.org/l/by-nc-sa/4.0/88x31.png\" />\n </a>\n</p>\n", "_____no_output_____" ], [ "# Introducción a Spark\n---", "_____no_output_____" ], [ "## Características\n\n* 100x más rápido que Hadoop MapReduce en memoria.\n* 10x más rápido en disco.\n\n ![](https://bitbucket.org/bigdata_famaf/diplodatos_bigdata/raw/HEAD/clases/01_intro_spark/spark_speed.png)\n \n* Multiplataforma\n * Corre en Hadoop Yarn, Mesos, standalone o en la nube (AWS, Azure, ...)\n * Acceso a datos en HDFS, Cassandra, HBase, Hive, Tachyon, JDBC, etc.\n![](https://bitbucket.org/bigdata_famaf/diplodatos_bigdata/raw/HEAD/clases/01_intro_spark/spark_multi_plataforma.png)", "_____no_output_____" ], [ "## Múltiples funcionalidades en una plataforma (Stack unificado)\n\n<img src=\"https://bitbucket.org/bigdata_famaf/diplodatos_bigdata/raw/b17129f7118b3389b8c7f2f85fd89c6238fe0edd/clases/01_intro_spark/unified_stack.png\" alt=\"Drawing\" style=\"width: 80%;\"/>", "_____no_output_____" ], [ "## Word Count (MapReduce)\n\n```java\npublic class WordCount {\n\tpublic static class Map extends MapReduceBase implements Mapper<LongWritable, Text, Text, IntWritable> {\n\n\t\tprivate final static IntWritable one = new IntWritable(1);\n\t\tprivate Text word = new Text();\n \n\t\tpublic void map(LongWritable key, Text value, OutputCollector<Text, IntWritable> output, Reporter reporter) throws IOException {\n\n\t\t\tString line = value.toString();\n\t\t\tStringTokenizer tokenizer = new StringTokenizer(line);\n\n\t\t\twhile (tokenizer.hasMoreTokens()) {\n\t\t\t\tword.set(tokenizer.nextToken());\n\t\t\t\toutput.collect(word, one);\n\t\t\t}\n\t\t}\n\t}\n\n\tpublic static class Reduce extends MapReduceBase implements Reducer<Text, IntWritable, Text, IntWritable> {\n\n\t\tpublic void reduce(Text key, Iterator values, OutputCollector<Text, IntWritable> output, Reporter reporter) throws IOException {\n\n\t\t\tint sum = 0;\n\t\t\twhile (values.hasNext()) {\n\t\t\t\tsum += values.next().get();\n\t\t\t}\n\t\t\toutput.collect(key, new IntWritable(sum));\n\t\t}\n\t}\n\n\tpublic static void main(String[] args) throws Exception {\n\n\t\tJobConf conf = new JobConf(WordCount.class);\n\t\tconf.setJobName(\"wordcount\");\n\t\tconf.setOutputKeyClass(Text.class);\n\t\tconf.setOutputValueClass(IntWritable.class);\n\t\tconf.setMapperClass(Map.class);\n\t\tconf.setCombinerClass(Reduce.class);\n\t\tconf.setReducerClass(Reduce.class);\n\t\tconf.setInputFormat(TextInputFormat.class);\n\t\tconf.setOutputFormat(TextOutputFormat.class);\n\t\tFileInputFormat.setInputPaths(conf, new Path(args[0]));\n\t\tFileOutputFormat.setOutputPath(conf, new Path(args[1]));\n\t\tJobClient.runJob(conf);\n\t}\n}\n```", "_____no_output_____" ] ], [ [ "from pyspark.sql import SparkSession", "_____no_output_____" ], [ "spark = SparkSession.builder.appName(\"01_intro\").getOrCreate()\nsc = spark.sparkContext", "_____no_output_____" ] ], [ [ "## Word Count (Spark)\n\n* `lines` es un **array distribuido** de lineas de texto (`RDD[str]`).\n - una parte del arreglo en cada **nodo del cluster**.\n\n* `lines` tiene el método `flatMap` (línea 6):\n - `flatMap(lambda line: line.split(\" \"))` toma cada cada elemento del `RDD` (linea), lo convierte en sequencia de palabras y concatena estas secuencias:\n - `lambda line: line.split(\" \")` es la **función** que toma una linea y la divide en una secuencia de palabras.\n \n - Su resultado es un array **distribuido** de palabras (`RDD[str]`).\n \n* Al resultado de `flatMap` se aplica el método `filter` (línea 7):\n - `filter(lambda word: word)` saca las palabras que son vacías (pueden aparecer?).\n - `lambda word: word` es la **función** que pregunta si la palabra es vacía.\n - `filter` devuelve un `RDD` que se almacena en `words`.\n\n* `words` tiene el método `map` (línea 11):\n - `map(lambda word: (word,1))` agrega a cada palabra de `words` un `1`.\n - El resultado es un **arreglo distribuido** de tuplas `RDD[(str, Int)]`.\n \n* A este `RDD` se le aplica el método `reduceByKey` (línea 12):\n - `reduceByKey(lambda n,m: n+m)` suma los `1`'s de las palabras iguales (la key es la palabra).\n", "_____no_output_____" ] ], [ [ "lines = sc.textFile(\"../inputs/README.md\")\n\nwords = lines \\\n .flatMap(lambda line: line.split(\" \")) \\\n .filter(lambda word: word)\n\n## filter-devuele true si no es vacia. \n#MapReduce\nwordCount = words \\\n .map(lambda word: (word,1)) \\\n .reduceByKey(lambda n,m: n+m)", "_____no_output_____" ] ], [ [ "## Resultado Word Count Spark", "_____no_output_____" ] ], [ [ "result = wordCount.sortBy((lambda p: p[1]), ascending = False) # ordena por cantidad\n\nlocal_result = result.collect() # Traigo desde cluster\n\nfor word, count in local_result[:10]: # tomo 10\n print(word, count) # los imprimo", "from 4\nApache 3\nZeppelin 3\nand 3\nto 3\n* 2\n### 2\nbinary 2\nPlease 2\n[User 2\n" ] ], [ [ "## RDD de entrada", "_____no_output_____" ] ], [ [ "inputRDD = sc.textFile(\"../inputs/logs/zeppelin-interpreter-spark-*-nabucodonosor.log\") \n# se crea un nuevo RDD:\nskarkRDD = inputRDD.filter(lambda line: \"Successfully started service 'SparkUI' on port \" in line) \n\nfor ln, l in enumerate(skarkRDD.collect()):\n print(\"Linea {}:\".format(ln), l)", "Linea 0: INFO [2020-10-31 10:15:43,582] ({pool-2-thread-3} Logging.scala[logInfo]:54) - Successfully started service 'SparkUI' on port 4044.\n" ] ], [ [ "## Ejercicio 0 (word count)\n* Crear una celda abajo de esta (poner mouse debajo de esta celda y seleccionar \"Add Paragraph\").\n* Copiar el programa `wordcount` anterior en la misma (esta en 2 celdas).\n - [`shift`]-[`flechas`] para seleccionar.\n - [`ctrl`]-[`c`] para copiar.\n - [`ctrl`]-[`v`] para pegar.\n* Modificarlo para leer todas la lineas de los archivos en `./licenses/`\n - Ayuda: si al método `textFile` se le indica el nombre de un directorio carga todos los archivo del mismo.\n* Ejecute la celda ([`shift`]-[`enter`])\n* Ver la cantidad de tareas en SparkUI", "_____no_output_____" ] ], [ [ "lines = sc.textFile(\"../inputs/licenses/\")\n\nwords = lines \\\n .flatMap(lambda line: line.split(\" \")) \\\n .filter(lambda word: word) \n \n## filter-devuele true si no es vacia. \n\n#MapReduce\nwordCount = words \\\n .map(lambda word: (word,1)) \\\n .reduceByKey(lambda n,m: n+m)\n\nlocal_result = wordCount.collect() # Traigo desde cluster\n\nfor word, count in local_result[:10]: # tomo 10\n print(word, count) # los imprimo", "ARE 17\nSERVICES; 17\ncommon 7\n(ii) 7\ncode, 10\n\"Not 5\nSubject 10\npublicly 10\nperform, 7\nlitigation 12\n" ] ], [ [ "Ver la cantidad de tareas en SparkUI - Rta: \n * Si no se ejecuta la instruccion `collect()` No se ejecuta ninguna tarea\n Estp es por que spark tiene el enfoque lazy, no evalua nada y no ejecuta nada\n hasta se necesite el resultado.\n * si se ejecuta, Utiliza 124/124 tareas ", "_____no_output_____" ], [ "## Ejecución de programas en Spark\n\n* En [Zeppelin](http://zeppelin.apache.org/) (como lo hacemos ahora)\n* En `pyspark` shell (tambien interactivo)\n* Como programa autónomo", "_____no_output_____" ], [ "### pyspark shell\n\n* Abrir una terminal\n* Ir a la instalación Spark\n```sh\ncd ~/spark/spark-2.3.4-bin-hadoop2.7\n```\n* Arrancar el shell\n```sh\n./bin/pyspark\n```\n* Escribir en shell (después apretar `Enter`)\n```python\n>>> lines = sc.textFile(\"README.md\")\n>>> lines.first()\n```", "_____no_output_____" ], [ "### Programa autónomo\n\n* Ir a programa\n```sh\ncd diplodatos_bigdata/prog/word_count\n```\n* Actualizar repositorio\n```sh\ngit pull\n```\n* Ver programa\n```sh\nless src/main/python/WordCount.py\n```\n (salir con [`q`])\n \n### Ejecucion de programa\n\n* Ejecutar\n```sh\n~/spark/spark-2.3.4-bin-hadoop2.7/bin/spark-submit --master local[4] src/main/python/WordCount.py \\\n ~/spark/spark-2.3.4-bin-hadoop2.7/licenses/\n```", "_____no_output_____" ], [ "### Versión Spark en Zeppelin", "_____no_output_____" ] ], [ [ "print(sc.version)", "3.0.1\n" ] ], [ [ "### Principales referencias online:\n\n* [Documentación Spark](http://spark.apache.org/docs/2.2.1/)\n* [API Spark Python](http://spark.apache.org/docs/2.2.1/api/python/index.html)", "_____no_output_____" ], [ "## Ejercicios MapReduce con Spark\n---", "_____no_output_____" ], [ "### Ejercicio 1\n\nModifique el programa *word count* siguiente para que cuente la **cantidad de apariciones de cada letra** en el archivo.\n\n* Ayuda: solo hay que modificar la linea 6", "_____no_output_____" ] ], [ [ "lines = sc.textFile(\"../inputs/README.md\")\n\nwords = lines \\\n .flatMap(lambda line: list(line)) \\\n .filter(lambda word: word != ' ')\n\n\n#MapReduce\nwordCount = words \\\n .map(lambda word: (word,1)) \\\n .reduceByKey(lambda n,m: n+m)\n\nresult = wordCount \\\n .sortBy((lambda p: p[1]), ascending=False) # ordena por cantidad\n\nlocal_result = result.collect() # Traigo desde cluster\n\nfor word, count in local_result[:10]: # tomo 10\n print(word, count) # los imprimo", "e 102\nt 92\na 83\ni 73\np 61\no 59\nn 58\ns 56\nl 54\nr 54\n" ] ], [ [ "### Ejercicio 2\n\nCada línea del archivo `~/diplodatos_bigdata/ds/links_raw.txt` contiene un url de una página web seguido de los links que posee a otras páginas web:\n```\n<url> <url link 1> <url link 2> ... <url link n>\n```\n\nBasándose en la utilización de la técnica de *MapReduce* que se mostró en el programa `word count` haga un programa en Spark que cuente la cantidad de links que apuntan a cada página.\n\n#### Ayuda\n\nA continuación está el comienzo del programa. Falta hacer el *MapReduce* y mostrar el resultado.\n", "_____no_output_____" ] ], [ [ "baseDir = \"../inputs\" # llenar con el directorio git\n\nlines = sc.textFile(baseDir + \"/ds/links_raw.txt\")\n\nlinksTo = lines \\\n .flatMap(lambda l: l.split(\" \")[1:]) # separo los links y tomo los apuntados\n\n# Ahora linksTo tiene las paginas apuntadas\n\n# Completar los ...\n\n# MapReduce\ninvLinkCount = linksTo.map(lambda link: (link,1)) \\\n .reduceByKey(lambda n,m: n+m)\n\nresult = invLinkCount.sortBy((lambda p: p[1]), ascending = False)\n\nlocal_result = result.collect() # Traigo desde cluster\n\nfor word, count in local_result[:10]: # tomo 10\n print(word, count) # los imprimo", "http://www.yahoo.com/ 199\nhttp://www.ca.gov/ 169\nhttp://www.leginfo.ca.gov/calaw.html 155\nhttp://www.linkexchange.com/ 134\nhttp://www.berkeley.edu/ 126\nhttp://www.sen.ca.gov/ 123\nhttp://home.netscape.com/comprod/mirror/index.html 109\nhttp://www.assembly.ca.gov/ 99\nhttp://www.epa.gov/ 95\nhttp://www.usgs.gov/ 84\n" ] ], [ [ "FIN", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ] ]

ec7ce47444f69641d5fdf7b7d8dd39610cc2fd2c

223,952

ipynb

Jupyter Notebook

examples/sensitivity_example_with_synthetic_data.ipynb

ibraaaa/causalml

340ff52488af5efd01d4a633107e231c88b8307d

[ "Apache-2.0" ]

2,919

2019-08-12T23:02:10.000Z

2022-03-31T21:59:34.000Z

examples/sensitivity_example_with_synthetic_data.ipynb

ibraaaa/causalml

340ff52488af5efd01d4a633107e231c88b8307d

[ "Apache-2.0" ]

317

2019-08-13T14:16:22.000Z

2022-03-26T08:44:06.000Z

examples/sensitivity_example_with_synthetic_data.ipynb

ibraaaa/causalml

340ff52488af5efd01d4a633107e231c88b8307d

[ "Apache-2.0" ]

466

2019-08-18T01:45:14.000Z

2022-03-31T08:11:53.000Z

91.934319

37,012

0.770888

[ [ [ "# Setup\nIn this notebook, we will generate some synthetic data to demonstrate how sensitivity analysis work by different methods.\n\n# Sensitivity Analysis\n## Methods\nWe provided five methods for sensitivity analysis including (Placebb Treatment, Random Cause, Subset Data, Random Replace and Selection Bias). \nThis notebook will walkthrough how to use the combined function sensitivity_analysis() to compare different method and also how to use each individual method separately:\n\n1. Placebo Treatment: Replacing treatment with a random variable\n2. Irrelevant Additional Confounder: Adding a random common cause variable\n3. Subset validation: Removing a random subset of the data\n4. Selection Bias method with One Sided confounding function and Alignment confounding function\n5. Random Replace: Random replace a covariate with an irrelevant variable", "_____no_output_____" ] ], [ [ "%matplotlib inline\n%load_ext autoreload\n%autoreload 2", "_____no_output_____" ], [ "import pandas as pd\nimport matplotlib.pyplot as plt\nfrom sklearn.linear_model import LinearRegression\nimport warnings\nimport matplotlib\nfrom causalml.inference.meta import BaseXLearner\nfrom causalml.dataset import synthetic_data\n\nfrom causalml.metrics.sensitivity import Sensitivity\nfrom causalml.metrics.sensitivity import SensitivityRandomReplace, SensitivitySelectionBias\n\nplt.style.use('fivethirtyeight')\nmatplotlib.rcParams['figure.figsize'] = [8, 8]\nwarnings.filterwarnings('ignore')\n\n# logging.basicConfig(level=logging.INFO)\n\npd.options.display.float_format = '{:.4f}'.format", "/Users/jing.pan/anaconda3/envs/causalml_3_6/lib/python3.6/site-packages/sklearn/utils/deprecation.py:144: FutureWarning: The sklearn.utils.testing module is deprecated in version 0.22 and will be removed in version 0.24. The corresponding classes / functions should instead be imported from sklearn.utils. Anything that cannot be imported from sklearn.utils is now part of the private API.\n warnings.warn(message, FutureWarning)\n" ] ], [ [ "# Data Pred", "_____no_output_____" ], [ "## Generate Synthetic data", "_____no_output_____" ] ], [ [ "# Generate synthetic data using mode 1\nnum_features = 6 \ny, X, treatment, tau, b, e = synthetic_data(mode=1, n=100000, p=num_features, sigma=1.0)", "_____no_output_____" ], [ "tau.mean()", "_____no_output_____" ] ], [ [ "## Define Features", "_____no_output_____" ] ], [ [ "# Generate features names\nINFERENCE_FEATURES = ['feature_' + str(i) for i in range(num_features)]\nTREATMENT_COL = 'target'\nOUTCOME_COL = 'outcome'\nSCORE_COL = 'pihat'", "_____no_output_____" ], [ "df = pd.DataFrame(X, columns=INFERENCE_FEATURES)\ndf[TREATMENT_COL] = treatment\ndf[OUTCOME_COL] = y\ndf[SCORE_COL] = e", "_____no_output_____" ], [ "df.head()", "_____no_output_____" ] ], [ [ "# Sensitivity Analysis", "_____no_output_____" ], [ "## With all Covariates", "_____no_output_____" ], [ "### Sensitivity Analysis Summary Report (with One-sided confounding function and default alpha)", "_____no_output_____" ] ], [ [ "# Calling the Base XLearner class and return the sensitivity analysis summary report\nlearner_x = BaseXLearner(LinearRegression())\nsens_x = Sensitivity(df=df, inference_features=INFERENCE_FEATURES, p_col='pihat',\n treatment_col=TREATMENT_COL, outcome_col=OUTCOME_COL, learner=learner_x)\n# Here for Selection Bias method will use default one-sided confounding function and alpha (quantile range of outcome values) input\nsens_sumary_x = sens_x.sensitivity_analysis(methods=['Placebo Treatment',\n 'Random Cause',\n 'Subset Data',\n 'Random Replace',\n 'Selection Bias'], sample_size=0.5)", "_____no_output_____" ], [ "# From the following results, refutation methods show our model is pretty robust; \n# When alpah > 0, the treated group always has higher mean potential outcomes than the control; when < 0, the control group is better off.\nsens_sumary_x", "_____no_output_____" ] ], [ [ "### Random Replace", "_____no_output_____" ] ], [ [ "# Replace feature_0 with an irrelevent variable\nsens_x_replace = SensitivityRandomReplace(df=df, inference_features=INFERENCE_FEATURES, p_col='pihat',\n treatment_col=TREATMENT_COL, outcome_col=OUTCOME_COL, learner=learner_x,\n sample_size=0.9, replaced_feature='feature_0')\ns_check_replace = sens_x_replace.summary(method='Random Replace')\ns_check_replace", "_____no_output_____" ] ], [ [ "### Selection Bias: Alignment confounding Function", "_____no_output_____" ] ], [ [ "sens_x_bias_alignment = SensitivitySelectionBias(df, INFERENCE_FEATURES, p_col='pihat', treatment_col=TREATMENT_COL,\n outcome_col=OUTCOME_COL, learner=learner_x, confound='alignment',\n alpha_range=None)", "_____no_output_____" ], [ "lls_x_bias_alignment, partial_rsqs_x_bias_alignment = sens_x_bias_alignment.causalsens()", "_____no_output_____" ], [ "lls_x_bias_alignment", "_____no_output_____" ], [ "partial_rsqs_x_bias_alignment", "_____no_output_____" ], [ "# Plot the results by confounding vector and plot Confidence Intervals for ATE\nsens_x_bias_alignment.plot(lls_x_bias_alignment, ci=True)", "_____no_output_____" ], [ "# Plot the results by rsquare with partial r-square results by each individual features\nsens_x_bias_alignment.plot(lls_x_bias_alignment, partial_rsqs_x_bias_alignment, type='r.squared', partial_rsqs=True)", "_____no_output_____" ] ], [ [ "## Drop One Confounder", "_____no_output_____" ] ], [ [ "df_new = df.drop('feature_0', axis=1).copy()\nINFERENCE_FEATURES_new = INFERENCE_FEATURES.copy()\nINFERENCE_FEATURES_new.remove('feature_0')\ndf_new.head()", "_____no_output_____" ], [ "INFERENCE_FEATURES_new", "_____no_output_____" ] ], [ [ "### Sensitivity Analysis Summary Report (with One-sided confounding function and default alpha)", "_____no_output_____" ] ], [ [ "sens_x_new = Sensitivity(df=df_new, inference_features=INFERENCE_FEATURES_new, p_col='pihat',\n treatment_col=TREATMENT_COL, outcome_col=OUTCOME_COL, learner=learner_x)\n# Here for Selection Bias method will use default one-sided confounding function and alpha (quantile range of outcome values) input\nsens_sumary_x_new = sens_x_new.sensitivity_analysis(methods=['Placebo Treatment',\n 'Random Cause',\n 'Subset Data',\n 'Random Replace',\n 'Selection Bias'], sample_size=0.5)", "_____no_output_____" ], [ "# Here we can see the New ATE restul from Random Replace method actually changed ~ 12.5%\nsens_sumary_x_new", "_____no_output_____" ] ], [ [ "### Random Replace", "_____no_output_____" ] ], [ [ "# Replace feature_0 with an irrelevent variable\nsens_x_replace_new = SensitivityRandomReplace(df=df_new, inference_features=INFERENCE_FEATURES_new, p_col='pihat',\n treatment_col=TREATMENT_COL, outcome_col=OUTCOME_COL, learner=learner_x,\n sample_size=0.9, replaced_feature='feature_1')\ns_check_replace_new = sens_x_replace_new.summary(method='Random Replace')\ns_check_replace_new", "_____no_output_____" ] ], [ [ "### Selection Bias: Alignment confounding Function", "_____no_output_____" ] ], [ [ "sens_x_bias_alignment_new = SensitivitySelectionBias(df_new, INFERENCE_FEATURES_new, p_col='pihat', treatment_col=TREATMENT_COL,\n outcome_col=OUTCOME_COL, learner=learner_x, confound='alignment',\n alpha_range=None)", "_____no_output_____" ], [ "lls_x_bias_alignment_new, partial_rsqs_x_bias_alignment_new = sens_x_bias_alignment_new.causalsens()", "_____no_output_____" ], [ "lls_x_bias_alignment_new", "_____no_output_____" ], [ "partial_rsqs_x_bias_alignment_new", "_____no_output_____" ], [ "# Plot the results by confounding vector and plot Confidence Intervals for ATE\nsens_x_bias_alignment_new.plot(lls_x_bias_alignment_new, ci=True)", "_____no_output_____" ], [ "# Plot the results by rsquare with partial r-square results by each individual features\nsens_x_bias_alignment_new.plot(lls_x_bias_alignment_new, partial_rsqs_x_bias_alignment_new, type='r.squared', partial_rsqs=True)", "_____no_output_____" ] ], [ [ "## Generate a Selection Bias Set", "_____no_output_____" ] ], [ [ "df_new_2 = df.copy()\ndf_new_2['treated_new'] = df['feature_0'].rank()\ndf_new_2['treated_new'] = [1 if i > df_new_2.shape[0]/2 else 0 for i in df_new_2['treated_new']]", "_____no_output_____" ], [ "df_new_2.head()", "_____no_output_____" ] ], [ [ "### Sensitivity Analysis Summary Report (with One-sided confounding function and default alpha)", "_____no_output_____" ] ], [ [ "sens_x_new_2 = Sensitivity(df=df_new_2, inference_features=INFERENCE_FEATURES, p_col='pihat',\n treatment_col='treated_new', outcome_col=OUTCOME_COL, learner=learner_x)\n# Here for Selection Bias method will use default one-sided confounding function and alpha (quantile range of outcome values) input\nsens_sumary_x_new_2 = sens_x_new_2.sensitivity_analysis(methods=['Placebo Treatment',\n 'Random Cause',\n 'Subset Data',\n 'Random Replace',\n 'Selection Bias'], sample_size=0.5)", "_____no_output_____" ], [ "sens_sumary_x_new_2", "_____no_output_____" ] ], [ [ "### Random Replace", "_____no_output_____" ] ], [ [ "# Replace feature_0 with an irrelevent variable\nsens_x_replace_new_2 = SensitivityRandomReplace(df=df_new_2, inference_features=INFERENCE_FEATURES, p_col='pihat',\n treatment_col='treated_new', outcome_col=OUTCOME_COL, learner=learner_x,\n sample_size=0.9, replaced_feature='feature_0')\ns_check_replace_new_2 = sens_x_replace_new_2.summary(method='Random Replace')\ns_check_replace_new_2", "_____no_output_____" ] ], [ [ "### Selection Bias: Alignment confounding Function", "_____no_output_____" ] ], [ [ "sens_x_bias_alignment_new_2 = SensitivitySelectionBias(df_new_2, INFERENCE_FEATURES, p_col='pihat', treatment_col='treated_new',\n outcome_col=OUTCOME_COL, learner=learner_x, confound='alignment',\n alpha_range=None)", "_____no_output_____" ], [ "lls_x_bias_alignment_new_2, partial_rsqs_x_bias_alignment_new_2 = sens_x_bias_alignment_new_2.causalsens()", "_____no_output_____" ], [ "lls_x_bias_alignment_new_2", "_____no_output_____" ], [ "partial_rsqs_x_bias_alignment_new_2", "_____no_output_____" ], [ "# Plot the results by confounding vector and plot Confidence Intervals for ATE\nsens_x_bias_alignment_new_2.plot(lls_x_bias_alignment_new_2, ci=True)", "_____no_output_____" ], [ "# Plot the results by rsquare with partial r-square results by each individual features\nsens_x_bias_alignment_new_2.plot(lls_x_bias_alignment_new, partial_rsqs_x_bias_alignment_new_2, type='r.squared', partial_rsqs=True)", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code" ] ]

ec7ceff9057b2d552f8002f2a81f523d1ff93003

43,805

ipynb

Jupyter Notebook

src_nlp/tensorflow/vae/train_cnn_tfd.ipynb

ashishpatel26/finch

bf2958c0f268575e5d51ad08fbc08b151cbea962

[ "MIT" ]

1

2019-02-12T09:22:00.000Z

src_nlp/tensorflow/vae/train_cnn_tfd.ipynb

loopzxl/finch

bf2958c0f268575e5d51ad08fbc08b151cbea962

[ "MIT" ]

null

src_nlp/tensorflow/vae/train_cnn_tfd.ipynb

loopzxl/finch

bf2958c0f268575e5d51ad08fbc08b151cbea962

[ "MIT" ]

1

2020-10-15T21:34:17.000Z

55.239596

673

0.63075

[ [ [ "Notebook written by [Zhedong Zheng](https://github.com/zhedongzheng)\n\n<img src=\"cnn_vae.png\" width=\"300\">", "_____no_output_____" ] ], [ [ "import tensorflow as tf\nimport numpy as np", "_____no_output_____" ], [ "PARAMS = {\n 'max_len': 15,\n 'vocab_size': 10000,\n 'embed_dims': 128,\n 'rnn_size': 128,\n 'cnn_size': 128,\n 'latent_size': 16,\n 'kernel_sz': 3,\n 'n_hidden_layer': 3,\n 'clip_norm': 5.0,\n 'anneal_max': 1.0,\n 'anneal_bias': 6000,\n 'batch_size': 128,\n 'n_epochs': 10,\n}", "_____no_output_____" ], [ "def build_vocab(index_from=4):\n PARAMS['word2idx'] = tf.keras.datasets.imdb.get_word_index()\n PARAMS['word2idx'] = {k: (v + index_from) for k, v in PARAMS['word2idx'].items()}\n PARAMS['word2idx']['<pad>'] = 0\n PARAMS['word2idx']['<start>'] = 1\n PARAMS['word2idx']['<unk>'] = 2\n PARAMS['word2idx']['<end>'] = 3\n PARAMS['idx2word'] = {i: w for w, i in PARAMS['word2idx'].items()}\n\n \ndef load_data(index_from=4):\n (X_train, _), (X_test, _) = tf.contrib.keras.datasets.imdb.load_data(\n num_words=PARAMS['vocab_size'], index_from=index_from)\n return (X_train, X_test)", "_____no_output_____" ], [ "word2idx = build_vocab()\nX = np.concatenate(load_data())\n\nX = np.concatenate((\n tf.keras.preprocessing.sequence.pad_sequences(\n X, PARAMS['max_len'], truncating='post', padding='post'),\n tf.keras.preprocessing.sequence.pad_sequences(\n X, PARAMS['max_len'], truncating='pre', padding='post')))\n\nenc_inp = X[:, 1:]\ndec_inp = X\ndec_out = np.concatenate([X[:, 1:], np.full([X.shape[0], 1], PARAMS['word2idx']['<end>'])], 1)", "_____no_output_____" ], [ "def kl_w_fn(global_step):\n return PARAMS['anneal_max'] * tf.sigmoid((10 / PARAMS['anneal_bias']) * (\n tf.to_float(global_step) - tf.constant(PARAMS['anneal_bias'] / 2)))\n\n\ndef clip_grads(loss):\n variables = tf.trainable_variables()\n grads = tf.gradients(loss, variables)\n clipped_grads, _ = tf.clip_by_global_norm(grads, PARAMS['clip_norm'])\n return zip(clipped_grads, variables)\n\n\ndef rnn_cell():\n return tf.nn.rnn_cell.GRUCell(PARAMS['rnn_size'],\n kernel_initializer=tf.orthogonal_initializer())\n\n\ndef cnn_block(x, dilation_rate, pad_sz):\n pad = tf.zeros([tf.shape(x)[0], pad_sz, x.get_shape()[-1].value])\n x = tf.layers.conv1d(inputs = tf.concat([pad, x, pad], 1),\n filters = PARAMS['cnn_size'],\n kernel_size = PARAMS['kernel_sz'],\n dilation_rate = dilation_rate)\n x = x[:, :-pad_sz, :]\n x = tf.nn.relu(x)\n return x\n\n\ndef cnn_forward(x, embedding):\n for i in range(PARAMS['n_hidden_layer']):\n dilation_rate = 2 ** i\n pad_sz = (PARAMS['kernel_sz'] - 1) * dilation_rate\n x += cnn_block(x, dilation_rate, pad_sz)\n logits = tf.reshape(x, [-1, PARAMS['cnn_size']])\n logits = tf.matmul(logits, embedding, transpose_b=True)\n logits = tf.reshape(logits, [tf.shape(x)[0], -1, PARAMS['vocab_size']])\n return logits\n\n\ndef autoregressive(embedding, z, input_proj):\n batch_sz = tf.shape(z)[0]\n \n def cond(i, x, temp):\n return i < PARAMS['max_len']\n\n def body(i, x, temp):\n sos = tf.fill([batch_sz, 1], PARAMS['word2idx']['<start>'])\n x = tf.concat([sos, x[:, :-1]], 1)\n \n x = tf.nn.embedding_lookup(embedding, x)\n x = input_proj(tf.concat([x, z], -1))\n logits = cnn_forward(x, embedding)\n ids = tf.argmax(logits, -1, output_type=tf.int32)[:, i]\n ids = tf.expand_dims(ids, -1)\n\n temp = tf.concat([temp[:, 1:], ids], -1)\n\n x = tf.concat([temp[:, -(i+1):], temp[:, :-(i+1)]], -1)\n x = tf.reshape(x, [batch_sz, PARAMS['max_len']])\n i += 1\n return i, x, temp\n \n x = tf.zeros([batch_sz, PARAMS['max_len']], tf.int32)\n _, res, _ = tf.while_loop(cond, body, [tf.constant(0), x, x])\n \n return res", "_____no_output_____" ], [ "def forward(inputs, labels, mode):\n is_training = (mode == tf.estimator.ModeKeys.TRAIN)\n enc_seq_len = tf.count_nonzero(inputs, 1, dtype=tf.int32)\n batch_sz = tf.shape(inputs)[0]\n \n with tf.variable_scope('Encoder'):\n embedding = tf.get_variable('lookup_table', [PARAMS['vocab_size'],\n PARAMS['embed_dims']])\n x = tf.nn.embedding_lookup(embedding, inputs)\n \n _, enc_state = tf.nn.dynamic_rnn(rnn_cell(), x, enc_seq_len, dtype=tf.float32)\n \n z_mean = tf.layers.dense(enc_state, PARAMS['latent_size'])\n z_var = tf.layers.dense(enc_state, PARAMS['latent_size'])\n \n posterior = tf.contrib.distributions.MultivariateNormalDiag(z_mean, z_var)\n prior = tf.contrib.distributions.MultivariateNormalDiag(tf.zeros_like(z_mean),\n tf.ones_like(z_var))\n \n with tf.variable_scope('Decoder'):\n input_proj = tf.layers.Dense(PARAMS['cnn_size'], tf.nn.relu)\n z = tf.tile(tf.expand_dims(posterior.sample(), 1), [1, PARAMS['max_len'], 1])\n \n if is_training:\n dec_inp = tf.nn.embedding_lookup(embedding, labels['dec_inp'])\n x = input_proj(tf.concat([dec_inp, z], -1))\n logits = cnn_forward(x, embedding)\n return logits, posterior, prior\n else:\n return autoregressive(embedding, z, input_proj)", "_____no_output_____" ], [ "def model_fn(features, labels, mode):\n logits_or_ids = forward(features, labels, mode) \n \n if mode == tf.estimator.ModeKeys.PREDICT:\n return tf.estimator.EstimatorSpec(mode, predictions=logits_or_ids)\n \n if mode == tf.estimator.ModeKeys.TRAIN:\n logits, posterior, prior = logits_or_ids\n \n out_dist = tf.distributions.Categorical(logits)\n \n global_step = tf.train.get_global_step()\n \n nll_loss = - tf.reduce_sum(out_dist.log_prob(labels['dec_out']))\n \n kl_w = kl_w_fn(global_step)\n \n kl_loss = tf.reduce_sum(tf.distributions.kl_divergence(posterior, prior))\n \n loss_op = nll_loss + kl_w * kl_loss\n \n train_op = tf.train.AdamOptimizer().apply_gradients(\n clip_grads(loss_op),\n global_step = global_step)\n \n lth = tf.train.LoggingTensorHook(\n {'nll_loss': nll_loss, 'kl_w': kl_w, 'kl_loss': kl_loss}, every_n_iter=100)\n \n return tf.estimator.EstimatorSpec(\n mode=mode, loss=loss_op, train_op=train_op, training_hooks=[lth])", "_____no_output_____" ], [ "def inf_inp(test_strs):\n x = [[PARAMS['word2idx'].get(w, 2) for w in s.split()] for s in test_strs]\n x = tf.keras.preprocessing.sequence.pad_sequences(\n x, PARAMS['max_len'], truncating='post', padding='post')\n return x\n\ndef demo(test_strs, pred_ids):\n for s, pred in zip(test_strs, pred_ids):\n print('\\nOriginal:', s)\n print('Reconstr:', ' '.join([PARAMS['idx2word'].get(idx, '<unk>') for idx in pred]))\n\n\ntest_strs = ['i love this film and i think it is one of the best films',\n 'this movie is a waste of time and there is no point to watch it']\n\nestimator = tf.estimator.Estimator(model_fn)\n\nfor _ in range(PARAMS['n_epochs']):\n estimator.train(tf.estimator.inputs.numpy_input_fn(\n x = enc_inp,\n y = {'dec_inp': dec_inp, 'dec_out': dec_out},\n batch_size = PARAMS['batch_size'],\n shuffle = True))\n \n pred_ids = list(estimator.predict(tf.estimator.inputs.numpy_input_fn(\n x = inf_inp(test_strs),\n shuffle = False)))\n\n demo(test_strs, pred_ids)\n \n print()", "INFO:tensorflow:Using default config.\nWARNING:tensorflow:Using temporary folder as model directory: /var/folders/sx/fv0r97j96fz8njp14dt5g7940000gn/T/tmpt6rxeanl\nINFO:tensorflow:Using config: {'_model_dir': '/var/folders/sx/fv0r97j96fz8njp14dt5g7940000gn/T/tmpt6rxeanl', '_tf_random_seed': None, '_save_summary_steps': 100, '_save_checkpoints_steps': None, '_save_checkpoints_secs': 600, '_session_config': None, '_keep_checkpoint_max': 5, '_keep_checkpoint_every_n_hours': 10000, '_log_step_count_steps': 100, '_train_distribute': None, '_service': None, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x11b432eb8>, '_task_type': 'worker', '_task_id': 0, '_global_id_in_cluster': 0, '_master': '', '_evaluation_master': '', '_is_chief': True, '_num_ps_replicas': 0, '_num_worker_replicas': 1}\nINFO:tensorflow:Calling model_fn.\nINFO:tensorflow:Done calling model_fn.\nINFO:tensorflow:Create CheckpointSaverHook.\nINFO:tensorflow:Graph was finalized.\nINFO:tensorflow:Running local_init_op.\nINFO:tensorflow:Done running local_init_op.\nINFO:tensorflow:Saving checkpoints for 1 into /var/folders/sx/fv0r97j96fz8njp14dt5g7940000gn/T/tmpt6rxeanl/model.ckpt.\nINFO:tensorflow:loss = 17757.797, step = 1\nINFO:tensorflow:nll_loss = 17684.46, kl_w = 0.006692851, kl_loss = 10957.21\nINFO:tensorflow:global_step/sec: 2.17765\nINFO:tensorflow:loss = 10933.477, step = 101 (45.922 sec)\nINFO:tensorflow:nll_loss = 10900.135, kl_w = 0.007897083, kl_loss = 4222.0977 (45.922 sec)\nINFO:tensorflow:global_step/sec: 2.26024\nINFO:tensorflow:loss = 10842.261, step = 201 (44.243 sec)\nINFO:tensorflow:nll_loss = 10804.835, kl_w = 0.009315956, kl_loss = 4017.3525 (44.243 sec)\nINFO:tensorflow:global_step/sec: 2.20411\nINFO:tensorflow:loss = 10243.188, step = 301 (45.370 sec)\nINFO:tensorflow:nll_loss = 10192.387, kl_w = 0.010986943, kl_loss = 4623.8013 (45.370 sec)\nINFO:tensorflow:global_step/sec: 2.18017\nINFO:tensorflow:loss = 9709.822, step = 401 (45.868 sec)\nINFO:tensorflow:nll_loss = 9650.805, kl_w = 0.012953726, kl_loss = 4556.036 (45.868 sec)\nINFO:tensorflow:global_step/sec: 2.23163\nINFO:tensorflow:loss = 9265.691, step = 501 (44.810 sec)\nINFO:tensorflow:nll_loss = 9190.885, kl_w = 0.015267149, kl_loss = 4899.8467 (44.810 sec)\nINFO:tensorflow:global_step/sec: 2.18961\nINFO:tensorflow:loss = 8816.474, step = 601 (45.670 sec)\nINFO:tensorflow:nll_loss = 8726.945, kl_w = 0.01798621, kl_loss = 4977.6035 (45.670 sec)\nINFO:tensorflow:global_step/sec: 2.33169\nINFO:tensorflow:loss = 8734.211, step = 701 (42.887 sec)\nINFO:tensorflow:nll_loss = 8631.089, kl_w = 0.021179108, kl_loss = 4869.068 (42.888 sec)\nINFO:tensorflow:Saving checkpoints for 782 into /var/folders/sx/fv0r97j96fz8njp14dt5g7940000gn/T/tmpt6rxeanl/model.ckpt.\nINFO:tensorflow:Loss for final step: 2049.7031.\nINFO:tensorflow:Calling model_fn.\nINFO:tensorflow:Done calling model_fn.\nINFO:tensorflow:Graph was finalized.\nINFO:tensorflow:Restoring parameters from /var/folders/sx/fv0r97j96fz8njp14dt5g7940000gn/T/tmpt6rxeanl/model.ckpt-782\nINFO:tensorflow:Running local_init_op.\nINFO:tensorflow:Done running local_init_op.\n\nOriginal: i love this film and i think it is one of the best films\nReconstr: i saw this movie for this movie i think that the <unk> <unk> i <end>\n\nOriginal: this movie is a waste of time and there is no point to watch it\nReconstr: this movie is a fan of time i just seen it movie about <unk> <end>\n\nINFO:tensorflow:Calling model_fn.\nINFO:tensorflow:Done calling model_fn.\nINFO:tensorflow:Create CheckpointSaverHook.\nINFO:tensorflow:Graph was finalized.\nINFO:tensorflow:Restoring parameters from /var/folders/sx/fv0r97j96fz8njp14dt5g7940000gn/T/tmpt6rxeanl/model.ckpt-782\nINFO:tensorflow:Running local_init_op.\nINFO:tensorflow:Done running local_init_op.\nINFO:tensorflow:Saving checkpoints for 783 into /var/folders/sx/fv0r97j96fz8njp14dt5g7940000gn/T/tmpt6rxeanl/model.ckpt.\nINFO:tensorflow:loss = 9291.661, step = 783\nINFO:tensorflow:nll_loss = 9169.796, kl_w = 0.024205629, kl_loss = 5034.592\nINFO:tensorflow:global_step/sec: 2.23615\nINFO:tensorflow:loss = 8433.8125, step = 883 (44.722 sec)\nINFO:tensorflow:nll_loss = 8289.746, kl_w = 0.028470589, kl_loss = 5060.168 (44.721 sec)\nINFO:tensorflow:global_step/sec: 2.24842\nINFO:tensorflow:loss = 8156.293, step = 983 (44.475 sec)\nINFO:tensorflow:nll_loss = 7992.166, kl_w = 0.033461247, kl_loss = 4904.9897 (44.475 sec)\nINFO:tensorflow:global_step/sec: 2.19425\nINFO:tensorflow:loss = 8059.392, step = 1083 (45.574 sec)\nINFO:tensorflow:nll_loss = 7862.749, kl_w = 0.039291352, kl_loss = 5004.737 (45.574 sec)\nINFO:tensorflow:global_step/sec: 2.14054\nINFO:tensorflow:loss = 7877.4155, step = 1183 (46.717 sec)\nINFO:tensorflow:nll_loss = 7650.47, kl_w = 0.04608883, kl_loss = 4924.08 (46.717 sec)\nINFO:tensorflow:global_step/sec: 2.13802\nINFO:tensorflow:loss = 8841.221, step = 1283 (46.772 sec)\nINFO:tensorflow:nll_loss = 8583.125, kl_w = 0.053996176, kl_loss = 4779.885 (46.772 sec)\nINFO:tensorflow:global_step/sec: 2.04279\nINFO:tensorflow:loss = 8351.722, step = 1383 (48.953 sec)\nINFO:tensorflow:nll_loss = 8054.1445, kl_w = 0.06317033, kl_loss = 4710.71 (48.953 sec)\nINFO:tensorflow:global_step/sec: 2.18052\nINFO:tensorflow:loss = 8543.8, step = 1483 (45.860 sec)\nINFO:tensorflow:nll_loss = 8207.943, kl_w = 0.07378165, kl_loss = 4552.0293 (45.860 sec)\nINFO:tensorflow:Saving checkpoints for 1564 into /var/folders/sx/fv0r97j96fz8njp14dt5g7940000gn/T/tmpt6rxeanl/model.ckpt.\nINFO:tensorflow:Loss for final step: 2099.5452.\nINFO:tensorflow:Calling model_fn.\nINFO:tensorflow:Done calling model_fn.\nINFO:tensorflow:Graph was finalized.\nINFO:tensorflow:Restoring parameters from /var/folders/sx/fv0r97j96fz8njp14dt5g7940000gn/T/tmpt6rxeanl/model.ckpt-1564\nINFO:tensorflow:Running local_init_op.\nINFO:tensorflow:Done running local_init_op.\n\nOriginal: i love this film and i think it is one of the best films\nReconstr: i was so bad and this movie is that all and all of time <end>\n\nOriginal: this movie is a waste of time and there is no point to watch it\nReconstr: this movie is a lot of movies that don't waste of this movie <unk> <end>\n\nINFO:tensorflow:Calling model_fn.\nINFO:tensorflow:Done calling model_fn.\nINFO:tensorflow:Create CheckpointSaverHook.\nINFO:tensorflow:Graph was finalized.\nINFO:tensorflow:Restoring parameters from /var/folders/sx/fv0r97j96fz8njp14dt5g7940000gn/T/tmpt6rxeanl/model.ckpt-1564\nINFO:tensorflow:Running local_init_op.\nINFO:tensorflow:Done running local_init_op.\nINFO:tensorflow:Saving checkpoints for 1565 into /var/folders/sx/fv0r97j96fz8njp14dt5g7940000gn/T/tmpt6rxeanl/model.ckpt.\nINFO:tensorflow:loss = 8396.324, step = 1565\nINFO:tensorflow:nll_loss = 8022.4976, kl_w = 0.08368247, kl_loss = 4467.2085\nINFO:tensorflow:global_step/sec: 2.21291\nINFO:tensorflow:loss = 8478.651, step = 1665 (45.190 sec)\nINFO:tensorflow:nll_loss = 8043.9316, kl_w = 0.09738124, kl_loss = 4464.1055 (45.190 sec)\nINFO:tensorflow:global_step/sec: 2.07123\nINFO:tensorflow:loss = 8133.461, step = 1765 (48.281 sec)\nINFO:tensorflow:nll_loss = 7641.1675, kl_w = 0.11304584, kl_loss = 4354.8145 (48.281 sec)\nINFO:tensorflow:global_step/sec: 2.21284\nINFO:tensorflow:loss = 7868.1704, step = 1865 (45.191 sec)\nINFO:tensorflow:nll_loss = 7321.884, kl_w = 0.13086486, kl_loss = 4174.4326 (45.191 sec)\nINFO:tensorflow:global_step/sec: 2.02446\nINFO:tensorflow:loss = 8176.475, step = 1965 (49.396 sec)\nINFO:tensorflow:nll_loss = 7531.4424, kl_w = 0.15101443, kl_loss = 4271.332 (49.395 sec)\nINFO:tensorflow:global_step/sec: 2.13666\nINFO:tensorflow:loss = 7860.392, step = 2065 (46.802 sec)\nINFO:tensorflow:nll_loss = 7156.92, kl_w = 0.17364664, kl_loss = 4051.1694 (46.802 sec)\nINFO:tensorflow:global_step/sec: 2.22284\nINFO:tensorflow:loss = 8611.154, step = 2165 (44.987 sec)\nINFO:tensorflow:nll_loss = 7857.939, kl_w = 0.19887616, kl_loss = 3787.358 (44.987 sec)\nINFO:tensorflow:global_step/sec: 2.18977\nINFO:tensorflow:loss = 8461.563, step = 2265 (45.667 sec)\nINFO:tensorflow:nll_loss = 7666.117, kl_w = 0.22676536, kl_loss = 3507.7935 (45.667 sec)\nINFO:tensorflow:Saving checkpoints for 2346 into /var/folders/sx/fv0r97j96fz8njp14dt5g7940000gn/T/tmpt6rxeanl/model.ckpt.\nINFO:tensorflow:Loss for final step: 2221.3452.\nINFO:tensorflow:Calling model_fn.\n" ] ] ]

[ "markdown", "code" ]

[ [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7d0ebac1a9968be1aa7c6f304147fb87a3d6cb

370,897

ipynb

Jupyter Notebook

All iPython Notebooks/10_W_4_Web_scraping.ipynb

2series/Analytics-And-Python

38cc6b17e0c946da6360a63025979d9bffddedfa

[ "MIT" ]

null

All iPython Notebooks/10_W_4_Web_scraping.ipynb

2series/Analytics-And-Python

38cc6b17e0c946da6360a63025979d9bffddedfa

[ "MIT" ]

null

All iPython Notebooks/10_W_4_Web_scraping.ipynb

2series/Analytics-And-Python

38cc6b17e0c946da6360a63025979d9bffddedfa

[ "MIT" ]

null

146.195112

97,222

0.701135

[ [ [ "<h2>Caveat</h2>\nWeb sites often change the format of their pages so this may not always work. If it doesn't, rework the examples after examining the html content of the page (most browsers will let you see the html source - look for a \"page source\" option - though you might have to turn on the developer mode in your browser preferences. For example, on Chrome you need to click the \"developer mode\" check box under Extensions in the Preferences/Options menu. \n\nFeel free to use the links below to navigate the notebook:\n- [**STEP 0**](#import): Import necessary modules (requests and BeautifulSoup)\n- [**STEP 1**](#step1): Check the http Request-Response cycle (**response.status_code == 200**)\n - [Model 0](#model0): Set up the BeautifulSoup object (**results_page**)\n- [**STEP 2**](#step2): BS4 functions\n - [Model 1](#model1): **find_all()**\n - [Model 2](#model2): **find()**\n - [Model 3](#model3): find() and find_all() qualification by CSS selectors\n - [Model 4](#model4): __get_text()__ returns the marked up text (the content) enclosed in a tag\n - [Model 5](#model5): __get()__ returns the __value__ of a tag attribute\n- [**STEP 3**](#step3): Dealing with tags (Example)\n - [Model 6](#model6): Build a new function (Example 1)\n - [Model 7](#model7): Construct a list of dictionaries\n- [**STEP 4**](#step4): Logging in to a web server\n - [Model 8](#model8): Get username and password\n - [Model 9](#model9): Construct an object that contains the data to be sent to the login page\n - [Model 10](#model10): Get the value of the login token\n - [Model 11](#model11): Setup a session, login, and get data", "_____no_output_____" ], [ "<a id='import'></a>\n\n<h3>Import necessary modules</h3>", "_____no_output_____" ] ], [ [ "import requests\nfrom bs4 import BeautifulSoup", "_____no_output_____" ] ], [ [ "<a id='step1'></a>\n<h3>The http Request-Response cycle</h3>", "_____no_output_____" ] ], [ [ "url = \"http://www.epicurious.com/search/Tofu Chili\"\nresponse = requests.get(url)\nif response.status_code == 200:\n print(\"Success\")\nelse:\n print(\"Failure\")", "Success\n" ], [ "keywords = input(\"Please enter the things you want to see in a recipe\")\nurl = \"http://www.epicurious.com/search/\" + keywords\nresponse = requests.get(url)\nif response.status_code == 200:\n print(\"Success\")\nelse:\n print(\"Failure\")", "Please enter the things you want to see in a recipe tofu chili\nSuccess\n" ] ], [ [ "<a id='model0'></a>\n<h3>Set up the BeautifulSoup object</h3>", "_____no_output_____" ] ], [ [ "results_page = BeautifulSoup(response.content,'lxml')\nprint(results_page.prettify())", "<!DOCTYPE html>\n<html>\n <head>\n <meta charset=\"utf-8\"/>\n <meta content=\"app-id=312101965\" name=\"apple-itunes-app\"/>\n <title>\n Search | Epicurious.com\n </title>\n <link href=\"//assets.adobedtm.com\" rel=\"dns-prefetch\"/>\n <link href=\"https://www.google-analytics.com\" rel=\"dns-prefetch\"/>\n <link href=\"//tpc.googlesyndication.com\" rel=\"dns-prefetch\"/>\n <link href=\"//static.parsely.com\" rel=\"dns-prefetch\"/>\n <link href=\"//condenast.demdex.net\" rel=\"dns-prefetch\"/>\n <link href=\"//capture.condenastdigital.com\" rel=\"dns-prefetch\"/>\n <link href=\"//pixel.condenastdigital.com\" rel=\"dns-prefetch\"/>\n <link href=\"//use.typekit.net\" rel=\"dns-prefetch\"/>\n <link href=\"//fonts.typekit.net\" rel=\"dns-prefetch\"/>\n <link href=\"//p.typekit.net\" rel=\"dns-prefetch\"/>\n <link href=\"//assets.epicurious.com\" rel=\"dns-prefetch\"/>\n <link href=\"//ad.doubleclick.net\" rel=\"dns-prefetch\"/>\n <link href=\"//pagead2.googlesyndication.com\" rel=\"dns-prefetch\"/>\n <link href=\"//z.moatads.com\" rel=\"dns-prefetch\"/>\n <meta content=\"en_US\" itemprop=\"inLanguage\" property=\"og:locale\"/>\n <meta content=\"IE=edge\" http-equiv=\"x-ua-compatible\"/>\n <meta content=\"no-cache\" http-equiv=\"cache-control\"/>\n <meta content=\"no-cache\" http-equiv=\"pragma\"/>\n <meta content=\"Search | Epicurious.com\" itemprop=\"name\"/>\n <meta content=\"https://www.epicurious.com/static/img/misc/epicurious-social-logo.png\" itemprop=\"logo\"/>\n <meta content=\"Easily search and browse more than 37,000 recipes, articles, galleries, menus, and videos from Epicurious.com, Bon Appétit, and other partners.\" name=\"description\"/>\n <meta content=\"Epicurious\" itemprop=\"author\"/>\n <link href=\"https://www.epicurious.com/search/%20tofu%20chili\" rel=\"canonical\"/>\n <meta content=\"Copyright (c) 2018 Conde Nast\" name=\"copyright\"/>\n <meta content=\"9c2002da922784afad64b638161c75f7\" name=\"p:domain_verify\"/>\n <meta content=\"Search | Epicurious.com\" property=\"og:title\"/>\n <meta content=\"website\" property=\"og:type\"/>\n <meta content=\"https://www.epicurious.com/search/%20tofu%20chili\" property=\"og:url\"/>\n <meta content=\"Easily search and browse more than 37,000 recipes, articles, galleries, menus, and videos from Epicurious.com, Bon Appétit, and other partners.\" property=\"og:description\"/>\n <meta content=\"https://www.epicurious.com/static/img/misc/epicurious-social-logo.png\" property=\"og:image\"/>\n <meta content=\"Epicurious\" property=\"og:site_name\"/>\n <meta content=\"1636080783276430\" property=\"fb:app_id\"/>\n <meta content=\"722582662\" property=\"fb:admins\"/>\n <meta content=\"774348857\" property=\"fb:admins\"/>\n <meta content=\"596666898\" property=\"fb:admins\"/>\n <meta content=\"837402\" property=\"fb:admins\"/>\n <meta content=\"685417657\" property=\"fb:admins\"/>\n <meta content=\"22500087\" property=\"fb:admins\"/>\n <meta content=\"1107036618\" property=\"fb:admins\"/>\n <meta content=\"1045857449\" property=\"fb:admins\"/>\n <meta content=\"14601235\" property=\"fb:admins\"/>\n <link href=\"https://plus.google.com/106968200752753566855\" rel=\"publisher\"/>\n <link href=\"/static/img/favicon.png\" rel=\"icon\" type=\"image/png\"/>\n <meta content=\"#f93f23\" name=\"theme-color\"/>\n <meta content=\"width=device-width, initial-scale=1.0\" name=\"viewport\"/>\n \n <script>\n var EPI = EPI || {\n barCnsCrtPage: false,\n onCompleteActions: []\n };\n\n document.onreadystatechange = function () {\n if (document.readyState === 'complete') {\n EPI.onCompleteActions.forEach(function (callback) {\n if (typeof callback === 'function') {\n callback();\n }\n });\n\n if (window.location.search.substr(1).indexOf('ui-regression-test=true') >= 0) {\n console.log('ui-regression-test-ready');\n }\n }\n };\n\n \n !function(a){var b=function(b,c){\"use strict\";var d=a.document.getElementsByTagName(\"script\")[0],e=a.document.createElement(\"script\");return e.src=b,e.async=!0,d.parentNode.insertBefore(e,d),c&&\"function\"==typeof c&&(e.onload=c),e};\"undefined\"!=typeof module?module.exports=b:a.loadJS=b}(\"undefined\"!=typeof global?global:this); // loadCSS\n (function(h){var d=function(d,e,n){function k(a){if(b.body)return a();setTimeout(function(){k(a)})}function f(){a.addEventListener&&a.removeEventListener(\"load\",f);a.media=n||\"all\"}var b=h.document,a=b.createElement(\"link\"),c;if(e)c=e;else{var l=(b.body||b.getElementsByTagName(\"head\")[0]).childNodes;c=l[l.length-1]}var m=b.styleSheets;a.rel=\"stylesheet\";a.href=d;a.media=\"only x\";k(function(){c.parentNode.insertBefore(a,e?c:c.nextSibling)});var g=function(b){for(var c=a.href,d=m.length;d--;)if(m[d].href===\n c)return b();setTimeout(function(){g(b)})};a.addEventListener&&a.addEventListener(\"load\",f);a.onloadcssdefined=g;g(f);return a};\"undefined\"!==typeof exports?exports.loadCSS=d:h.loadCSS=d})(\"undefined\"!==typeof global?global:this);\n \n (function(a){if(a.loadCSS){var b=loadCSS.relpreload={};b.support=function(){try{return a.document.createElement(\"link\").relList.supports(\"preload\")}catch(b){return!1}};b.poly=function(){for(var b=a.document.getElementsByTagName(\"link\"),d=0;d<b.length;d++){var c=b[d];\"preload\"===c.rel&&\"style\"===c.getAttribute(\"as\")&&(a.loadCSS(c.href,c),c.rel=null)}};if(!b.support()){b.poly();var e=a.setInterval(b.poly,300);a.addEventListener&&a.addEventListener(\"load\",function(){a.clearInterval(e)});a.attachEvent&&\n a.attachEvent(\"onload\",function(){a.clearInterval(e)})}}})(this);\n </script>\n <link as=\"style\" href=\"https://use.typekit.net/zpl6zji.css\" onload=\"this.rel='stylesheet'\" rel=\"preload\" type=\"text/css\"/>\n <style>\n @font-face{font-family:'Gotham SSm 5r';src:url(\"/static/fonts/GothamSSm5r.woff2\") format(\"woff2\"),url(\"/static/fonts/GothamSSm5r.woff\") format(\"woff\");font-weight:normal;font-style:normal}@font-face{font-family:'Gotham SSm 5i';src:url(\"/static/fonts/GothamSSm5i.woff2\") format(\"woff2\"),url(\"/static/fonts/GothamSSm5i.woff\") format(\"woff\");font-weight:normal;font-style:normal}@font-face{font-family:'Gotham SSm 7r';src:url(\"/static/fonts/GothamSSm7r.woff2\") format(\"woff2\"),url(\"/static/fonts/GothamSSm7r.woff\") format(\"woff\");font-weight:normal;font-style:normal}@font-face{font-family:'Gotham SSm 7i';src:url(\"/static/fonts/GothamSSm7i.woff2\") format(\"woff2\"),url(\"/static/fonts/GothamSSm7i.woff\") format(\"woff\");font-weight:normal;font-style:normal}@font-face{font-family:'Gotham Cond SSm 5r';src:url(\"/static/fonts/GothamCondSSm5r.woff2\") format(\"woff2\"),url(\"/static/fonts/GothamCondSSm5r.woff\") format(\"woff\");font-weight:normal;font-style:normal}@font-face{font-family:'Gotham Cond SSm 6r';src:url(\"/static/fonts/GothamCondSSm6r.woff2\") format(\"woff2\"),url(\"/static/fonts/GothamCondSSm6r.woff\") format(\"woff\");font-weight:normal;font-style:normal}header[role=\"banner\"],.search-dialog-form input[type=\"text\"]{font-family:\"Gotham SSm 7r\", sans-serif;font-weight:normal}[class$=\"-flash-message\"]{font-family:\"utopia-std-display\", serif;font-style:italic;font-weight:400}.icon-share{background-image:url(\"data:image/svg+xml;charset=US-ASCII,%3Csvg%20xmlns%3D%22http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg%22%20viewBox%3D%22-64%20741%20128%20102%22%3E%3Cpath%20d%3D%22M37%20813.7v19.5h-88.7v-63.4h19.6l13.6-11.8H-64v87H49.4v-43.2zm-11.7-19.8s-54.7-14.2-69.4%2034.6c0%200-5.3-69.9%2069.4-69.9V741L64%20775.8l-38.7%2036v-17.9z%22%2F%3E%3C%2Fsvg%3E\");background-repeat:no-repeat}footer[role=\"contentinfo\"] .corporate-info .conde-nast-brands .nav-title:after{background-repeat:no-repeat;background-size:contain;height:1em;width:.7em}footer[role=\"contentinfo\"] .corporate-info .conde-nast-brands .nav-title:after{content:\"\";display:inline-block;position:relative;top:1px;vertical-align:baseline}footer[role=\"contentinfo\"] .corporate-info .conde-nast-brands .nav-title:after{background-position:100% 0%;margin-left:.5rem}footer[role=\"contentinfo\"] .corporate-info .conde-nast-brands .nav-title:after{background-image:url(\"data:image/svg+xml;base64,PHN2ZyB2aWV3Qm94PSIwIDAgNzAgMTAwIiB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciPjxwYXRoIGQ9Ik03MCA1MC44NTlMMCAwbDIxLjU4IDUwLjg1OUwwIDEwMHoiIGZpbGw9IiNDQ0MiIGZpbGwtcnVsZT0iZXZlbm9kZCIvPjwvc3ZnPg==\")}footer[role=\"contentinfo\"] .corporate-info>.section-title,.homepage-content-channel-link,.lightreg-dialog fieldset legend,header[role=\"banner\"] .epicurious-logo,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .nav-title,.main-navigation .facebook-social-channel-link,.branding .facebook-social-channel-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .facebook-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .feed-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .googleplus-link,header[role=\"banner\"] .show-main-navigation:hover,header[role=\"banner\"] .show-main-navigation,header[role=\"banner\"] .show-search-button:hover,.search-dialog-form [type=\"submit\"]:hover,header[role=\"banner\"] .show-search-button,.search-dialog-form [type=\"submit\"],.user-status[data-user-type=\"anonymous\"] .recipebox-status:hover,.user-status[data-user-type=\"anonymous\"] .recipebox-status,.user-status[data-user-type=\"authenticated\"] .recipebox-status,.main-navigation .instagram-social-channel-link,.branding .instagram-social-channel-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .instagram-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .pinterest-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .tumblr-link,.main-navigation .twitter-social-channel-link,.branding .twitter-social-channel-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .twitter-link,.search-dialog-form [type=\"reset\"]:hover,[class$=\"-form\"] .close-button,.search-dialog-form [type=\"reset\"],.dismiss-main-navigation:hover,[class$=\"-comment-form\"] .close-button,.dismiss-main-navigation,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .youtube-link{background-color:transparent;background-position:50% 50%;background-repeat:no-repeat;background-size:contain}footer[role=\"contentinfo\"] .corporate-info>.section-title{background-image:url(\"data:image/svg+xml;base64,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\")}.homepage-content-channel-link{background-image:url(\"data:image/svg+xml;base64,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\")}.lightreg-dialog fieldset legend,header[role=\"banner\"] .epicurious-logo{background-image:url(\"data:image/svg+xml;base64,PHN2ZyB2aWV3Qm94PSIwIDAgNjEyIDEyOCIgeG1sbnM9Imh0dHA6Ly93d3cudzMub3JnLzIwMDAvc3ZnIj48cGF0aCBkPSJNMTAwLjYgODguOWMtMi44IDAtNC41LTEuMS01LjUtMi44VjQwLjhjMS43LTEuMyAzLjgtMi4yIDYuMy0yLjIgOC43IDAgMTEuMSAxMS4yIDExLjEgMjQuMS0uMSAxOS4xLTUuNCAyNi4yLTExLjkgMjYuMnptMTAuNS02MC43Yy02LjYgMC0xMi4xIDMtMTYuMSA3LjZsLS41LTYuM0g3MS4ybC0uMSA5Ny45aDI0Vjk0LjdjMy4zIDIuOCA4IDQuNCAxMyA0LjQgMTYgMCAyOC0xNSAyOC0zNS41IDAtMTktOC4xLTM1LjQtMjUtMzUuNHptMjY2LjIgMS41aDIzLjl2NjcuOGgtMjMuOVYyOS43ek0zNzEgNDIuM2MwIDcuOS00LjUgMTMuMy0xMS44IDEzLjMtNy4yIDAtMTEtNS4yLTExLTEwLjIgMC0xLjcuNi01IDIuMy03LjUtMy0xLjYtMTEuNCAxLjgtMTQuNCAxMS42djQ4aC0yMy45VjI5LjdoMjMuNUwzMzYgNDFjNC4zLTguMSAxMC44LTEyLjggMTguOS0xMi44IDEwLjUgMCAxNi4xIDYuNiAxNi4xIDE0LjF6bS05Mi4yLTEyLjZoMjMuOXY2Ny44aC0yMy45di0xNWMtMi40IDYuNi04LjIgMTYuNi0yMC41IDE2LjYtMTIuOCAwLTE5LTEwLjEtMTktMjAuOVYzOS41YzAtNS4xLTEuNC03LjYtMy4zLTguNXYtMS4zaDI3LjJ2NDEuNmMwIDEwLjMgNC42IDExIDcuMyAxMSAzLjggMCA2LjYtMy4zIDguMy04VjI5Ljd6TTYxMiA3NS41YzAgMTMuNy0xMi4yIDIzLjQtMjkuNyAyMy40LTkuMiAwLTE4LTQuNi0yMy00LjYtMy42IDAtNC41IDEtNC41IDFsLS44LTE4LjhoMi4xYzkuNiAxMi43IDE2LjIgMTguMyAyNi4xIDE4LjMgNi40IDAgMTEuOC0zLjIgMTEuOC04LjcgMC01LjktNS4zLTcuMy0xMy05LjItMjAuNy01LjMtMjYuNS0xNC4zLTI2LjUtMjYuNCAwLTEzLjQgMTEuNy0yMi41IDI2LjMtMjIuNSA1IDAgOS43IDEgMTMuMyAyLjIgMi4yLjcgOC42IDMuNSAxMi45IDMuMnYxOC4yaC0yQzU5Ni45IDM5LjggNTg5LjkgMzIgNTgxIDMyYy01LjYgMC04LjggMy41LTguOCA4IDAgNS41IDUuMiA3IDE0LjUgOS43IDE2LjggNS4yIDI1LjMgMTEuNyAyNS4zIDI1Ljh6bS04OS4zLTQ1LjhoMjMuOXY2Ny44aC0yMy45di0xNWMtMi40IDYuNi04LjIgMTYuNi0yMC41IDE2LjYtMTIuOCAwLTE5LTEwLjEtMTktMjAuOVYzOS41YzAtNS4xLTEuNC03LjYtMy4zLTguNXYtMS4zaDI3LjJ2NDEuNmMwIDEwLjMgNC42IDExIDcuMyAxMSAzLjggMCA2LjYtMy4zIDguMy04VjI5Ljd6bS04MC43IDY1Yy02LjkgMC05LjQtOS40LTkuNC0zMS4yIDAtMjEuOCAyLjUtMzEuMiA5LjQtMzEuMnM5LjQgOS40IDkuNCAzMS4yYzAgMjEuOC0yLjYgMzEuMi05LjQgMzEuMnptMC02Ni41Yy0yMS42IDAtMzQuMSAxNS44LTM0LjEgMzUuMyAwIDE5LjUgMTIuNSAzNS4zIDM0LjEgMzUuM3MzNC0xNS44IDM0LTM1LjNjMC0xOS41LTEyLjUtMzUuMy0zNC0zNS4zem0tMjk5LjEgMS41aDIzLjl2NjcuOGgtMjMuOVYyOS43em04OSA0NS4zbDEuNS41Yy0xLjcgOS0xMCAyMy41LTI4LjcgMjMuNS0yMS4xIDAtMzEuMi0xNi40LTMxLjItMzUuNCAwLTIwLjQgMTMuNC0zNS41IDMzLjQtMzUuNSAxNC44IDAgMjMuOSA3LjkgMjMuOSAxNi40IDAgNS4yLTMuOSA5LjQtOS43IDkuNC00LjIgMC04LjUtMy45LTguNS04LjcgMC00LjkgMy03LjYgNi41LTguMi0xLjItMi4yLTMuMy0zLjktNy4xLTMuOS02LjUgMC0xMy45IDguMS0xMy45IDI3LjYgMCAxMi44IDYuMiAyNC4xIDE4LjkgMjQuMSAxMC4yIDAgMTQuOS05LjggMTQuOS05Ljh6bTE0NS42LTYzLjdjMC02IDUuMi0xMC44IDExLjctMTAuOCA2LjQgMCAxMS43IDQuOCAxMS43IDEwLjggMCA2LTUuMiAxMC44LTExLjcgMTAuOC02LjUgMC0xMS43LTQuOC0xMS43LTEwLjh6bS0yMzQuNCAwYzAtNiA1LjItMTAuOCAxMS43LTEwLjggNi40IDAgMTEuNyA0LjggMTEuNyAxMC44IDAgNi01LjIgMTAuOC0xMS43IDEwLjgtNi41IDAtMTEuNy00LjgtMTEuNy0xMC44ek0yNi44IDY0LjhDMjQuNyA1MC4xIDI4LjkgMzQgMzQuOSAzNGM2LjEgMCA4LjYgMTIuMSA4LjcgMjYuMmwtMTYuOCA0LjZ6bTcuOC0zNi41QzEyLjkgMjguMyAwIDQ0LjggMCA2My43IDAgNzUgNC41IDg1LjQgMTIuNiA5MmM1LjMgNC42IDEyLjMgNy40IDIwLjkgNy40IDIyLjIgMCAyOS43LTE4LjMgMzAuNy0yMy41aC0uMWMwLS4xIDAtLjIuMS0uM2wtMS41LS41cy03LjQgOS44LTE4LjYgOS44Yy0zLjUgMC02LjgtMS4yLTkuNi0zLjItMy4zLTIuNy01LjYtNi43LTYuNC0xMC45TDY1LjUgNjB2LS4zYy0uOC0xOC4zLTEzLjgtMzEuNC0zMC45LTMxLjR6IiBmaWxsPSIjRUY0MjI5Ii8+PC9zdmc+\")}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .nav-title{background-image:url(\"data:image/svg+xml;base64,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\")}.main-navigation .facebook-social-channel-link{background-image:url(\"data:image/svg+xml;base64,PHN2ZyB2aWV3Qm94PSIwIDAgMTM4IDI1NiIgeG1sbnM9Imh0dHA6Ly93d3cudzMub3JnLzIwMDAvc3ZnIj48cGF0aCBkPSJNMCAxMzcuMDIyaDM5LjE5M1YyNTZoNDkuMDExVjEzNy4wMjJoNDkuMDEyVjg4LjA5SDg4LjIwNFY2My4yN2MwLTguMjkyIDUuMjU3LTE1LjA1NSA5Ljc5Ni0xNS4wNTVoMzkuMjE2VjBIOThDNjUuNTU3LS4wMTIgMzkuMTkzIDI4LjYwNyAzOS4xOTMgNjMuOTc2Vjg4LjA5SDB2NDguOTMyeiIgZmlsbC1ydWxlPSJub256ZXJvIiBmaWxsPSIjRjkzRjIzIi8+PC9zdmc+\")}.branding .facebook-social-channel-link{background-image:url(\"data:image/svg+xml;base64,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\")}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .facebook-link{background-image:url(\"data:image/svg+xml;base64,PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHZpZXdCb3g9Ii02NCA3MjggMTI4IDEyOCI+PHBhdGggZmlsbD0iI2ZmZmZmZiIgZD0iTS02NCA3OTJjMC0zNS4yIDI4LjUtNjQgNjQtNjRzNjQgMjguOCA2NCA2NC0yOC41IDY0LTY0IDY0LTY0LTI4LjYtNjQtNjR6bTQ3LjMgMi4xaDkuNHYyOC44SDQuNXYtMjguOGgxMS44di0xMS44SDQuNXYtNi4xYzAtMi4xIDEuMi0zLjYgMi40LTMuNmg5LjR2LTExLjVINi45Yy03LjkgMC0xNC4zIDctMTQuMyAxNS41djUuOGgtOS40bC4xIDExLjd6Ij48L3BhdGg+PC9zdmc+\")}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .feed-link{background-image:url(\"data:image/svg+xml;base64,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\")}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .googleplus-link{background-image:url(\"data:image/svg+xml;base64,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\")}header[role=\"banner\"] .show-main-navigation:hover{background-image:url(\"data:image/svg+xml;base64,PHN2ZyB2aWV3Qm94PSIwIDAgMjUgMTgiIHhtbG5zPSJodHRwOi8vd3d3LnczLm9yZy8yMDAwL3N2ZyI+PGcgZmlsbD0iIzMzMyIgZmlsbC1ydWxlPSJldmVub2RkIj48cGF0aCBkPSJNMCAwaDI1djJIMHpNMCA4aDI1djJIMHpNMCAxNmgyNXYySDB6Ii8+PC9nPjwvc3ZnPg==\")}header[role=\"banner\"] .show-main-navigation{background-image:url(\"data:image/svg+xml;base64,PHN2ZyB2aWV3Qm94PSIwIDAgMjUgMTgiIHhtbG5zPSJodHRwOi8vd3d3LnczLm9yZy8yMDAwL3N2ZyI+PGcgZmlsbD0iI0ExQTFBMSIgZmlsbC1ydWxlPSJldmVub2RkIj48cGF0aCBkPSJNMCAwaDI1djJIMHpNMCA4aDI1djJIMHpNMCAxNmgyNXYySDB6Ii8+PC9nPjwvc3ZnPg==\")}header[role=\"banner\"] .show-search-button:hover,.search-dialog-form [type=\"submit\"]:hover{background-image:url(\"data:image/svg+xml;base64,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\")}header[role=\"banner\"] .show-search-button,.search-dialog-form [type=\"submit\"]{background-image:url(\"data:image/svg+xml;base64,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\")}.user-status[data-user-type=\"anonymous\"] .recipebox-status:hover{background-image:url(\"data:image/svg+xml;base64,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\")}.user-status[data-user-type=\"anonymous\"] .recipebox-status{background-image:url(\"data:image/svg+xml;base64,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\")}.user-status[data-user-type=\"authenticated\"] .recipebox-status{background-image:url(\"data:image/svg+xml;base64,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\")}.main-navigation .instagram-social-channel-link{background-image:url(\"data:image/svg+xml;base64,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\")}.branding .instagram-social-channel-link{background-image:url(\"data:image/svg+xml;base64,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\")}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .instagram-link{background-image:url(\"data:image/svg+xml;base64,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\")}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .pinterest-link{background-image:url(\"data:image/svg+xml;base64,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\")}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .tumblr-link{background-image:url(\"data:image/svg+xml;base64,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\")}.main-navigation .twitter-social-channel-link{background-image:url(\"data:image/svg+xml;base64,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\")}.branding .twitter-social-channel-link{background-image:url(\"data:image/svg+xml;base64,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\")}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .twitter-link{background-image:url(\"data:image/svg+xml;base64,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\")}.search-dialog-form [type=\"reset\"]:hover{background-image:url(\"data:image/svg+xml;base64,PHN2ZyB2aWV3Qm94PSIwIDAgMjUgMjUiIHhtbG5zPSJodHRwOi8vd3d3LnczLm9yZy8yMDAwL3N2ZyI+PGcgZmlsbD0iIzMzMyI+PHBhdGggZD0iTTIyLjI0NCAwTDI1IDIuNzU2IDIuNzU2IDI1IDAgMjIuMjQ0IDIyLjI0NCAweiIvPjxwYXRoIGQ9Ik0yLjc1NSAwTDI1IDIyLjI0NSAyMi4yNDMgMjUgMCAyLjc1NSAyLjc1NSAweiIvPjwvZz48L3N2Zz4=\")}[class$=\"-form\"] .close-button,.search-dialog-form [type=\"reset\"],.dismiss-main-navigation:hover,[class$=\"-comment-form\"] .close-button{background-image:url(\"data:image/svg+xml;base64,PHN2ZyB2aWV3Qm94PSIwIDAgMjUgMjUiIHhtbG5zPSJodHRwOi8vd3d3LnczLm9yZy8yMDAwL3N2ZyI+PGcgZmlsbD0iI0ExQTFBMSI+PHBhdGggZD0iTTIyLjI0NCAwTDI1IDIuNzU2IDIuNzU2IDI1IDAgMjIuMjQ0IDIyLjI0NCAweiIvPjxwYXRoIGQ9Ik0yLjc1NSAwTDI1IDIyLjI0NSAyMi4yNDMgMjUgMCAyLjc1NSAyLjc1NSAweiIvPjwvZz48L3N2Zz4=\")}.dismiss-main-navigation{background-image:url(\"data:image/svg+xml;base64,PHN2ZyB2aWV3Qm94PSIwIDAgMjUgMjUiIHhtbG5zPSJodHRwOi8vd3d3LnczLm9yZy8yMDAwL3N2ZyI+PGcgZmlsbD0iI0NDQyI+PHBhdGggZD0iTTIyLjI0NCAwTDI1IDIuNzU2IDIuNzU2IDI1IDAgMjIuMjQ0IDIyLjI0NCAweiIvPjxwYXRoIGQ9Ik0yLjc1NSAwTDI1IDIyLjI0NSAyMi4yNDMgMjUgMCAyLjc1NSAyLjc1NSAweiIvPjwvZz48L3N2Zz4=\")}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .youtube-link{background-image:url(\"data:image/svg+xml;base64,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\")}footer[role=\"contentinfo\"] .corporate-info>.section-title,.lightreg-dialog fieldset legend,header[role=\"banner\"] .epicurious-logo,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .nav-title,.main-navigation .facebook-social-channel-link,.branding .facebook-social-channel-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .facebook-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .feed-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .googleplus-link,header[role=\"banner\"] .show-main-navigation,header[role=\"banner\"] .show-search-button,.search-dialog-form [type=\"submit\"],.user-status[data-user-type=\"anonymous\"] .recipebox-status:hover,.user-status[data-user-type=\"anonymous\"] .recipebox-status,.user-status[data-user-type=\"authenticated\"] .recipebox-status,.main-navigation .instagram-social-channel-link,.branding .instagram-social-channel-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .instagram-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .pinterest-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .tumblr-link,.main-navigation .twitter-social-channel-link,.branding .twitter-social-channel-link,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .twitter-link,[class$=\"-form\"] .close-button,.search-dialog-form [type=\"reset\"],.dismiss-main-navigation:hover,[class$=\"-comment-form\"] .close-button,.dismiss-main-navigation,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .youtube-link{border:none;direction:ltr;display:inline-block;overflow:hidden;padding:0;text-align:left;text-indent:-9999px}footer[role=\"contentinfo\"] .corporate-info>.section-title>a,.lightreg-dialog fieldset legend>a,header[role=\"banner\"] .epicurious-logo>a,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .nav-title>a,.main-navigation .facebook-social-channel-link>a,.branding .facebook-social-channel-link>a,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .facebook-link>a,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .feed-link>a,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .googleplus-link>a,header[role=\"banner\"] .show-main-navigation>a,header[role=\"banner\"] .show-search-button>a,.search-dialog-form [type=\"submit\"]>a,.user-status[data-user-type=\"anonymous\"] .recipebox-status:hover>a,.user-status[data-user-type=\"anonymous\"] .recipebox-status>a,.user-status[data-user-type=\"authenticated\"] .recipebox-status>a,.main-navigation .instagram-social-channel-link>a,.branding .instagram-social-channel-link>a,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .instagram-link>a,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .pinterest-link>a,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .tumblr-link>a,.main-navigation .twitter-social-channel-link>a,.branding .twitter-social-channel-link>a,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .twitter-link>a,[class$=\"-form\"] .close-button>a,.search-dialog-form [type=\"reset\"]>a,.dismiss-main-navigation:hover>a,[class$=\"-comment-form\"] .close-button>a,.dismiss-main-navigation>a,footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .youtube-link>a{display:block;height:100%;width:100%}html,body,div,span,applet,object,iframe,h1,h2,h3,h4,h5,h6,p,blockquote,pre,a,abbr,acronym,address,big,cite,code,del,dfn,em,img,ins,kbd,q,s,samp,small,strike,strong,sub,sup,tt,var,b,u,i,center,dl,dt,dd,ol,ul,li,fieldset,form,label,legend,table,caption,tbody,tfoot,thead,tr,th,td,article,aside,canvas,details,embed,figure,figcaption,footer,header,hgroup,menu,nav,output,ruby,section,summary,time,mark,audio,video{margin:0;padding:0;border:0;font:inherit;font-size:100%;vertical-align:baseline}html{line-height:1}ol,ul{list-style:none}table{border-collapse:collapse;border-spacing:0}caption,th,td{text-align:left;font-weight:normal;vertical-align:middle}q,blockquote{quotes:none}q:before,q:after,blockquote:before,blockquote:after{content:\"\";content:none}a img{border:none}article,aside,details,figcaption,figure,footer,header,hgroup,main,menu,nav,section,summary{display:block}html{box-sizing:border-box;font-size:16px;line-height:1.625em}*,*:before,*:after{box-sizing:inherit}* html{font-size:100%}body{background:#fff;font-family:utopia-std, serif;font-weight:400;position:relative;color:#333;-webkit-font-smoothing:antialiased;-moz-osx-font-smoothing:grayscale}@media only screen and (max-width: 640px){body{-webkit-text-size-adjust:none;-ms-text-size-adjust:none;width:100%;min-width:0}}html,body{height:100%}h1,h2,h3,h4,h5,h6{line-height:1.15538em;padding:0.273em 0;text-rendering:optimizeLegibility}p{font-family:\"utopia-std\",serif;font-weight:400;font-size:1rem;margin-bottom:6px;line-height:.75em}[class$=\"-flash-message\"]{border:1px solid #eee;font-size:1rem;line-height:1.25;margin:1rem 0;padding:1rem}.confirm-flash-message{border-color:#83b838;color:#83b838}.error-flash-message{border-color:#f93f23;color:#f93f23}ol{list-style:decimal}img{max-width:100%;height:auto}.apply-social{line-height:0}.social-img picture,.apply-social picture{display:inline-block;line-height:0;position:relative}strong{font-weight:700;line-height:inherit}em{font-style:italic;line-height:inherit}hr{border:1px solid #ccc;clear:both;margin:1rem 0 1.125rem;height:0}.btn,.skiplink{display:inline-block;width:auto;-webkit-appearance:none;text-align:center}.btn>a,.btn input,.btn button,.skiplink>a,.skiplink input,.skiplink button{display:block;height:100%}.btn input,.btn button,.skiplink input,.skiplink button{background:none;border:none;width:100%;font-size:100%;cursor:pointer;-webkit-appearance:none;-moz-appearance:none;appearance:none}html{-webkit-overflow-scrolling:touch}a{color:#333;line-height:normal;text-decoration:none}a:hover,a:active{text-decoration:underline}button{border:none}button:focus{outline:0}button:hover{cursor:pointer}fieldset{border:none;margin:0;padding:0}.field .input{position:relative;padding:0 .625rem;background:#fff;border:1px solid #d8d8d8}img.photo,.photo-wrap img{max-width:none}img{-webkit-transition:opacity .7s;-moz-transition:opacity .7s;-o-transition:opacity .7s;transition:opacity .7s}img.pending{opacity:0}img.loaded{opacity:1}.component-lazy img{opacity:0}.component-lazy.img-loaded img,.component-lazy.img-error img{opacity:1}input{border-radius:0;-webkit-appearance:none}input:focus{outline:0}ul,li{margin:0;padding:0}.flag-inappropriate-content-button{font-family:\"Gotham SSm 7r\",sans-serif;font-weight:normal;color:#a1a1a1;font-size:.75rem;line-height:0rem;background-color:#fff;border:1px solid #a1a1a1;display:block;margin:0 auto;padding:1.4375rem 0 1.375rem;text-transform:uppercase;width:15rem}.flag-inappropriate-content-button:hover{border-color:#333;color:#333}.flagged.flag-inappropriate-content-button,.flagged.flag-inappropriate-content-button:hover{border-color:#f93f23;color:#f93f23;cursor:default}[class$=\"-form\"]{background-color:#fff;display:block;margin:0 auto;padding:2rem 0;width:100%;z-index:1}[class$=\"-form\"] ::selection{background-color:transparent}[class$=\"-form\"] a{color:inherit}[class$=\"-form\"] fieldset,[class$=\"-form\"] label,[class$=\"-form\"] p{font-family:\"Gotham SSm 5r\",sans-serif;font-weight:normal;color:#333;font-size:.875rem;line-height:1.25rem}[class$=\"-form\"] fieldset,[class$=\"-form\"] p{border:none;margin:0;padding:.5rem 2rem}[class$=\"-form\"] p{padding:0}[class$=\"-form\"] legend{font-family:\"Gotham Cond SSm 5r\",sans-serif;font-weight:normal;color:#333;font-size:1.875rem;line-height:1.875rem;margin:0 auto;padding:.5rem 0 .75rem}[class$=\"-form\"] label{cursor:default;display:none}[class$=\"-form\"] .description,[class$=\"-form\"] .warning{margin:1rem 0}[class$=\"-form\"] .warning{color:#f93f23}[class$=\"-form\"] input[type=\"checkbox\"],[class$=\"-form\"] input[type=\"checkbox\"]:focus{border:1px solid #ccc;border-radius:0;height:.75rem;width:.75rem}[class$=\"-form\"] input[type=\"checkbox\"]:checked,[class$=\"-form\"] input[type=\"checkbox\"]:focus:checked{background-color:#f93f23}[class$=\"-form\"] input[type=\"checkbox\"]:hover,[class$=\"-form\"] input[type=\"checkbox\"]:focus:hover{border-color:#333}[class$=\"-form\"] input[type=\"email\"],[class$=\"-form\"] input[type=\"password\"],[class$=\"-form\"] input[type=\"text\"],[class$=\"-form\"] textarea{font-family:\"Gotham SSm 5r\",sans-serif;font-weight:normal;color:#333;font-size:.875rem;line-height:.875rem;border:1px solid #ccc;display:block;margin:1rem 0;padding:1rem .5rem;width:100%}[class$=\"-form\"] input[type=\"email\"].error,[class$=\"-form\"] input[type=\"email\"]:invalid,[class$=\"-form\"] input[type=\"password\"].error,[class$=\"-form\"] input[type=\"password\"]:invalid,[class$=\"-form\"] input[type=\"text\"].error,[class$=\"-form\"] input[type=\"text\"]:invalid,[class$=\"-form\"] textarea.error,[class$=\"-form\"] textarea:invalid{border-color:#f93f23}[class$=\"-form\"] input[type=\"email\"]:focus,[class$=\"-form\"] input[type=\"password\"]:focus,[class$=\"-form\"] input[type=\"text\"]:focus,[class$=\"-form\"] textarea:focus{border-color:#999;outline:none}[class$=\"-form\"] input[type=\"email\"]::selection,[class$=\"-form\"] input[type=\"password\"]::selection,[class$=\"-form\"] input[type=\"text\"]::selection,[class$=\"-form\"] textarea::selection{background-color:#ccc}[class$=\"-form\"] textarea{overflow-y:auto;resize:none}[class$=\"-form\"] .close-button{background-color:transparent;background-repeat:no-repeat;background-position:50% 50%;background-size:.9375rem,.9375rem;border:none;direction:ltr;display:inline-block;height:.9375rem;overflow:hidden;padding:0;text-indent:-9999px;width:.9375rem;position:absolute;right:13px;top:11px}[class$=\"-form\"] .close-button>a{display:block;height:100%;width:100%}[class$=\"-form\"] .cancel-button,[class$=\"-form\"] .save-button,[class$=\"-form\"] .submit-button{font-family:\"Gotham SSm 7r\",sans-serif;font-weight:normal;color:#fff;font-size:.875rem;line-height:0rem;background-color:#333;border:1px solid #333;display:inline-block;margin:36px auto;padding:22px 0 21px;text-transform:uppercase;width:100%}[class$=\"-form\"] .cancel-button:focus,[class$=\"-form\"] .cancel-button:hover,[class$=\"-form\"] .save-button:focus,[class$=\"-form\"] .save-button:hover,[class$=\"-form\"] .submit-button:focus,[class$=\"-form\"] .submit-button:hover{background-color:#292929}[class$=\"-form\"] .cancel-button{background-color:#fff;border-color:#f93f23;color:#f93f23}[class$=\"-form\"] .cancel-button:focus,[class$=\"-form\"] .cancel-button:hover{background-color:#f93f23;color:#fff}[class$=\"-modal-dialog\"] [class$=\"-form\"]{background-color:#fff;display:block;margin:0 auto;padding:25px 0 0;width:90%}@media only screen and (min-width: 768px){[class$=\"-form\"] .close-button{right:15px;top:15px}[class$=\"-form\"] .submit-button{width:158px}}.contact-us-comment-form,.flag-inappropriate-content-comment-form{border:1px solid #ccc}[class$=\"-login-form\"],[class$=\"-registration-form\"],[class$=\"-reset-password-form\"],[class$=\"-update-password-form\"]{text-align:center}[class$=\"-login-form\"] .fine-print,[class$=\"-registration-form\"] .fine-print,[class$=\"-reset-password-form\"] .fine-print,[class$=\"-update-password-form\"] .fine-print{color:#9B9B9B;font-size:.625rem;text-align:center}[class$=\"-login-form\"] .fine-print a,[class$=\"-registration-form\"] .fine-print a,[class$=\"-reset-password-form\"] .fine-print a,[class$=\"-update-password-form\"] .fine-print a{text-decoration:underline}[class$=\"-login-form\"] .goto-counterpart a,[class$=\"-registration-form\"] .goto-counterpart a,[class$=\"-reset-password-form\"] .goto-counterpart a,[class$=\"-update-password-form\"] .goto-counterpart a{text-decoration:underline}[class$=\"-login-form\"] .forgot-password-link{display:block;margin:1rem 0;text-align:right;text-decoration:underline;text-transform:capitalize}.piano-ad-blocker-registration-form .close-button{display:none}.piano-engaged-user-registration-form .close-button{font-family:\"Gotham SSm 7r\",sans-serif;font-weight:normal;color:#ccc;font-size:.625rem;line-height:0rem;background:none;background-color:#e5e5e5;bottom:1rem;height:auto;margin-right:-50px;padding:15px 0 14px;right:50%;text-indent:0;top:initial;width:100px}.piano-engaged-user-registration-form .close-button:hover{color:#333}.piano-ad-blocker-modal-dialog,.piano-login-modal-dialog,.piano-reset-password-modal-dialog{font-family:\"Gotham SSm 5r\",sans-serif;font-weight:normal;color:#333;font-size:.875rem;line-height:1.25rem;padding:2rem 0 0;text-align:center}.piano-ad-blocker-modal-dialog .dialog-title,.piano-login-modal-dialog .dialog-title,.piano-reset-password-modal-dialog .dialog-title{font-family:\"Gotham Cond SSm 5r\",sans-serif;font-weight:normal;color:#f93f23;font-size:1.875rem;line-height:1.875rem;margin:0 auto;padding:8px 0 12px;width:40%}.piano-ad-blocker-modal-dialog .description,.piano-ad-blocker-modal-dialog .your-ad-blocker,.piano-login-modal-dialog .description,.piano-login-modal-dialog .your-ad-blocker,.piano-reset-password-modal-dialog .description,.piano-reset-password-modal-dialog .your-ad-blocker{font-family:\"Gotham SSm 5r\",sans-serif;font-weight:normal;color:#333;font-size:.875rem;line-height:1.25rem;display:block;margin:0 auto;padding:0 2rem}.piano-ad-blocker-modal-dialog .your-ad-blocker .what,.piano-login-modal-dialog .your-ad-blocker .what,.piano-reset-password-modal-dialog .your-ad-blocker .what{background-color:#ccc;display:block;margin:1rem 0;padding:1rem}.piano-ad-blocker-modal-dialog .your-ad-blocker .how,.piano-login-modal-dialog .your-ad-blocker .how,.piano-reset-password-modal-dialog .your-ad-blocker .how{display:block;text-decoration:underline}@media only screen and (min-width: 768px){.piano-ad-blocker-modal-dialog,.piano-login-modal-dialog,.piano-reset-password-modal-dialog{height:96%;top:2%}.piano-ad-blocker-modal-dialog .description,.piano-ad-blocker-modal-dialog .your-ad-blocker,.piano-login-modal-dialog .description,.piano-login-modal-dialog .your-ad-blocker,.piano-reset-password-modal-dialog .description,.piano-reset-password-modal-dialog .your-ad-blocker{padding:0 6rem}}.ReactModalPortal>div{opacity:0;transition:opacity 0.25s}.ReactModalPortal>div.ReactModal__Overlay--after-open{opacity:1}body[data-overlay=\"modal\"],body.ReactModal__Body--open{height:100%;overflow:hidden;padding-right:15px}body[data-overlay=\"modal\"] .content-wrapper,body[data-overlay=\"modal\"] .page-wrap,body.ReactModal__Body--open .content-wrapper,body.ReactModal__Body--open .page-wrap{display:block;filter:blur(5px);-moz-filter:blur(5px);-ms-filter:blur(5px);-o-filter:blur(5px);-webkit-filter:blur(5px)}[class$=\"-modal-dialog-wrapper\"],[class$=\"-modal-confirmation-wrapper\"]{background:rgba(0,0,0,0.7);display:block;height:100%;left:0;position:fixed;top:0;width:100%;z-index:9999999}[class$=\"-modal-confirmation-wrapper\"]{background:none}[class$=\"-modal-confirmation-wrapper\"] .recipbox-modal-confirmation{display:flex;justify-content:center;align-items:center}[class$=\"-modal-confirmation-wrapper\"] .recipbox-modal-confirmation p{font-family:\"Gotham SSm 5r\",sans-serif;font-weight:normal;color:#f93f23;font-size:.875rem;line-height:1.125rem;margin:0}[class$=\"-modal-dialog\"],[class$=\"-modal-confirmation\"]{background-color:#ffffff;display:block;height:90%;left:50%;margin-left:-45%;outline:none;position:fixed;top:5%;width:90%;z-index:999999}[class$=\"-modal-confirmation\"]{height:auto;margin-left:-30%;padding:1.1875rem;text-align:center;top:30%;width:14.5rem}@media only screen and (min-width: 768px){[class$=\"-modal-dialog\"],[class$=\"-modal-confirmation\"]{border:1px solid #ccc;height:80%;left:50%;margin-left:-300px;top:10%;width:600px}[class$=\"-modal-confirmation-wrapper\"] .recipbox-modal-confirmation p{font-family:\"Gotham SSm 5r\",sans-serif;font-weight:normal;color:#f93f23;font-size:1.0625rem;line-height:1.375rem}[class$=\"-modal-confirmation\"]{height:auto;left:50%;margin-left:-10%;padding:1.875rem;top:50%;width:24.875rem}.error-modal-dialog{height:auto}}.content-wrapper{margin:0 auto}.content-wrapper .main-content{-webkit-transition:-webkit-transform .3s linear;-moz-transition:-moz-transform .3s linear;-o-transition:-o-transform .3s linear;transition:transform .3s linear}.content-and-rail{*zoom:1}.content-and-rail:before,.content-and-rail:after{content:\"\";display:table}.content-and-rail:after{clear:both}.fullwidth{width:100%}.hide{display:none}.show{display:block}.locked{position:fixed}.modal.show{visibility:visible;opacity:1}.ad-container{line-height:0}.embed-video{*zoom:1;clear:both;position:relative}.embed-video:before,.embed-video:after{content:\"\";display:table}.embed-video:after{clear:both}.embed-video .video-container{position:relative;padding-bottom:56.25%;padding-top:30px;height:0;overflow:hidden}.embed-video .video-container iframe,.embed-video .video-container object,.embed-video .video-container embed{position:absolute;top:0;left:0;width:100%;height:100%}.embed-social{*zoom:1;clear:both;position:relative}.embed-social:before,.embed-social:after{content:\"\";display:table}.embed-social:after{clear:both}.embed-social .embed-facebook{margin-bottom:20px;position:relative;padding-top:0;padding-bottom:0}.embed-social .embed-facebook iframe,.embed-social .embed-facebook object,.embed-social .embed-facebook embed{position:absolute;top:0;left:0;width:100%;height:100%}.embed-social .embed-instagram{margin-bottom:20px;position:relative;padding-top:30px;padding-bottom:113%}.embed-social .embed-instagram blockquote{margin-left:0;margin-right:0;width:100%}.embed-social .embed-instagram blockquote:before{content:\"\"}@media only screen and (max-width: 640px){.embed-social .embed-instagram{padding-bottom:122%}}.embed-social .embed-instagram iframe,.embed-social .embed-instagram object,.embed-social .embed-instagram embed{position:absolute;top:0;left:0;width:100%;height:100%}.embed-social .embed-twitter{position:relative;padding-top:0px;padding-bottom:0px;text-align:-webkit-center}.embed-social .embed-twitter iframe,.embed-social .embed-twitter object,.embed-social .embed-twitter embed{position:absolute;top:0;left:0;width:100%;height:100%}.cnid-loader{position:fixed;top:45%;left:50%;margin-left:-8px;margin-top:-8px;background-color:#000;width:50px;height:50px;border-radius:10px;opacity:.8;z-index:1101}.cnid-loader img{position:absolute;top:16%;left:16%}.cnid-ios-scroll{-webkit-overflow-scrolling:touch;overflow-y:auto}div.social-img{display:inline-block;position:relative;line-height:0}.social-buttons{position:absolute;z-index:10;bottom:.625rem;right:.625rem;*zoom:1}.social-buttons:before,.social-buttons:after{content:\"\";display:table}.social-buttons:after{clear:both}.social-buttons div{float:left}.social-buttons div.facebook{margin-right:.375rem}.social-buttons div.social{display:block;line-height:0;height:1.75rem;width:1.75rem}.browsehappy{text-align:center;font-weight:normal;font-size:16px;font-family:\"Gotham SSm 5r\";background-color:#f1f2f2;position:fixed;height:26px;bottom:-14px;z-index:99999;width:100%}[class^=\"icon-\"]{background-position:50% 50%;background-repeat:no-repeat;background-size:100% 100%;cursor:pointer;display:block;direction:ltr;overflow:hidden;padding:0;text-indent:-9999px}[class^=\"icon-\"] a{display:block;height:100%;width:100%}.icon-epicurious{height:63px;margin:0 auto;overflow:initial;position:relative;width:304px}.print-controls{display:none}.content-container{width:100%;max-width:1024px;min-width:320px;margin:0 auto;padding-left:10px;padding-right:10px;*zoom:1}.content-container:before,.content-container:after{content:\"\";display:table}.content-container:after{clear:both}.content-container>*:first-child{margin-left:0}@media only screen and (max-width: 640px){.content-container{width:auto;min-width:0;margin-left:0;margin-right:0}}.rail-promo img{width:300px;min-height:300px}.main.has-rail{margin-left:1.953125%;float:left;min-height:1px;position:relative;-webkit-box-sizing:border-box;-moz-box-sizing:border-box;box-sizing:border-box;width:68.1396484375%;margin-left:0}@media only screen and (max-width: 640px){.main.has-rail{float:left;margin-left:0;width:100%}}.main.no-rail{padding:0;margin:0}.main.no-rail:before,.main.no-rail:after{content:\"\";display:table}.main.no-rail:after{clear:both}@media only screen and (min-width: 1024px){.mobile{display:none}body>.content-wrapper>div.main{margin-top:1.25rem}}.rail{margin-left:1.953125%;float:left;min-height:1px;position:relative;-webkit-box-sizing:border-box;-moz-box-sizing:border-box;box-sizing:border-box;width:29.9072265625%;position:static}@media only screen and (max-width: 640px){.rail{float:left;margin-left:0;width:100%}}.rail div.video{margin:0;padding:0}.video-container,#cne-interlude-player-container{width:100%}.video-container.sticky,#cne-interlude-player-container.sticky{position:fixed;width:18.75rem;z-index:15}.weather{width:300px;height:400px;background-color:#ccc}.rail div{margin-bottom:20px}.epi-sticky.stuck{position:fixed;margin-left:0;transform:translateZ(0);-webkit-transform:translateZ(0);z-index:9}.byline,.byline-divider,.pub-date{font:normal .6326875rem \"Gotham SSm 7r\";color:#f93f23;text-transform:uppercase;letter-spacing:normal}.byline.sponsored,.byline-divider.sponsored,.pub-date.sponsored{color:#f93f23;display:inline-block;border:1px solid #f93f23;padding:.375rem .75rem}.byline.mobile,.byline-divider.mobile,.pub-date.mobile{display:none}.pub-date,.byline-divider{color:#a4a4a4}.sponsor-logo{display:block;margin-bottom:.5rem}h3.tag,.sponsor-logo{text-transform:uppercase;margin:0 0 1.375rem 0;padding:0}h3.tag a,.sponsor-logo a{color:#a1a1a1;font:normal 20.169px \"Gotham Cond SSm 5r\"}.sponsor-logo{margin:0;padding:0 0 0.5rem}.fullwidth .section-header{max-width:62.75rem !important}.section-header{width:100%;max-width:1024px;min-width:320px;margin:0 auto;padding-left:10px;padding-right:10px;*zoom:1;text-align:center;overflow:hidden;max-width:62.75rem !important;margin:0 auto;padding:0}.section-header:before,.section-header:after{content:\"\";display:table}.section-header:after{clear:both}.section-header>*:first-child{margin-left:0}@media only screen and (max-width: 640px){.section-header{width:auto;min-width:0;margin-left:0;margin-right:0}}.section-header h2{display:inline-block;position:relative;letter-spacing:.04em}.section-header h2:before{content:\"\";position:absolute;border-top:2px solid #ccc;width:25rem;height:2px;display:block;left:-25.375rem;top:52%}.section-header h2:after{content:\"\";position:absolute;border-top:2px solid #ccc;width:25rem;height:2px;display:block;right:-25.375rem;top:52%}.top5 h2{color:#333;font:normal 50px \"Gotham Cond SSm 5r\";line-height:3.125rem;text-transform:uppercase}@media only screen and (min-width: 641px) and (max-width: 1023px){.mobile{display:none}.nav.mobile{display:block}.content-container{padding:0 1.625rem}.rail{float:none;margin:0;padding:0;text-align:center;width:100%}.content-block{padding:0}.tablet-ads1{*zoom:1;margin:0 0 1.75rem;text-align:center}.tablet-ads1:before,.tablet-ads1:after{content:\"\";display:table}.tablet-ads1:after{clear:both}.tablet-ads1 .ad,.tablet-ads1 .weather,.tablet-ads1 .cook-it{max-width:18.75rem}.tablet-ads1 .ad{margin-left:auto;margin-right:auto}.tablet-ads2{text-align:center}.section-header{overflow:hidden}.section-header h2:before{width:104%;left:-106%}.section-header h2:after{width:104%;right:-106%}.fullwidth .section-header{margin:0 3.25%;width:93.5%}}@media only screen and (max-width: 640px){body{overflow-x:hidden}body.modal-open{overflow:hidden}.content-container{padding:0}.rail{float:none;margin:0;padding:0 .625rem}.rail div{margin-left:auto;margin-right:auto;text-align:center}.rail div.zergnet{max-width:20rem;overflow:hidden}#zergnet-maincontainer{margin:0 auto !important}.content-block{padding:0 1rem}h3.tag a{font:normal 18.369px \"Gotham Cond SSm 5r\"}.section-header h2:before,.section-header h2:after{display:none}.top5 h2{font:normal 40px \"Gotham Cond SSm 5r\";line-height:2.5rem}.progression{*zoom:1}.progression:before,.progression:after{content:\"\";display:table}.progression:after{clear:both}.progression.mobile{display:inline-block}.progression li{text-indent:-9999px;float:left;margin:0 .75rem 0 0;background:url(\"/static/img/mobile-sprite.png\") 0 0 no-repeat;height:7px;width:7px}.progression li:last-child{margin-right:0}.progression li.on{background:url(\"/static/img/mobile-sprite.png\") 0 -7px no-repeat}.byline.sponsored{padding:.375rem .75rem;width:85%;text-align:center;margin-bottom:.25rem}.ad.ad-footer{display:block;position:fixed;bottom:0;margin-bottom:0;width:100%;background-color:#f1f2f2;text-align:center;padding:.375rem 0;z-index:9999}body.parallax-loaded .ad.ad-footer{position:relative}}.lightreg-overlay{background-color:rgba(0,0,0,0.8);bottom:0;top:0;left:0;position:fixed;right:0;z-index:9999999}.lightreg-dialog{background-color:#ffffff;background-clip:padding-box;font:normal .75rem \"Gotham SSm 5r\",sans-serif;height:80%;left:0;margin-left:0;outline:none;position:fixed;top:10%;width:100%}.lightreg-dialog a{color:#333;font-size:.75rem;text-decoration:none}.lightreg-dialog form{padding-top:2.25rem}.lightreg-dialog fieldset{margin:0 auto;padding:0;width:18.4375rem}.lightreg-dialog fieldset legend{display:block;height:3.625rem;margin:0 auto .5rem;width:17.5rem}.lightreg-dialog input:focus,.lightreg-dialog button:focus{outline:0}.lightreg-dialog input::-webkit-input-placeholder{color:#ccc;font-weight:normal;text-transform:uppercase}.lightreg-dialog input:-moz-placeholder{color:#ccc;font-weight:normal;text-transform:uppercase}.lightreg-dialog input::-moz-placeholder{color:#ccc;font-weight:normal;text-transform:uppercase}.lightreg-dialog input:-ms-input-placeholder{color:#ccc;font-weight:normal;text-transform:uppercase}.lightreg-dialog input[type=text],.lightreg-dialog input[type=email],.lightreg-dialog input[type=password]{border:1px solid #ccc;border-radius:0;color:#ccc;display:block;font:normal 1rem/1.125rem \"Gotham SSm 5r\",sans-serif;height:2.8125rem;margin:0 0 .9375rem;padding:0 0 0 .75rem;width:100%;-webkit-appearance:none}.lightreg-dialog input[type=password]{margin-bottom:.875rem}.lightreg-dialog input[type=text].error,.lightreg-dialog input[type=email].error,.lightreg-dialog input[type=password].error,.lightreg-dialog input:focus:invalid:focus,.lightreg-dialog textarea:focus:invalid:focus,.lightreg-dialog select:focus:invalid:focus{color:#f93f23;border:2px solid #f93f23}.lightreg-dialog textarea:focus,.lightreg-dialog input[type=\"text\"]:focus,.lightreg-dialog input[type=\"password\"]:focus,.lightreg-dialog input[type=\"datetime\"]:focus,.lightreg-dialog input[type=\"datetime-local\"]:focus,.lightreg-dialog input[type=\"date\"]:focus,.lightreg-dialog input[type=\"month\"]:focus,.lightreg-dialog input[type=\"time\"]:focus,.lightreg-dialog input[type=\"week\"]:focus,.lightreg-dialog input[type=\"number\"]:focus,.lightreg-dialog input[type=\"email\"]:focus,.lightreg-dialog input[type=\"url\"]:focus,.lightreg-dialog input[type=\"search\"]:focus,.lightreg-dialog input[type=\"tel\"]:focus,.lightreg-dialog input[type=\"color\"]:focus,.lightreg-dialog .uneditable-input:focus{border-color:#ccc}.lightreg-dialog .error-message{color:#f93f23;text-align:left;position:relative;top:-.75rem}.lightreg-dialog button.submit-button{font:normal 400 12.274px \"Gotham SSm 5r\";background:#333;border:0;border-radius:.25rem;color:#fff;cursor:pointer;display:block;font-size:.75rem;height:2.8125rem;margin:1.875rem 0 .8125rem;padding:.375rem .75rem;touch-action:manipulation;user-select:none;width:100%}.lightreg-dialog button.submit-button:hover,.lightreg-dialog button.submit-button:focus,.lightreg-dialog button.submit-button:active,.lightreg-dialog button.submit-button.active,.lightreg-dialog button.submit-button.disabled,.lightreg-dialog button.submit-button[disabled]{background-color:#333;color:#fff}.lightreg-dialog .description{font-family:\"Gotham SSm 5r\", sans-serif;font-size:.75rem;line-height:1rem;margin:0 0 1.25rem}.lightreg-dialog .forgot-password-link{margin:.8125rem 0 0}.lightreg-dialog .goto-counterpart{font-size:.75rem;text-align:center}.lightreg-dialog .fine-print{margin-top:.5625rem}.lightreg-dialog .fine-print,.lightreg-dialog .fine-print a{color:#9B9B9B;font-size:.6875rem;line-height:normal;padding:0;text-align:center}.lightreg-dialog .loading-message{margin:1.5rem 0 0}.lightreg-dialog p.success-message,.lightreg-dialog p.description,.lightreg-dialog p.fine-print{font-family:\"Gotham SSm 5r\", sans-serif}.lightreg-dialog .base-registration-form .fine-print{margin-top:1.5625rem}.lightreg-dialog .base-reset-password-form .success-message,.lightreg-dialog .base-update-password-form .success-message{font-size:1rem;margin-bottom:1.875rem}@media only screen and (min-width: 1024px){.lightreg-dialog{border:1px solid rgba(0,0,0,0.3);left:50%;margin-left:-18.75rem;width:37.5rem}}.loading-message-wrapper{background-color:rgba(255,255,255,0.7);height:100%;left:0;position:absolute;top:0;width:100%;z-index:1}.loading-message-wrapper:hover{cursor:wait}body[data-overlay=\"modal\"] .loading-message-wrapper{background-color:rgba(0,0,0,0.7);position:fixed;z-index:9999}@keyframes loading-spin{100%{transform:rotate(360deg);-webkit-transform:rotate(360deg);-moz-transform:rotate(360deg);-ms-transform:rotate(360deg);-o-transform:rotate(360deg)}}@-webkit-keyframes loading-spin{100%{-webkit-transform:rotate(360deg)}}@-moz-keyframes loading-spin{100%{-moz-transform:rotate(360deg)}}@-ms-keyframes loading-spin{100%{-ms-transform:rotate(360deg)}}@-o-keyframes loading-spin{100%{-o-transform:rotate(360deg)}}.loading-message{animation:loading-spin 0.7s infinite linear;border-bottom:7px solid rgba(161,161,161,0.2);border-left:7px solid rgba(161,161,161,0.2);border-radius:100%;border-right:7px solid rgba(161,161,161,0.2);border-top:7px solid #f93f23;direction:ltr;display:block;height:3rem;left:50%;margin:-1.5rem 0 0 -1.5rem;overflow:hidden;position:absolute;text-indent:-9999px;top:50%;width:3rem;-webkit-animation:loading-spin 0.7s infinite linear;-moz-animation:loading-spin 0.7s infinite linear;-ms-animation:loading-spin 0.7s infinite linear;-o-animation:loading-spin 0.7s infinite linear}.loading-message:hover{cursor:wait}.site-header-wrap{display:block;height:3.75rem}header[role=\"banner\"]{background-color:#fff;box-shadow:0 1px 3px 0 rgba(0,0,0,0.1);font-size:1rem;height:3.75rem;left:0;top:0;transition:top .5s;width:100%;z-index:20}header[role=\"banner\"] ::selection{background-color:transparent}header[role=\"banner\"] a{color:inherit}header[role=\"banner\"] fieldset,header[role=\"banner\"] form,header[role=\"banner\"] legend{margin:0;padding:0}header[role=\"banner\"] .nav-title,header[role=\"banner\"] .section-title{display:none}header[role=\"banner\"] .show-main-navigation{height:1.125rem;left:1.125rem;position:absolute;top:1.375rem;width:1.5625rem}header[role=\"banner\"] .epicurious-logo{height:1.8125rem;left:3.75rem;position:absolute;top:1rem;width:8.625rem}header[role=\"banner\"] .epicurious-logo a{display:block}header[role=\"banner\"] .show-search-button{height:1.625rem;outline:none;position:absolute;right:1.125rem;top:1.0625rem;width:1.625rem}body[data-bannerad-type=\"leaderboard\"][data-bannerad-flow=\"fixed\"][data-siteheader-flow=\"fixed\"][data-bannerad-display=\"hidden\"] header[role=\"banner\"]{top:0}body[data-bannerad-type=\"leaderboard\"][data-bannerad-flow=\"fixed\"][data-siteheader-flow=\"fixed\"][data-bannerad-display=\"visible\"] header[role=\"banner\"]{top:7.125rem}body[data-overlay=\"modal\"] header[role=\"banner\"],body.ReactModal__Body--open header[role=\"banner\"]{left:-8px}body[data-siteheader-flow=\"fixed\"] header[role=\"banner\"]{position:fixed}body[data-siteheader-flow=\"scroll\"] header[role=\"banner\"]{position:relative}@media only screen and (min-width: 768px){header[role=\"banner\"] .show-main-navigation{left:1.5rem}header[role=\"banner\"] .epicurious-logo{left:50%;transform:translateX(-50%)}header[role=\"banner\"] .epicurious-logo:hover{height:1.9375rem;top:.9375rem;width:8.75rem}header[role=\"banner\"] .show-search-button{right:1.25rem;top:1rem}}@media only screen and (min-width: 1024px){header[role=\"banner\"] .show-main-navigation{left:1.3125rem}header[role=\"banner\"] .show-search-button{background-position:100% 50%;background-size:1.3125rem;color:#a1a1a1;font-size:.8125rem;height:0;line-height:0;padding:.6875rem 1.625rem .625rem 0;right:1.4375rem;text-indent:0;top:1.3125rem;width:auto}header[role=\"banner\"] .show-search-button:hover{background-position:100% 50%;color:#333;text-decoration:none}}input:-webkit-autofill,select:-webkit-autofill,textarea:-webkit-autofill{background-color:inherit;background-image:inherit;color:inherit}.global-search-modal-dialog-wrapper{background:rgba(255,255,255,0.9)}.global-search-modal-dialog{background-color:transparent;border:none;height:100%;margin-left:-50%;top:0;width:100%}.global-search-modal-dialog .search-dialog-form{background-color:transparent;margin:0 .9375rem;padding:0;width:calc(100% - 30px)}.search-dialog-form{border-bottom:1px solid #ccc;position:relative}.search-dialog-form fieldset{padding:0}.search-dialog-form input[type=\"text\"]{background-color:transparent;border:0;font-size:1rem;line-height:1;margin:2.25rem 0 .8125rem 1.3125rem;padding:0;width:calc(100% - 20px)}.search-dialog-form input[type=\"text\"]:focus{color:#333}.search-dialog-form input[type=\"text\"]::placeholder{color:#ccc}.search-dialog-form input[type=\"text\"].error{color:#f30}.search-dialog-form [type=\"submit\"]{height:1rem;left:0;position:absolute;top:2.25rem;width:1rem}.search-dialog-form [type=\"submit\"]:hover{cursor:pointer}.search-dialog-form [type=\"reset\"]{background-size:.9375rem .9375rem;height:1.4375rem;position:absolute;right:0;top:.9375rem;width:1.4375rem}.search-dialog-form [type=\"reset\"]:hover{background-size:.9375rem .9375rem}@media only screen and (min-width: 768px){.global-search-modal-dialog .search-dialog-form{margin:0 auto;width:40rem}.search-dialog-form input[type=\"text\"]{font-size:2.6875rem;margin:11.5rem 0 2.125rem 2.875rem;width:calc(100% - 46px)}.search-dialog-form [type=\"submit\"]{height:2rem;left:0;top:12.125rem;width:2rem}.search-dialog-form [type=\"reset\"]{height:1.625rem;top:5.875rem;width:1.625rem}}@media only screen and (min-width: 1024px){.global-search-modal-dialog .search-dialog-form{width:51.875rem}.search-dialog-form [type=\"reset\"]{height:2rem;top:5.875rem;width:2rem}}.site-pushdown-ad-wrap{display:none}body[data-bannerad-type=\"crown\"] .site-header-ad-wrap{display:none}body[data-bannerad-type=\"crown\"] .site-header-ad-wrap.loaded{display:block}body[data-bannerad-type=\"mobile-banner\"],body[data-bannerad-type=\"mobile-leaderboard\"]{padding-bottom:4.625rem}body[data-bannerad-type=\"mobile-banner\"] .site-header-ad-wrap>div,body[data-bannerad-type=\"mobile-banner\"].homepage .site-pushdown-ad-wrap>div,body[data-bannerad-type=\"mobile-leaderboard\"] .site-header-ad-wrap>div,body[data-bannerad-type=\"mobile-leaderboard\"].homepage .site-pushdown-ad-wrap>div{background-color:#f1f2f2;bottom:0;height:3.875rem;overflow:hidden;position:fixed;padding:.375rem 0;text-align:center;width:100%;z-index:20}body[data-bannerad-type=\"leaderboard\"] .site-header-ad-wrap,body[data-bannerad-type=\"leaderboard\"].homepage .site-pushdown-ad-wrap{padding:0}body[data-bannerad-type=\"leaderboard\"] .site-header-ad-wrap>div,body[data-bannerad-type=\"leaderboard\"].homepage .site-pushdown-ad-wrap>div{height:7.125rem;overflow:hidden;padding:.75rem 0;transition:top .5s;width:100%;z-index:20}body[data-bannerad-type=\"leaderboard\"][data-bannerad-flow=\"fixed\"] .site-header-ad-wrap,body[data-bannerad-type=\"leaderboard\"][data-bannerad-flow=\"fixed\"].homepage .site-pushdown-ad-wrap{height:7.125rem}body[data-bannerad-type=\"leaderboard\"][data-bannerad-flow=\"fixed\"] .site-header-ad-wrap>div,body[data-bannerad-type=\"leaderboard\"][data-bannerad-flow=\"fixed\"].homepage .site-pushdown-ad-wrap>div{background-color:#f1f2f2;position:fixed}body[data-bannerad-type=\"leaderboard\"][data-bannerad-display=\"hidden\"] .site-header-ad-wrap>div,body[data-bannerad-type=\"leaderboard\"][data-bannerad-display=\"hidden\"].homepage .site-pushdown-ad-wrap>div{top:-7.125rem}body[data-bannerad-type=\"leaderboard\"][data-bannerad-display=\"visible\"] .site-header-ad-wrap>div,body[data-bannerad-type=\"leaderboard\"][data-bannerad-display=\"visible\"].homepage .site-pushdown-ad-wrap>div{top:0}@media only screen and (min-width: 768px){.site-header-ad-wrap{background-color:#f1f2f2;text-align:center;width:100%}.site-header-ad-wrap>div{display:block;left:0;margin:0 auto;top:0}body[data-bannerad-type=\"pushdown\"] .site-pushdown-ad-wrap{margin-top:1.875rem;margin-bottom:1.875rem}body[data-bannerad-type=\"other\"][data-bannerad-display=\"visible\"] .site-header-ad-wrap{margin-top:4.8125rem}body[data-bannerad-type=\"other\"][data-bannerad-display=\"visible\"] [id^='adUnitContainer']{margin-top:4.8125rem !important}}@media screen, print{.printable .site-header-ad-wrap{display:none}}.main-navigation{font-family:\"Gotham SSm 7r\",sans-serif;font-weight:normal;color:#333;font-size:.875rem;line-height:1em;background-color:#333;color:#fff;height:100%;left:-16.6875rem;letter-spacing:0.1px;line-height:3.125rem;padding-top:1.625rem;position:fixed;top:0;text-align:center;transition:transform 0.3s linear;width:16.6875rem;z-index:22}.main-navigation .logout-user-action{border:1px solid #f93f23;color:#f93f23;padding:.5rem 2rem}.main-navigation[aria-expanded=\"true\"]{-webkit-transform:translateX(16.6875rem);-moz-transform:translateX(16.6875rem);-ms-transform:translateX(16.6875rem);-o-transform:translateX(16.6875rem);transform:translateX(16.6875rem);overflow-y:auto}.dismiss-main-navigation{height:1.0625rem;left:1.625rem;position:absolute;top:1.375rem;width:1.0625rem}@media only screen and (min-width: 768px){.main-navigation{left:-20.125rem;padding-top:2.8125rem;width:20.125rem}.main-navigation[aria-expanded=\"true\"]{-webkit-transform:translateX(20.125rem);-moz-transform:translateX(20.125rem);-ms-transform:translateX(20.125rem);-o-transform:translateX(20.125rem);transform:translateX(20.125rem)}}@media screen, print{.printable .main-navigation{display:none}}.content-channel-links{text-transform:uppercase}[class$=\"-content-channel-link\"]{color:#fff;display:block;margin:1.875rem 0}[class$=\"-content-channel-link\"] a{display:block}[class$=\"-content-channel-link\"] a:hover{color:#f93f23;text-decoration:none}.homepage-content-channel-link{background-color:transparent;background-repeat:no-repeat;background-position:50% 50%;background-size:cover;border:none;direction:ltr;display:inline-block;height:3.4375rem;overflow:hidden;padding:0;text-indent:-9999px;width:3.4375rem;border:1px solid #f93f23;border-radius:100%;display:block;margin:0 auto 2.1875rem}.homepage-content-channel-link>a{display:block;height:100%;width:100%}.curated-content-channel-link{color:#83b838}body.recipes-menus .recipes-menus-content-channel-link,body.expert-advice .expert-advice-content-channel-link,body.ingredients .ingredients-content-channel-link,body.holidays-events .holidays-events-content-channel-link,body.community .community-content-channel-link,body.video .video-content-channel-link{color:#f93f23}@media only screen and (min-width: 768px){.homepage-content-channel-link{height:5.625rem;width:5.625rem}}.main-navigation .social-channel-links .section-title{display:block;font-size:.8125rem;padding:1.25rem 0 .5rem}.main-navigation .facebook-social-channel-link{width:.8125rem}.main-navigation .twitter-social-channel-link{width:1.875rem}.main-navigation .instagram-social-channel-link{width:1.5625rem}.main-navigation [class$=\"-social-channel-link\"]{height:1.5625rem;margin:0 .75rem}.branding .social-channel-links{display:none}@media only screen and (min-width: 1024px){.branding .social-channel-links{display:block;position:absolute;right:7.5rem;top:1rem}.branding .social-channel-links .section-title{color:#a1a1a1;display:inline-block;font-size:.8125rem;height:1em;line-height:1;overflow:hidden;padding:0;vertical-align:middle;width:50px}.branding .social-channel-links ul{display:inline-block;line-height:0;vertical-align:middle}.branding [class$=\"-social-channel-link\"]{height:1.5625rem;margin:0 .375rem;width:1.5625rem}}header[role=\"banner\"] .user-actions{background-color:inherit;font-size:12px;height:auto;padding:0}header[role=\"banner\"] .user-actions [class$=\"-user-action\"]{color:#333}header[role=\"banner\"] .user-actions[data-user-type=\"anonymous\"] [class$=\"-user-action\"]{display:none;position:absolute}header[role=\"banner\"] .user-actions[data-user-type=\"authenticated\"]{left:6.25rem;line-height:1em;position:absolute;top:2.1875rem;white-space:nowrap}header[role=\"banner\"] .user-actions[data-user-type=\"authenticated\"] .user-action-list{border-bottom:1px solid #e5e5e5;display:none;padding-top:16px}header[role=\"banner\"] .user-actions[data-user-type=\"authenticated\"] .user-action-list [class$=\"-user-action\"]{background-color:#fff;border:1px solid #e5e5e5;border-bottom:none;transition:background-color .5s}header[role=\"banner\"] .user-actions[data-user-type=\"authenticated\"] .user-action-list [class$=\"-user-action\"]:hover{background-color:#f93f23;color:#fff}header[role=\"banner\"] .user-actions[data-user-type=\"authenticated\"] .user-action-list [class$=\"-user-action\"] a{display:block;padding:15px}header[role=\"banner\"] .user-actions[data-user-type=\"authenticated\"][data-show-user-actions=\"true\"] .user-action-list{display:block}.user-status{display:block;position:absolute;right:3.8125rem}.user-status .loading-message-wrapper{height:2rem;left:10px;opacity:.25;top:6px;width:2rem}.user-status .loading-message-wrapper .loading-message{height:2rem;width:2rem}.user-status .recipebox-status{display:block;outline:none}.user-status[data-user-type=\"anonymous\"]{top:1.25rem}.user-status[data-user-type=\"anonymous\"] .recipebox-status{height:1.3125rem;width:1.3125rem}.user-status[data-user-type=\"anonymous\"] .login-register-actions{color:#a1a1a1;display:none;font-size:.8125rem;letter-spacing:0.1px;line-height:1;vertical-align:top}.user-status[data-user-type=\"anonymous\"] .login-register-actions a:hover{color:#333;text-decoration:none}.user-status[data-user-type=\"anonymous\"] .login-register-actions [title=\"Log-in\"]:after{color:#a1a1a1;content:\"/\"}.user-status[data-user-type=\"authenticated\"]{outline:none;top:.9375rem}.user-status[data-user-type=\"authenticated\"] .recipebox-status{height:1.75rem;width:1.75rem}.user-status[data-user-type=\"authenticated\"] .recipebox-size{font-family:\"Gotham SSm 7r\",sans-serif;font-weight:normal;color:#f93f23;font-size:.625rem;line-height:.625rem;background-color:#fff;border:solid 1px #a1a1a1;border-radius:100%;bottom:.5rem;display:block;height:1.25rem;overflow:hidden;padding:.25rem 0;position:absolute;right:-.625rem;text-align:center;width:1.25rem}.user-status[data-user-type=\"authenticated\"] .user-greeting{color:#f93f23;display:none}.user-status[data-user-type=\"authenticated\"]:hover .recipebox-size,.user-status[data-user-type=\"authenticated\"]:hover .user-greeting{text-decoration:underline}.main-navigation .user-status[data-user-type=\"anonymous\"]{margin-top:1.125rem;position:static}.main-navigation .user-status[data-user-type=\"anonymous\"] .recipebox-status{display:none}.main-navigation .user-status[data-user-type=\"anonymous\"] .login-register-actions{border:solid 1px #979797;color:#979797;display:inline-block;margin:0 auto;padding:.5rem 2rem}.main-navigation .user-status[data-user-type=\"anonymous\"] .login-register-actions a:hover{color:#f93f23}.main-navigation .user-status[data-user-type=\"authenticated\"]{display:none}@media only screen and (min-width: 768px){.user-status[data-user-type=\"anonymous\"]{top:1.0625rem}.user-status[data-user-type=\"anonymous\"] .recipebox-status{height:1.5rem;width:1.5rem}.user-status[data-user-type=\"authenticated\"]{right:4.3125rem}}@media only screen and (min-width: 1024px){.user-status{left:4.1875rem;right:initial;z-index:21}.user-status[data-user-type=\"anonymous\"]{top:1.3125rem}.user-status[data-user-type=\"anonymous\"] .recipebox-status{height:1.3125rem;width:1.3125rem}.user-status[data-user-type=\"anonymous\"] .login-register-actions{display:inline-block;margin:4px 0 0 12px}.user-status[data-user-type=\"authenticated\"]{right:initial}.user-status[data-user-type=\"authenticated\"] .user-greeting{display:inline-block;font-size:.8125rem;font-weight:normal;left:2.8125rem;letter-spacing:0.1px;position:absolute;top:.375rem;white-space:nowrap}}@media screen, print{.printable header[role=\"banner\"]{width:auto}.printable header[role=\"banner\"] .show-main-nvaigation,.printable header[role=\"banner\"] .user-status,.printable header[role=\"banner\"] .search-dialog{display:none}}[class$=\"-comment-form\"]{left:50%;margin-left:-45%;padding:25px 0 0;position:absolute;top:0}[class$=\"-comment-form\"] .title{font-family:\"Gotham Cond SSm 5r\",sans-serif;font-weight:normal;color:#333;font-size:1.875rem;line-height:1.875rem;padding:0 16px}[class$=\"-comment-form\"] .description,[class$=\"-comment-form\"] .warning{font-family:\"Gotham SSm 5r\",sans-serif;font-weight:normal;color:#333;font-size:.875rem;line-height:1.25rem;margin:12px 0;padding:0 16px}[class$=\"-comment-form\"] .warning{color:#f93f23}[class$=\"-comment-form\"] .actions,[class$=\"-comment-form\"] .fields{border:none;margin:0;padding:0 16px}[class$=\"-comment-form\"] .comment,[class$=\"-comment-form\"] .email,[class$=\"-comment-form\"] .name{font-family:\"Gotham SSm 5r\",sans-serif;font-weight:normal;color:#333;font-size:.875rem;line-height:.875rem;display:block;width:100%;border:1px solid #ccc;margin:.5rem 0;padding:.5rem}[class$=\"-comment-form\"] .comment.error,[class$=\"-comment-form\"] .comment:invalid,[class$=\"-comment-form\"] .email.error,[class$=\"-comment-form\"] .email:invalid,[class$=\"-comment-form\"] .name.error,[class$=\"-comment-form\"] .name:invalid{border-color:#f93f23}[class$=\"-comment-form\"] .comment:focus,[class$=\"-comment-form\"] .email:focus,[class$=\"-comment-form\"] .name:focus{border-color:#999;outline:none}[class$=\"-comment-form\"] .comment{overflow-y:auto;resize:none}[class$=\"-comment-form\"] .close-button{background-color:transparent;background-repeat:no-repeat;background-position:50% 50%;background-size:.9375rem,.9375rem;border:none;direction:ltr;display:inline-block;height:.9375rem;overflow:hidden;padding:0;text-indent:-9999px;width:.9375rem;position:absolute;right:13px;top:11px}[class$=\"-comment-form\"] .close-button>a{display:block;height:100%;width:100%}[class$=\"-comment-form\"] .submit-button{font-family:\"Gotham SSm 7r\",sans-serif;font-weight:normal;color:#fff;font-size:.875rem;line-height:0rem;background-color:#333;display:block;margin:36px auto;padding:23px 0 22px;width:100%}@media only screen and (min-width: 768px){[class$=\"-comment-form\"]{margin-left:-189px;padding:30px 0 0;width:378px}[class$=\"-comment-form\"] .title,[class$=\"-comment-form\"] .description,[class$=\"-comment-form\"] .warning{padding:0 30px}[class$=\"-comment-form\"] .fields,[class$=\"-comment-form\"] .actions{padding:0 30px}[class$=\"-comment-form\"] .name,[class$=\"-comment-form\"] .email,[class$=\"-comment-form\"] .comment{margin:8px 0}[class$=\"-comment-form\"] .close-button{right:15px;top:15px}[class$=\"-comment-form\"] .submit-button{width:158px}}footer[role=\"contentinfo\"]{-webkit-transition:-webkit-transform .3s linear;-moz-transition:-moz-transform .3s linear;-o-transition:-o-transform .3s linear;transition:transform .3s linear;background-color:#f93f23}footer[role=\"contentinfo\"] ::selection{background-color:transparent}footer[role=\"contentinfo\"] ul,footer[role=\"contentinfo\"] li{display:inline;margin:0;padding:0}footer[role=\"contentinfo\"] .epicurious-links,footer[role=\"contentinfo\"] .corporate-info{display:block;margin:0 auto;padding:52px 16px;width:100%}footer[role=\"contentinfo\"] .epicurious-links [class$=\"-link\"]{font-family:\"Gotham SSm 5r\",sans-serif;font-weight:normal;color:#fff;font-size:.625rem;line-height:1.5rem}footer[role=\"contentinfo\"] .epicurious-links>.section-title{display:none}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .nav-title{background-position:0 0;height:1.875rem;margin-bottom:1.25rem;vertical-align:top;width:100%}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .nav-title:hover{cursor:pointer}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .links-list [class$=\"-link\"]{display:inline-block;height:1.75rem;margin-right:.5rem;width:1.75rem}footer[role=\"contentinfo\"] .epicurious-links .channel-links,footer[role=\"contentinfo\"] .epicurious-links .helpful-links,footer[role=\"contentinfo\"] .epicurious-links .fig-links{display:none}footer[role=\"contentinfo\"] .epicurious-links .channel-links .nav-title,footer[role=\"contentinfo\"] .epicurious-links .channel-links .nav-title a,footer[role=\"contentinfo\"] .epicurious-links .helpful-links .nav-title,footer[role=\"contentinfo\"] .epicurious-links .helpful-links .nav-title a,footer[role=\"contentinfo\"] .epicurious-links .fig-links .nav-title,footer[role=\"contentinfo\"] .epicurious-links .fig-links .nav-title a{font-family:\"Gotham SSm 7r\",sans-serif;font-weight:normal;color:#fff;font-size:.875rem;line-height:1.125rem;margin-bottom:18px;text-transform:uppercase}footer[role=\"contentinfo\"] .corporate-info-wrap{background-color:#292929;padding-bottom:62px}footer[role=\"contentinfo\"] .corporate-info{font-family:\"Gotham SSm 5r\",sans-serif;font-weight:normal;color:#868686;font-size:.8125rem;line-height:.8125rem}footer[role=\"contentinfo\"] .corporate-info a:link,footer[role=\"contentinfo\"] .corporate-info a:visited{color:#868686}footer[role=\"contentinfo\"] .corporate-info>.section-title{display:block;height:23px;margin:0 0 17px;width:160px}footer[role=\"contentinfo\"] .corporate-info .conde-nast-brands{display:inline-block;position:relative}footer[role=\"contentinfo\"] .corporate-info .conde-nast-brands .nav-title{font-family:\"Gotham SSm 5r\",sans-serif;font-weight:normal;color:#868686;font-size:.875rem;line-height:.875rem;border:1px solid #5f5d5c;display:inline-block;height:0;line-height:0;padding:19px 13px 19px 0;text-align:center;vertical-align:middle;width:183px}footer[role=\"contentinfo\"] .corporate-info .conde-nast-brands .nav-title:after{height:8px;margin-left:5px;margin-top:-5px;position:absolute;top:50%;transform:rotate(90deg)}footer[role=\"contentinfo\"] .corporate-info .conde-nast-brands .nav-title:hover{cursor:pointer}footer[role=\"contentinfo\"] .corporate-info .conde-nast-brands .conde-nast-brands-list{background-color:#292929;border:1px solid #5f5d5c;bottom:41px;display:none;line-height:21px;padding:10px 0;position:absolute;width:183px}footer[role=\"contentinfo\"] .corporate-info .conde-nast-brands .conde-nast-brands-list .conde-nast-brand{display:block;margin:0 0 0 16px}footer[role=\"contentinfo\"] .corporate-info .conde-nast-brands[aria-hidden=\"false\"] .nav-title:after{margin-top:-1px;transform:rotate(-90deg)}footer[role=\"contentinfo\"] .corporate-info .conde-nast-brands[aria-hidden=\"false\"] .conde-nast-brands-list{display:block}footer[role=\"contentinfo\"] .corporate-info .conde-nast-services,footer[role=\"contentinfo\"] .corporate-info .legal-notice{margin-top:27px}footer[role=\"contentinfo\"] .corporate-info .conde-nast-services{display:none}footer[role=\"contentinfo\"] .corporate-info .conde-nast-services .conde-nast-service{font-family:\"Gotham SSm 7r\",sans-serif;font-weight:normal;color:#868686;font-size:.6875rem;line-height:.6875rem}footer[role=\"contentinfo\"] .corporate-info .legal-notice .title{display:none}footer[role=\"contentinfo\"] .corporate-info .legal-notice p{font-family:\"Gotham SSm 5r\",sans-serif;font-weight:normal;color:#868686;font-size:.75rem;line-height:.875rem;display:inline;margin-right:.375rem}@media only screen and (min-width: 768px){footer[role=\"contentinfo\"] .epicurious-links,footer[role=\"contentinfo\"] .corporate-info{padding:52px 0;width:44.75rem}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links{border-bottom:1px solid #fff;margin-bottom:1.25rem;padding-bottom:1.25rem}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .nav-title{margin:0;width:50%}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .links-list{display:inline-block;text-align:right;width:50%}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .links-list [class$=\"-link\"]{height:1.75rem;margin-left:.5rem;margin-right:0;width:1.75rem}footer[role=\"contentinfo\"] .epicurious-links .channel-links,footer[role=\"contentinfo\"] .epicurious-links .helpful-links,footer[role=\"contentinfo\"] .epicurious-links .fig-links{display:inline-block;width:34%;vertical-align:top}footer[role=\"contentinfo\"] .epicurious-links .channel-links [class$=\"-link\"],footer[role=\"contentinfo\"] .epicurious-links .helpful-links [class$=\"-link\"],footer[role=\"contentinfo\"] .epicurious-links .fig-links [class$=\"-link\"]{display:block}footer[role=\"contentinfo\"] .epicurious-links .fig-links{float:right;width:auto}footer[role=\"contentinfo\"] .corporate-info-wrap{padding-bottom:0}footer[role=\"contentinfo\"] .corporate-info>.section-title{display:inline-block;margin:0 17px 0 0;position:relative;top:5px}footer[role=\"contentinfo\"] .corporate-info .conde-nast-services{display:block}footer[role=\"contentinfo\"] .corporate-info .conde-nast-services .nav-title{display:none}footer[role=\"contentinfo\"] .corporate-info .conde-nast-services .conde-nast-service{display:inline-block;margin:0 16px 0 0;text-transform:uppercase}}@media only screen and (min-width: 1024px){footer[role=\"contentinfo\"] .epicurious-links,footer[role=\"contentinfo\"] .corporate-info{width:62.75rem}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links{border-bottom:none;display:inline-block;width:40%}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .nav-title{margin-bottom:1.25rem;width:50%}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .links-list{display:block;text-align:left;width:100%}footer[role=\"contentinfo\"] .epicurious-links .epi-social-links .links-list [class$=\"-link\"]{font-size:11px;height:1.75rem;margin-left:0;margin-right:.5rem}footer[role=\"contentinfo\"] .epicurious-links .channel-links,footer[role=\"contentinfo\"] .epicurious-links .helpful-links{width:19%}footer[role=\"contentinfo\"] .epicurious-links .fig-links{width:auto}footer[role=\"contentinfo\"] .epicurious-links .helpful-links{margin-left:0}footer[role=\"contentinfo\"] .corporate-info .conde-nast-services{float:right;margin-top:13px}footer[role=\"contentinfo\"] .corporate-info .conde-nast-services .conde-nast-service{margin:0 0 0 16px}}@media screen, print{.printable footer[role=\"contentinfo\"]{display:none}}\n </style>\n <link href=\"/static/css/search__3.11.1.css\" rel=\"stylesheet\"/>\n <script type=\"text/javascript\">\n var digitalData = {\n \"display\" : \"Search Results\",\"pageType\" : \"search\",\"section\" : \"search\",\"canonical\" : \"https://www.epicurious.com/search/%20tofu%20chili\",\"pageValue\" : \"all\",\n };\n </script>\n <script defer=\"\" src=\"https://assets.adobedtm.com/6372cf21ef88ee60bc2977a4898dcb5c7945a212/satelliteLib-b8937e0c536ed51fe92efb5ca25fc6f76fdeef0e.js\">\n </script>\n <script type=\"text/javascript\">\n !function (e){\"use strict\";var t=e||{};window.cns=window.cns||{async:function(e,t){this.queue[e]||(this.queue[e]=[]),\"function\"==typeof t&&this.queue[e].push(t)},queue:{},embed:{backup:t.overwrite}};var n=document.createElement(\"script\");t.isAsync?n.async=!0:n.defer=!0,n.src=function(){if(t.overwrite)return t.overwrite;var e=location.hostname,n=\"\"!==location.port?\":\"+location.port:\"\",o=location.search.match(/CNS_SB=([^&\\s]+)[&\\s]?/)||[],s=o.length?\"-sandbox/\"+o[1]:\"\",c=!t.bucketer||s?\"\":t.bucketer(),i=\"\";if(!1!==t.bustCache){var r=new Date;r.setMinutes(Math.floor(r.getMinutes()/15),0,0),i+=\"?\"+r.getTime()}return\"//\"+(e+n)+\"/cns\"+s+c+\"/services.min.js\"+i}(),document.head.appendChild(n)}({});\n </script>\n </head>\n <body class=\"search-page\">\n \n <span class=\"page-wrap\" id=\"react-app\">\n <span class=\"page\" data-react-checksum=\"1783234613\" data-reactid=\"1\" data-reactroot=\"\">\n <div class=\"header-wrapper\" data-reactid=\"2\">\n <div class=\"header\" data-reactid=\"3\" role=\"banner\">\n <h2 data-reactid=\"4\" itemtype=\"https://schema.org/Organization\">\n <a data-reactid=\"5\" href=\"/\" itemprop=\"url\" title=\"Epicurious\">\n Epicurious\n </a>\n </h2>\n <div class=\"search-form-container\" data-reactid=\"6\">\n <form action=\"/search/\" autocomplete=\"off\" data-reactid=\"7\" method=\"get\" role=\"search\">\n <fieldset data-reactid=\"8\">\n <button class=\"submit\" data-reactid=\"9\" type=\"submit\">\n search\n </button>\n <input autocomplete=\"off\" data-reactid=\"10\" maxlength=\"120\" name=\"terms\" placeholder=\"Find a Recipe\" type=\"text\" value=\" tofu chili\"/>\n <button class=\"filter mobile\" data-reactid=\"11\">\n filters\n </button>\n <button class=\"filter tablet\" data-reactid=\"12\">\n filter results\n </button>\n </fieldset>\n </form>\n <div class=\"ingredient-filters\" data-reactid=\"13\">\n <h3 data-reactid=\"14\">\n Include/Exclude Ingredients\n </h3>\n <form class=\"include-ingredients\" data-reactid=\"15\">\n <fieldset data-reactid=\"16\">\n <input aria-label=\"Include ingredients\" data-reactid=\"17\" placeholder=\"Include ingredients:\" type=\"text\"/>\n <button data-reactid=\"18\">\n include\n </button>\n </fieldset>\n </form>\n <form class=\"exclude-ingredients\" data-reactid=\"19\">\n <fieldset data-reactid=\"20\">\n <input aria-label=\"Exclude ingredients\" data-reactid=\"21\" placeholder=\"Exclude ingredients:\" type=\"text\"/>\n <button data-reactid=\"22\">\n exclude\n </button>\n </fieldset>\n </form>\n </div>\n <button class=\"filter ingredient-filter\" data-reactid=\"23\">\n include/exclude ingredients\n </button>\n <div class=\"search-tags\" data-reactid=\"24\">\n <button class=\"clear-all\" data-reactid=\"25\">\n Clear all\n </button>\n </div>\n </div>\n </div>\n <div class=\"filters\" data-reactid=\"26\">\n <div data-reactid=\"27\">\n \n <div class=\"filter-action-panel\" data-reactid=\"29\">\n <p data-reactid=\"30\">\n \n 18\n \n \n matching results\n \n </p>\n <button data-reactid=\"33\">\n Apply\n </button>\n <button data-reactid=\"34\">\n Cancel\n </button>\n </div>\n </div>\n </div>\n </div>\n <div class=\"search-results-status\" data-reactid=\"35\">\n <div class=\"search-tags\" data-reactid=\"36\">\n <button class=\"clear-all\" data-reactid=\"37\">\n Clear all\n </button>\n </div>\n <p class=\"results-message\" data-reactid=\"38\">\n <span class=\"matching-count\" data-reactid=\"39\">\n 18\n </span>\n \n matching results for\n \n <em class=\"terms\" data-reactid=\"41\">\n tofu chili\n </em>\n </p>\n <div class=\"search-sort-container\" data-reactid=\"42\">\n <div class=\"content-type-filter-options-menu\" data-reactid=\"43\">\n <label aria-expanded=\"false\" data-reactid=\"44\" for=\"content-type-filter-options-menu\" role=\"button\">\n <span class=\"options-menu-title\" data-reactid=\"45\">\n Show\n </span>\n <span class=\"selected-option-text\" data-reactid=\"46\">\n All Content\n </span>\n </label>\n <select data-reactid=\"47\" id=\"content-type-filter-options-menu\" name=\"content-type-filter\">\n <option data-reactid=\"48\" selected=\"\" value=\"all\">\n All Content\n </option>\n <option data-reactid=\"49\" value=\"recipe\">\n Recipes\n </option>\n <option data-reactid=\"50\" value=\"article\">\n Articles\n </option>\n <option data-reactid=\"51\" value=\"package\">\n Guides\n </option>\n <option data-reactid=\"52\" value=\"menu\">\n Menus\n </option>\n <option data-reactid=\"53\" value=\"gallery\">\n Galleries\n </option>\n <option data-reactid=\"54\" value=\"cnevideo\">\n Videos\n </option>\n </select>\n </div>\n <div class=\"sort-filter-options-menu\" data-reactid=\"55\">\n <label aria-expanded=\"false\" data-reactid=\"56\" for=\"sort-filter-options-menu\" role=\"button\">\n <span class=\"options-menu-title\" data-reactid=\"57\">\n Sort by\n </span>\n <span class=\"selected-option-text\" data-reactid=\"58\">\n Relevance\n </span>\n </label>\n <select data-reactid=\"59\" id=\"sort-filter-options-menu\" name=\"sort-filter\">\n <option data-reactid=\"60\" selected=\"\" value=\"relevance\">\n Relevance\n </option>\n <option data-reactid=\"61\" value=\"newest\">\n Newest\n </option>\n <option data-reactid=\"62\" value=\"highestRated\">\n Highest Rated\n </option>\n <option data-reactid=\"63\" value=\"mostReviewed\">\n Most Reviewed\n </option>\n <option data-reactid=\"64\" value=\"makeItAgain\">\n Make Again %\n </option>\n </select>\n </div>\n </div>\n </div>\n <section data-query=\" tofu chili\" data-reactid=\"65\" role=\"main\">\n <h3 class=\"section-title\" data-reactid=\"66\">\n Search Results\n </h3>\n <div class=\"results-group\" data-group-number=\"1\" data-reactid=\"67\">\n <article class=\"recipe-content-card\" data-has-quickview=\"false\" data-index=\"0\" data-reactid=\"68\" itemscope=\"\" itemtype=\"https://schema.org/Recipe\">\n <header class=\"summary\" data-reactid=\"69\">\n <strong class=\"tag\" data-reactid=\"70\">\n recipe\n </strong>\n <h4 class=\"hed\" data-reactid=\"71\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"72\" href=\"/recipes/food/views/spicy-lemongrass-tofu-233844\">\n Spicy Lemongrass Tofu\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"73\" data-truncate=\"1\">\n Dau hu xa ot\nEditor's note: The recipe and introductory text below are excerpted from Pleasures of the Vietnamese Table by Mai Pham and are part of our story on Lunar New Year.\nWhile traveling on a train one time to the coastal town of Nha Trang, I sat next to an elderly nun. Over the course of our bumpy eight-hour ride, she shared stories of life at the temple and the difficult years after the end of the war when the Communist government cracked down on religious factions. Toward the end of our chat, she pulled out a bag of food she'd prepared for the trip. It was tofu that had been cooked in chilies, lemongrass and la lot, an aromatic leaf also known as pepper leaf. When she gave me a taste, I knew immediately that I had to learn how to make it. This is my rendition of that fabulous dish. Make sure to pat the tofu dry before marinating it and use very fresh lemongrass. I always love serving this to friends who think tofu dishes are bland.\n </p>\n <dl class=\"recipes-ratings-summary\" data-reactid=\"74\" data-reviews-count=\"17\" data-reviews-rating=\"3.38\" itemprop=\"aggregateRating\" itemscope=\"\" itemtype=\"https://schema.org/AggregateRating\">\n <dt class=\"rating-label\" data-reactid=\"75\">\n Average user rating\n </dt>\n <span class=\"reviews-count-container\" data-reactid=\"76\">\n <dd class=\"rating\" data-rating=\"3.5\" data-reactid=\"77\">\n <span data-reactid=\"78\" itemprop=\"ratingValue\">\n 3.5\n </span>\n \n /\n \n <span data-reactid=\"80\" itemprop=\"bestRating\">\n 4\n </span>\n <meta content=\"0\" data-reactid=\"81\" itemprop=\"worstRating\"/>\n </dd>\n <dt class=\"reviews-count-label\" data-reactid=\"82\">\n Reviews\n </dt>\n <dd class=\"reviews-count\" data-reactid=\"83\" itemprop=\"reviewCount\">\n 17\n </dd>\n </span>\n <span class=\"make-again-container\" data-reactid=\"84\">\n <dt class=\"make-again-percentage-label\" data-reactid=\"85\">\n Percentage of reviewers who will make this recipe again\n </dt>\n <dd class=\"make-again-percentage\" data-reactid=\"86\">\n \n 88\n \n \n %\n \n </dd>\n </span>\n </dl>\n </header>\n <a class=\"photo-link\" data-reactid=\"89\" href=\"/recipes/food/views/spicy-lemongrass-tofu-233844\">\n <div class=\"photo-wrap\" data-reactid=\"90\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"91\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"92\" href=\"/recipes/food/views/spicy-lemongrass-tofu-233844\" itemprop=\"url\" title=\"Spicy Lemongrass Tofu\">\n \n View “\n \n \n Spicy Lemongrass Tofu\n \n \n ”\n \n </a>\n <div class=\"recipe-panel \" data-reactid=\"96\">\n <a class=\"view-complete-item\" data-reactid=\"97\" href=\"/recipes/food/views/spicy-lemongrass-tofu-233844\">\n View Recipe\n </a>\n <div class=\"controls\" data-reactid=\"98\">\n <a class=\"show-quick-view\" data-reactid=\"99\" href=\"/recipes/food/views/spicy-lemongrass-tofu-233844\" title=\"Spicy Lemongrass Tofu\">\n Quick view\n </a>\n <a class=\"toggle-compare-state\" data-reactid=\"100\">\n Compare Recipe\n </a>\n </div>\n </div>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"1\" data-reactid=\"101\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"102\">\n <strong class=\"tag\" data-reactid=\"103\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"104\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"105\" href=\"/expert-advice/ground-turkey-breast-versus-ground-dark-meat-article\">\n Ground Turkey Breast Must Be Stopped\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"106\" data-truncate=\"1\">\n Those packages of pink meat might not be as healthy as you think.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"107\" href=\"/expert-advice/ground-turkey-breast-versus-ground-dark-meat-article\">\n <div class=\"photo-wrap\" data-reactid=\"108\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"109\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"110\" href=\"/expert-advice/ground-turkey-breast-versus-ground-dark-meat-article\" itemprop=\"url\" title=\"Ground Turkey Breast Must Be Stopped\">\n \n View “\n \n \n Ground Turkey Breast Must Be Stopped\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"2\" data-reactid=\"114\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"115\">\n <strong class=\"tag\" data-reactid=\"116\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"117\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"118\" href=\"/expert-advice/how-to-make-pad-thai-in-22-minutes-article\">\n How to Make Better-Than-Takeout Pad Thai in 22 Minutes\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"119\" data-truncate=\"1\">\n Takeout is convenient and all. But in the time it takes to wait for delivery, you can make lighter, brighter, fresher pad Thai at home.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"120\" href=\"/expert-advice/how-to-make-pad-thai-in-22-minutes-article\">\n <div class=\"photo-wrap\" data-reactid=\"121\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"122\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"123\" href=\"/expert-advice/how-to-make-pad-thai-in-22-minutes-article\" itemprop=\"url\" title=\"How to Make Better-Than-Takeout Pad Thai in 22 Minutes\">\n \n View “\n \n \n How to Make Better-Than-Takeout Pad Thai in 22 Minutes\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"3\" data-reactid=\"127\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"128\">\n <strong class=\"tag\" data-reactid=\"129\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"130\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"131\" href=\"/ingredients/four-hot-sauces-beyond-sriracha-article\">\n Forget Sriracha: New Ways to Spice Up Everything\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"132\" data-truncate=\"1\">\n Sriracha's great and all, but we've moved on.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"133\" href=\"/ingredients/four-hot-sauces-beyond-sriracha-article\">\n <div class=\"photo-wrap\" data-reactid=\"134\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"135\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"136\" href=\"/ingredients/four-hot-sauces-beyond-sriracha-article\" itemprop=\"url\" title=\"Forget Sriracha: New Ways to Spice Up Everything\">\n \n View “\n \n \n Forget Sriracha: New Ways to Spice Up Everything\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"4\" data-reactid=\"140\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"141\">\n <strong class=\"tag\" data-reactid=\"142\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"143\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"144\" href=\"/expert-advice/its-time-to-up-your-instant-ramen-game-article\">\n It's Time To Up Your Instant Ramen Game\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"145\" data-truncate=\"1\">\n Upgrade your instant ramen with help from Chef Bill Kim of UrbanBelly in Chicago.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"146\" href=\"/expert-advice/its-time-to-up-your-instant-ramen-game-article\">\n <div class=\"photo-wrap\" data-reactid=\"147\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"148\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"149\" href=\"/expert-advice/its-time-to-up-your-instant-ramen-game-article\" itemprop=\"url\" title=\"It's Time To Up Your Instant Ramen Game\">\n \n View “\n \n \n It's Time To Up Your Instant Ramen Game\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"5\" data-reactid=\"153\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"154\">\n <strong class=\"tag\" data-reactid=\"155\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"156\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"157\" href=\"/ingredients/frozen-ethnic-international-thai-chinese-indian-ingredients-article\">\n The International Ingredients You Need in Your Freezer Right Now\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"158\" data-truncate=\"1\">\n With these ingredients always on hand, there's no Thai (or Indian) (or Chinese) recipe you can't try.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"159\" href=\"/ingredients/frozen-ethnic-international-thai-chinese-indian-ingredients-article\">\n <div class=\"photo-wrap\" data-reactid=\"160\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"161\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"162\" href=\"/ingredients/frozen-ethnic-international-thai-chinese-indian-ingredients-article\" itemprop=\"url\" title=\"The International Ingredients You Need in Your Freezer Right Now\">\n \n View “\n \n \n The International Ingredients You Need in Your Freezer Right Now\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"6\" data-reactid=\"166\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"167\">\n <strong class=\"tag\" data-reactid=\"168\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"169\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"170\" href=\"/recipes-menus/easy-vegetarian-dinners-for-every-night-of-the-week-article\">\n Easy Vegetarian Dinners for Every Night of the Week\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"171\" data-truncate=\"1\">\n A week of healthy meat-free dinners to get ready for spring.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"172\" href=\"/recipes-menus/easy-vegetarian-dinners-for-every-night-of-the-week-article\">\n <div class=\"photo-wrap\" data-reactid=\"173\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"174\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"175\" href=\"/recipes-menus/easy-vegetarian-dinners-for-every-night-of-the-week-article\" itemprop=\"url\" title=\"Easy Vegetarian Dinners for Every Night of the Week\">\n \n View “\n \n \n Easy Vegetarian Dinners for Every Night of the Week\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"7\" data-reactid=\"179\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"180\">\n <strong class=\"tag\" data-reactid=\"181\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"182\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"183\" href=\"/expert-advice/how-to-make-grain-noodle-bowls-lukas-volger-article\">\n How to Become a Pro Bowler\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"184\" data-truncate=\"1\">\n Bowl food doesn't get beautiful (and nutritious) (and filling) automatically—it takes a plan. And there's nobody better to get that plan from than the guy who literally wrote the book on the subject.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"185\" href=\"/expert-advice/how-to-make-grain-noodle-bowls-lukas-volger-article\">\n <div class=\"photo-wrap\" data-reactid=\"186\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"187\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"188\" href=\"/expert-advice/how-to-make-grain-noodle-bowls-lukas-volger-article\" itemprop=\"url\" title=\"How to Become a Pro Bowler\">\n \n View “\n \n \n How to Become a Pro Bowler\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"8\" data-reactid=\"192\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"193\">\n <strong class=\"tag\" data-reactid=\"194\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"195\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"196\" href=\"/expert-advice/emily-gould-cooking-diary-article\">\n What the Writer-Publisher-Mom Emily Gould Cooks in a Week\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"197\" data-truncate=\"1\">\n The novelist and independent book publisher relies on intuition, leftovers, and some very good cookbooks to make dinner happen.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"198\" href=\"/expert-advice/emily-gould-cooking-diary-article\">\n <div class=\"photo-wrap\" data-reactid=\"199\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"200\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"201\" href=\"/expert-advice/emily-gould-cooking-diary-article\" itemprop=\"url\" title=\"What the Writer-Publisher-Mom Emily Gould Cooks in a Week \">\n \n View “\n \n \n What the Writer-Publisher-Mom Emily Gould Cooks in a Week\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"9\" data-reactid=\"205\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"206\">\n <strong class=\"tag\" data-reactid=\"207\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"208\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"209\" href=\"/expert-advice/how-to-throw-hot-pot-party-at-home-article\">\n How to Throw a Hot Pot Party With a Slow Cooker\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"210\" data-truncate=\"1\">\n It's the all-you-can-eat fondue party that, thankfully, isn't a fondue party at all.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"211\" href=\"/expert-advice/how-to-throw-hot-pot-party-at-home-article\">\n <div class=\"photo-wrap\" data-reactid=\"212\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"213\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"214\" href=\"/expert-advice/how-to-throw-hot-pot-party-at-home-article\" itemprop=\"url\" title=\"How to Throw a Hot Pot Party With a Slow Cooker\">\n \n View “\n \n \n How to Throw a Hot Pot Party With a Slow Cooker\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"10\" data-reactid=\"218\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"219\">\n <strong class=\"tag\" data-reactid=\"220\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"221\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"222\" href=\"/expert-advice/hearty-vegetarian-soup-miso-tahini-recipe-vegan-article\">\n The Unlikely Ingredient That's Key to Making This Hearty Vegetarian Soup\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"223\" data-truncate=\"1\">\n You probably have it in the fridge already.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"224\" href=\"/expert-advice/hearty-vegetarian-soup-miso-tahini-recipe-vegan-article\">\n <div class=\"photo-wrap\" data-reactid=\"225\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"226\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"227\" href=\"/expert-advice/hearty-vegetarian-soup-miso-tahini-recipe-vegan-article\" itemprop=\"url\" title=\"The Unlikely Ingredient That's Key to Making This Hearty Vegetarian Soup\">\n \n View “\n \n \n The Unlikely Ingredient That's Key to Making This Hearty Vegetarian Soup\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"11\" data-reactid=\"231\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"232\">\n <strong class=\"tag\" data-reactid=\"233\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"234\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"235\" href=\"/expert-advice/how-to-encourage-your-kids-to-try-spicy-foods-article\">\n How to Get Your Kids to Eat—and Maybe Even Like—Spicy Foods\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"236\" data-truncate=\"1\">\n For starters, don't underestimate their taste preferences.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"237\" href=\"/expert-advice/how-to-encourage-your-kids-to-try-spicy-foods-article\">\n <div class=\"photo-wrap\" data-reactid=\"238\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"239\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"240\" href=\"/expert-advice/how-to-encourage-your-kids-to-try-spicy-foods-article\" itemprop=\"url\" title=\"How to Get Your Kids to Eat—and Maybe Even Like—Spicy Foods\">\n \n View “\n \n \n How to Get Your Kids to Eat—and Maybe Even Like—Spicy Foods\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"12\" data-reactid=\"244\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"245\">\n <strong class=\"tag\" data-reactid=\"246\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"247\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"248\" href=\"/expert-advice/what-is-the-low-fodmap-diet-article\">\n What the Heck Is the Low FODMAP Diet?\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"249\" data-truncate=\"1\">\n This is not your average low-carb eating plan.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"250\" href=\"/expert-advice/what-is-the-low-fodmap-diet-article\">\n <div class=\"photo-wrap\" data-reactid=\"251\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"252\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"253\" href=\"/expert-advice/what-is-the-low-fodmap-diet-article\" itemprop=\"url\" title=\"What the Heck Is the Low FODMAP Diet?\">\n \n View “\n \n \n What the Heck Is the Low FODMAP Diet?\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"13\" data-reactid=\"257\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"258\">\n <strong class=\"tag\" data-reactid=\"259\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"260\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"261\" href=\"/expert-advice/10-vegan-groceries-to-buy-every-single-week-recipes-article\">\n 10 Groceries to Buy Every Single Week If You're Vegan\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"262\" data-truncate=\"1\">\n From dried lentils to cashew butter, here are the 10 things vegans should never leave the store without.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"263\" href=\"/expert-advice/10-vegan-groceries-to-buy-every-single-week-recipes-article\">\n <div class=\"photo-wrap\" data-reactid=\"264\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"265\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"266\" href=\"/expert-advice/10-vegan-groceries-to-buy-every-single-week-recipes-article\" itemprop=\"url\" title=\"10 Groceries to Buy Every Single Week If You're Vegan\">\n \n View “\n \n \n 10 Groceries to Buy Every Single Week If You're Vegan\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"14\" data-reactid=\"270\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"271\">\n <strong class=\"tag\" data-reactid=\"272\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"273\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"274\" href=\"/expert-advice/why-my-kids-eat-indian-food-almost-every-night-article\">\n Here's Why My Kids Eat Indian Food Almost Every Night\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"275\" data-truncate=\"1\">\n Chicago-based cookbook author Anupy Singla changed careers in order to change her family's eating habits. It worked.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"276\" href=\"/expert-advice/why-my-kids-eat-indian-food-almost-every-night-article\">\n <div class=\"photo-wrap\" data-reactid=\"277\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"278\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"279\" href=\"/expert-advice/why-my-kids-eat-indian-food-almost-every-night-article\" itemprop=\"url\" title=\"Here's Why My Kids Eat Indian Food Almost Every Night\">\n \n View “\n \n \n Here's Why My Kids Eat Indian Food Almost Every Night\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"15\" data-reactid=\"283\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"284\">\n <strong class=\"tag\" data-reactid=\"285\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"286\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"287\" href=\"/ingredients/ultimate-guide-to-buying-pickles-article\">\n How to Navigate the Pickle Aisle\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"288\" data-truncate=\"1\">\n Everything you need to know when buying pickles.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"289\" href=\"/ingredients/ultimate-guide-to-buying-pickles-article\">\n <div class=\"photo-wrap\" data-reactid=\"290\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"291\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"292\" href=\"/ingredients/ultimate-guide-to-buying-pickles-article\" itemprop=\"url\" title=\"How to Navigate the Pickle Aisle\">\n \n View “\n \n \n How to Navigate the Pickle Aisle\n \n \n ”\n \n </a>\n </article>\n <article class=\"article-content-card\" data-has-quickview=\"false\" data-index=\"16\" data-reactid=\"296\" itemscope=\"\" itemtype=\"https://schema.org/ItemPage\">\n <header class=\"summary\" data-reactid=\"297\">\n <strong class=\"tag\" data-reactid=\"298\">\n article\n </strong>\n <h4 class=\"hed\" data-reactid=\"299\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"300\" href=\"/recipes-menus/fire-cider-recipe-cooking-tips-article\">\n How to Make Fire Cider at Home—and Cook With It\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"301\" data-truncate=\"1\">\n This old-school herbal remedy can be turned into new-school cocktails, salads, and more.\n </p>\n </header>\n <a class=\"photo-link\" data-reactid=\"302\" href=\"/recipes-menus/fire-cider-recipe-cooking-tips-article\">\n <div class=\"photo-wrap\" data-reactid=\"303\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"304\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"305\" href=\"/recipes-menus/fire-cider-recipe-cooking-tips-article\" itemprop=\"url\" title=\"How to Make Fire Cider at Home—and Cook With It\">\n \n View “\n \n \n How to Make Fire Cider at Home—and Cook With It\n \n \n ”\n \n </a>\n </article>\n <article class=\"recipe-content-card\" data-has-quickview=\"false\" data-index=\"17\" data-reactid=\"309\" itemscope=\"\" itemtype=\"https://schema.org/Recipe\">\n <header class=\"summary\" data-reactid=\"310\">\n <strong class=\"tag\" data-reactid=\"311\">\n recipe\n </strong>\n <h4 class=\"hed\" data-reactid=\"312\" data-truncate=\"3\" itemprop=\"name\">\n <a data-reactid=\"313\" href=\"/recipes/food/views/chinese-egg-noodles-with-smoked-duck-and-snow-peas-354302\">\n Chinese Egg Noodles with Smoked Duck and Snow Peas\n </a>\n </h4>\n <p class=\"dek\" data-reactid=\"314\" data-truncate=\"1\">\n If you live near a Chinese market, pick up barbecued or smoked duck there. Otherwise, smoked chicken or turkey from the supermarket (or leftover roast chicken) would be terrific tossed with the noodles. To make it a meal, add a platter of chilled silken tofu. Drizzle the tofu with soy sauce and chili sauce, then top with chopped green onions. Coconut ice cream with fresh berries and lychees would make a nice dessert.\n </p>\n <dl class=\"recipes-ratings-summary\" data-reactid=\"315\" data-reviews-count=\"4\" data-reviews-rating=\"2.25\" itemprop=\"aggregateRating\" itemscope=\"\" itemtype=\"https://schema.org/AggregateRating\">\n <dt class=\"rating-label\" data-reactid=\"316\">\n Average user rating\n </dt>\n <span class=\"reviews-count-container\" data-reactid=\"317\">\n <dd class=\"rating\" data-rating=\"2.5\" data-reactid=\"318\">\n <span data-reactid=\"319\" itemprop=\"ratingValue\">\n 2.5\n </span>\n \n /\n \n <span data-reactid=\"321\" itemprop=\"bestRating\">\n 4\n </span>\n <meta content=\"0\" data-reactid=\"322\" itemprop=\"worstRating\"/>\n </dd>\n <dt class=\"reviews-count-label\" data-reactid=\"323\">\n Reviews\n </dt>\n <dd class=\"reviews-count\" data-reactid=\"324\" itemprop=\"reviewCount\">\n 4\n </dd>\n </span>\n <span class=\"make-again-container\" data-reactid=\"325\">\n <dt class=\"make-again-percentage-label\" data-reactid=\"326\">\n Percentage of reviewers who will make this recipe again\n </dt>\n <dd class=\"make-again-percentage\" data-reactid=\"327\">\n \n 50\n \n \n %\n \n </dd>\n </span>\n </dl>\n </header>\n <a class=\"photo-link\" data-reactid=\"330\" href=\"/recipes/food/views/chinese-egg-noodles-with-smoked-duck-and-snow-peas-354302\">\n <div class=\"photo-wrap\" data-reactid=\"331\">\n <div class=\"component-lazy pending\" data-component=\"Lazy\" data-reactid=\"332\">\n </div>\n </div>\n </a>\n <a class=\"view-complete-item\" data-reactid=\"333\" href=\"/recipes/food/views/chinese-egg-noodles-with-smoked-duck-and-snow-peas-354302\" itemprop=\"url\" title=\"Chinese Egg Noodles with Smoked Duck and Snow Peas\">\n \n View “\n \n \n Chinese Egg Noodles with Smoked Duck and Snow Peas\n \n \n ”\n \n </a>\n <div class=\"recipe-panel \" data-reactid=\"337\">\n <a class=\"view-complete-item\" data-reactid=\"338\" href=\"/recipes/food/views/chinese-egg-noodles-with-smoked-duck-and-snow-peas-354302\">\n View Recipe\n </a>\n <div class=\"controls\" data-reactid=\"339\">\n <a class=\"show-quick-view\" data-reactid=\"340\" href=\"/recipes/food/views/chinese-egg-noodles-with-smoked-duck-and-snow-peas-354302\" title=\"Chinese Egg Noodles with Smoked Duck and Snow Peas\">\n Quick view\n </a>\n <a class=\"toggle-compare-state\" data-reactid=\"341\">\n Compare Recipe\n </a>\n </div>\n </div>\n </article>\n <footer class=\"results-group-footer\" data-reactid=\"342\">\n </footer>\n </div>\n <nav class=\"common-pagination\" data-reactid=\"343\" data-total-pages=\"1\" role=\"navigation\">\n <h6 class=\"nav-title\" data-reactid=\"344\">\n Pagination\n </h6>\n <span class=\"the-current-page\" data-reactid=\"345\">\n 1\n </span>\n </nav>\n </section>\n \n \n \n <footer class=\"site-footer\" data-reactid=\"348\" role=\"contentinfo\">\n <section class=\"corporate-info-wrap\" data-reactid=\"349\">\n <span class=\"corporate-info\" data-reactid=\"350\">\n <h3 class=\"section-title\" data-reactid=\"351\">\n Condé Nast\n </h3>\n <nav aria-hidden=\"true\" class=\"conde-nast-brands\" data-reactid=\"352\">\n <h4 class=\"nav-title\" data-reactid=\"353\">\n Condé Nast Websites\n </h4>\n <ul class=\"conde-nast-brands-list\" data-reactid=\"354\">\n <li data-reactid=\"355\">\n <a class=\"conde-nast-brand\" data-reactid=\"356\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.allure.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Allure\">\n Allure\n </a>\n </li>\n <li data-reactid=\"357\">\n <a class=\"conde-nast-brand\" data-reactid=\"358\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.architecturaldigest.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Architectural Digest\">\n Architectural Digest\n </a>\n </li>\n <li data-reactid=\"359\">\n <a class=\"conde-nast-brand\" data-reactid=\"360\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.arstechnica.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Ars Technica\">\n Ars Technica\n </a>\n </li>\n <li data-reactid=\"361\">\n <a class=\"conde-nast-brand\" data-reactid=\"362\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.bonappetit.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Bon Appétit\">\n Bon Appétit\n </a>\n </li>\n <li data-reactid=\"363\">\n <a class=\"conde-nast-brand\" data-reactid=\"364\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.brides.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Brides.com\">\n Brides.com\n </a>\n </li>\n <li data-reactid=\"365\">\n <a class=\"conde-nast-brand\" data-reactid=\"366\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.concierge.com/cntraveler\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Condé Nast Traveler\">\n Condé Nast Traveler\n </a>\n </li>\n <li data-reactid=\"367\">\n <a class=\"conde-nast-brand\" data-reactid=\"368\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.concierge.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Concierge\">\n Concierge\n </a>\n </li>\n <li data-reactid=\"369\">\n <a class=\"conde-nast-brand\" data-reactid=\"370\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.details.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Details\">\n Details\n </a>\n </li>\n <li data-reactid=\"371\">\n <a class=\"conde-nast-brand\" data-reactid=\"372\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.glamour.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Glamour\">\n Glamour\n </a>\n </li>\n <li data-reactid=\"373\">\n <a class=\"conde-nast-brand\" data-reactid=\"374\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.golfdigest.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Golf Digest\">\n Golf Digest\n </a>\n </li>\n <li data-reactid=\"375\">\n <a class=\"conde-nast-brand\" data-reactid=\"376\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.golfworld.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Golf World\">\n Golf World\n </a>\n </li>\n <li data-reactid=\"377\">\n <a class=\"conde-nast-brand\" data-reactid=\"378\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.gq.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"GQ\">\n GQ\n </a>\n </li>\n <li data-reactid=\"379\">\n <a class=\"conde-nast-brand\" data-reactid=\"380\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.hotelchatter.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Hotel Chatter\">\n Hotel Chatter\n </a>\n </li>\n <li data-reactid=\"381\">\n <a class=\"conde-nast-brand\" data-reactid=\"382\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.jaunted.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Jaunted\">\n Jaunted\n </a>\n </li>\n <li data-reactid=\"383\">\n <a class=\"conde-nast-brand\" data-reactid=\"384\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.luckymag.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Lucky\">\n Lucky\n </a>\n </li>\n <li data-reactid=\"385\">\n <a class=\"conde-nast-brand\" data-reactid=\"386\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.nutritiondata.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"NutritionData\">\n NutritionData\n </a>\n </li>\n <li data-reactid=\"387\">\n <a class=\"conde-nast-brand\" data-reactid=\"388\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.reddit.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Reddit\">\n Reddit\n </a>\n </li>\n <li data-reactid=\"389\">\n <a class=\"conde-nast-brand\" data-reactid=\"390\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.self.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Self\">\n Self\n </a>\n </li>\n <li data-reactid=\"391\">\n <a class=\"conde-nast-brand\" data-reactid=\"392\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.style.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Style\">\n Style\n </a>\n </li>\n <li data-reactid=\"393\">\n <a class=\"conde-nast-brand\" data-reactid=\"394\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.teenvogue.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Teen Vogue\">\n Teen Vogue\n </a>\n </li>\n <li data-reactid=\"395\">\n <a class=\"conde-nast-brand\" data-reactid=\"396\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.newyorker.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"The New Yorker\">\n The New Yorker\n </a>\n </li>\n <li data-reactid=\"397\">\n <a class=\"conde-nast-brand\" data-reactid=\"398\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://thescene.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"The Scene\">\n The Scene\n </a>\n </li>\n <li data-reactid=\"399\">\n <a class=\"conde-nast-brand\" data-reactid=\"400\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.vanityfair.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Vanity Fair\">\n Vanity Fair\n </a>\n </li>\n <li data-reactid=\"401\">\n <a class=\"conde-nast-brand\" data-reactid=\"402\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.vegaschatter.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Vegas Chatter\">\n Vegas Chatter\n </a>\n </li>\n <li data-reactid=\"403\">\n <a class=\"conde-nast-brand\" data-reactid=\"404\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.vogue.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Vogue\">\n Vogue\n </a>\n </li>\n <li data-reactid=\"405\">\n <a class=\"conde-nast-brand\" data-reactid=\"406\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.wmagazine.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"W\">\n W\n </a>\n </li>\n <li data-reactid=\"407\">\n <a class=\"conde-nast-brand\" data-reactid=\"408\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.wired.com\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Wired\">\n Wired\n </a>\n </li>\n </ul>\n </nav>\n <nav class=\"conde-nast-services\" data-reactid=\"409\">\n <h4 class=\"nav-title\" data-reactid=\"410\">\n Condé Nast Services\n </h4>\n <ul data-reactid=\"411\">\n <li data-reactid=\"412\">\n <a class=\"conde-nast-service\" data-reactid=\"413\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"/services/subscriptions\" title=\"Subscription Services\">\n Subscription Services\n </a>\n </li>\n <li data-reactid=\"414\">\n <a class=\"conde-nast-service\" data-reactid=\"415\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.condenast.com/careers\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Condé Nast Careers\">\n Careers\n </a>\n </li>\n <li data-reactid=\"416\">\n <a class=\"conde-nast-service\" data-reactid=\"417\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.condenaststore.com/\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Condé Nast Store\">\n Condé Nast Store\n </a>\n </li>\n <li data-reactid=\"418\">\n <a class=\"conde-nast-service\" data-reactid=\"419\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.condenast.com/reprints-permissions\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Reprints/Permissions\">\n Reprints/Permissions\n </a>\n </li>\n </ul>\n </nav>\n <article class=\"legal-notice\" data-reactid=\"420\">\n <h4 class=\"title\" data-reactid=\"421\">\n Legal Notice\n </h4>\n <p data-reactid=\"422\">\n \n ©\n \n \n 2018\n \n \n Condé Nast. All rights reserved\n \n </p>\n <p data-reactid=\"426\">\n \n Use of this site constitutes acceptance of our\n \n <a data-reactid=\"428\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.condenast.com/privacy-policy\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"User Agreement\">\n User Agreement\n </a>\n \n (effective 1/2/2014) and\n \n <a data-reactid=\"430\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.condenast.com/privacy-policy#privacypolicy\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Privacy Policy\">\n Privacy Policy\n </a>\n \n (effective 1/2/2014)\n \n </p>\n <p data-reactid=\"432\">\n <a data-reactid=\"433\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.condenast.com/privacy-policy#privacypolicy-california\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Your California Privacy Rights\">\n Your California Privacy Rights\n </a>\n </p>\n <p data-reactid=\"434\">\n The material on this site may not be reproduced, distributed, transmitted, cached or otherwise used, except with the prior written permission of Condé Nast.\n </p>\n <a class=\"ad-choice\" data-reactid=\"435\" data-track-location=\"footer\" data-track-source=\"navigation\" href=\"http://www.condenast.com/privacy-policy#privacypolicy-optout\" rel=\"noopener noreferrer\" target=\"_blank\" title=\"Ad Choices\">\n Ad Choices\n </a>\n </article>\n </span>\n </section>\n </footer>\n \n \n </span>\n </span>\n <div aria-hidden=\"true\" aria-labelledby=\"cnid-modalLabel\" class=\"modal cnid-modal hide fade\" id=\"cnid-modal\" role=\"dialog\" tabindex=\"-1\">\n <div class=\"frameContainer modal-body cnid-modal-body cnid-ios-scroll\">\n <button aria-hidden=\"true\" class=\"close cnid-close\" data-dismiss=\"modal\" type=\"button\">\n x\n </button>\n <iframe class=\"cnid-signin-iframe\" frameborder=\"0\" height=\"100%\" id=\"cnidClient\" marginheight=\"0\" marginwidth=\"0\" src=\"\" width=\"100%\">\n </iframe>\n </div>\n \n <div class=\"loader-container hide\">\n <div class=\"cnid-overlay-bg\">\n </div>\n <div class=\"cnid-loader\">\n <img src=\"https://cnid.condenastdigital.com/client/assets/img/spinner.gif\"/>\n </div>\n </div>\n \n \n <div class=\"cnid-alerts-container hide\">\n <div class=\"cnid-overlay-bg\">\n </div>\n <div class=\"cnid-alert-message alert text-center alert-success\">\n <p class=\"alert-txt\">\n </p>\n <button class=\"btn-conde okay-btn cnid-btn\" type=\"button\">\n Okay\n </button>\n </div>\n </div>\n </div>\n <div class=\"modal-backdrop\">\n </div>\n <script>\n var _app = {\"cookieDomain\":\".epicurious.com\",\"endpoints\":{\"contextual\":\"/api/dataintelligence/v1/content/recommended\",\"rotd\":\"/api/content/v1/rotd/count/1\"},\"env\":\"PROD\",\"facebook\":{\"appId\":\"1636080783276430\"},\"https\":\"true\",\"media\":{\"bp\":{\"xs\":0,\"s\":600,\"m\":768,\"l\":1024,\"xl\":1360,\"w\":1710}},\"newsletters\":{\"current\":[{\"description\":\"Become a better cook instantly with this weekly report of our ten most helpful tips, tricks, and kitchen secrets. Don't miss it!\",\"id\":\"5\",\"name\":\"The Top Ten\"},{\"description\":\"Love recipes, but hate searching? We do the work for you. You'll get our favorite seasonal recipe plus collections of our exclusive editors' picks.\",\"id\":\"5117\",\"name\":\"Cook This Now\"},{\"description\":\"Get a daily dose of the hottest recipes from Epicurious, Bon Appétit, and other great sites.\",\"id\":\"195169\",\"name\":\"Trending Recipes\"},{\"id\":\"248781\",\"name\":\"Announcements\"},{\"id\":\"248789\",\"name\":\"Cook 90\"},{\"id\":\"248818\",\"name\":\"Diabetes Friendly\"},{\"id\":\"248842\",\"name\":\"Small Plates\"}]},\"server\":\"https://www.epicurious.com\",\"services\":{\"endpoints\":{\"articles\":{\"latest\":\"/api/search/v1/query?size=20&content=article&sort=newest&q=\"},\"branded\":{\"article\":\"/api/branded/v1/article/:brandCode/:slug\"},\"content\":{\"channel\":\"/api/content/v1/channel/\",\"component\":\"/api/content/v1/component\",\"contextual\":\"/api/dataintelligence/v1/content/recommended\",\"featuredIn\":\"/api/content/v1/featuredin/\",\"homepage\":\"/api/content/v1/homepage\",\"latestRecipes\":\"/api/content/v1/component/latestrecipes\",\"memberLookup\":\"/api/users/v2/user/\",\"navItems\":\"/api/content/v1/component/navitems\",\"nutritionalInfo\":\"/api/nutritiondata/v2/\",\"reviews\":\"/api/recipes/v2/\",\"reviewsFeed\":\"/api/recipes/v3/\",\"reviewsSuffix\":\"/reviews\",\"reviewsFeedSuffix\":\"/reviews/feed\",\"rotd\":\"/api/content/v1/rotd/count/1\",\"search\":\"/api/content/v1/search\",\"videoSearchByKeywords\":\"/api/video/v1/keywords/\"},\"contributors\":{\"details\":\"/api/contributors/v1/:contributor/bundle\",\"feed\":\"/api/contributors/v1/:contributor/feed\"},\"cookbooks\":{\"details\":\"/api/cookbooks/v1/:slug\",\"index\":\"/api/cookbooks/v1\"},\"menus\":{\"details\":\"/api/users/v1/:userId/menus/:menuId\",\"index\":\"/api/users/v1/:userId/menus\"},\"photos\":{\"instagramPosts\":\"/api/instagram/v1\",\"menuCover\":\"/api/photos/v1/menu-cover\",\"searchCard\":\"/api/photos/v1/search-card\",\"searchUtility\":\"/api/photos/v1/search-utility\"},\"recipeBox\":{\"search\":\"/api/recipe-box/v1/:userId/search/:term\"},\"recipes\":{\"byKeywords\":\"/api/recipes/v2/keywords/\",\"byKeywordsForMenus\":\"/api/menus/v1/keywords/\",\"latest\":\"/api/recipes/v2/latest\"},\"search\":{\"details\":\"/api/recipes/v3/detail\",\"facets\":\"/api/search/v1/facets\",\"query\":\"/api/search/v1/query\",\"suggest\":\"/api/search/v1/suggest\"},\"users\":{\"authenticate\":\"/api/users/v2/authenticate\",\"info\":\"/api/users/v2/user/:uuid\",\"newsletterSubscriptions\":\"/api/newsletters/v1/subscription/:email\",\"recipeBoxSearch\":\"/api/users/v1/:userId/recipe-box/search/:query\",\"recipes\":\"/api/users/v1/:userId/recipes\",\"reviews\":\"/api/users/v1/:userId/reviews\",\"updateEmail\":\"/api/users/v2/user/email\",\"updatePassword\":\"/api/users/v2/user/password\",\"updateUsername\":\"/api/users/v2/user/displayname\",\"updateSettings\":\"/api/users/v2/user\"}}},\"servicesHost\":\"https://services.epicurious.com\",\"simpleReach\":{\"pid\":\"570d6b7d736b79bf1d000d27\"},\"user\":{\"cookies\":{\"keys\":{\"id\":\"amg_user_partner\",\"username\":\"amg_user_info\"},\"names\":[\"amg_user_ext\",\"amg_user_info\",\"amg_user_partner\",\"amg_user_update\",\"cn_uid\"]},\"email\":{\"regExp\":{}},\"password\":{\"regExp\":{},\"messages\":{\"requirements\":\"A password must be at least 6 characters long. It cannot begin or end with a space.\"}},\"serviceKey\":\"NtibqP3y1qSJM/Gsy3blJgNWt/o=\",\"serviceHost\":\"https://user-service.condenastdigital.com\"},\"userServiceHost\":\"https://user-service.condenastdigital.com\",\"userServiceKey\":\"NtibqP3y1qSJM/Gsy3blJgNWt/o=\",\"vulcan\":{\"host\":\"https://assets.epicurious.com\",\"path\":\"/photos/\"}};\n\n \n (function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){\n (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),\n m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m)\n })(window,document,'script','https://www.google-analytics.com/analytics.js','ga');\n \n ga('create', 'UA-2624365-1', 'auto');\n ga('send', 'pageview');\n  \n function insertOptimizely() {\n var projectId = 2716630229;\n var protocol = ('https:' == document.location.protocol ?\n 'https://' : 'http://');\n var scriptTag = document.createElement('script');\n scriptTag.type = 'text/javascript';\n scriptTag.async = true;\n scriptTag.src = protocol + 'cdn.optimizely.com/js/' +\n projectId + '.js';\n var s = document.getElementsByTagName('script')[0];\n s.parentNode.insertBefore(scriptTag, s);\n };\n \n function optimizelyTimeout() {\n window.optimizely = window.optimizely|| [];\n if (!window.optimizely.data) {\n window.optimizely.push(\"timeout\");\n }\n }\n \n if (false) {\n insertOptimizely();\n setTimeout(optimizelyTimeout, 1000);\n }\n </script>\n <script type=\"text/javascript\">\n window._state = \"%7B%22context%22%3A%7B%22dispatcher%22%3A%7B%22stores%22%3A%7B%22AdStore%22%3A%7B%22displayLeaderboard%22%3Atrue%2C%22pageCreated%22%3Afalse%2C%22pageSlots%22%3A%7B%7D%2C%22env%22%3A%22production%22%2C%22pageConfig%22%3A%7B%22channel%22%3A%22search%22%2C%22slug%22%3A%22search%22%2C%22sub_channel%22%3A%22%22%2C%22targeting%22%3A%7B%22tags%22%3A%5B%22search%22%2C%22%252520tofu%252520chili%22%2C%22food%22%2C%22content%22%5D%7D%2C%22templateType%22%3A%22search%22%7D%7D%2C%22ApplicationStore%22%3A%7B%22analyticsLocation%22%3A%22section1%22%2C%22analyticsSource%22%3A%22section1%22%2C%22config%22%3A%7B%22cookieDomain%22%3A%22.epicurious.com%22%2C%22endpoints%22%3A%7B%22contextual%22%3A%22%2Fapi%2Fdataintelligence%2Fv1%2Fcontent%2Frecommended%22%2C%22rotd%22%3A%22%2Fapi%2Fcontent%2Fv1%2Frotd%2Fcount%2F1%22%7D%2C%22env%22%3A%22PROD%22%2C%22facebook%22%3A%7B%22appId%22%3A%221636080783276430%22%7D%2C%22https%22%3A%22true%22%2C%22media%22%3A%7B%22bp%22%3A%7B%22xs%22%3A0%2C%22s%22%3A600%2C%22m%22%3A768%2C%22l%22%3A1024%2C%22xl%22%3A1360%2C%22w%22%3A1710%7D%7D%2C%22newsletters%22%3A%7B%22current%22%3A%5B%7B%22description%22%3A%22Become%20a%20better%20cook%20instantly%20with%20this%20weekly%20report%20of%20our%20ten%20most%20helpful%20tips%2C%20tricks%2C%20and%20kitchen%20secrets.%20Don't%20miss%20it!%22%2C%22id%22%3A%225%22%2C%22name%22%3A%22The%20Top%20Ten%22%7D%2C%7B%22description%22%3A%22Love%20recipes%2C%20but%20hate%20searching%3F%20We%20do%20the%20work%20for%20you.%20You'll%20get%20our%20favorite%20seasonal%20recipe%20plus%20collections%20of%20our%20exclusive%20editors'%20picks.%22%2C%22id%22%3A%225117%22%2C%22name%22%3A%22Cook%20This%20Now%22%7D%2C%7B%22description%22%3A%22Get%20a%20daily%20dose%20of%20the%20hottest%20recipes%20from%20Epicurious%2C%20Bon%20App%C3%A9tit%2C%20and%20other%20great%20sites.%22%2C%22id%22%3A%22195169%22%2C%22name%22%3A%22Trending%20Recipes%22%7D%2C%7B%22id%22%3A%22248781%22%2C%22name%22%3A%22Announcements%22%7D%2C%7B%22id%22%3A%22248789%22%2C%22name%22%3A%22Cook%2090%22%7D%2C%7B%22id%22%3A%22248818%22%2C%22name%22%3A%22Diabetes%20Friendly%22%7D%2C%7B%22id%22%3A%22248842%22%2C%22name%22%3A%22Small%20Plates%22%7D%5D%7D%2C%22server%22%3A%22https%3A%2F%2Fwww.epicurious.com%22%2C%22services%22%3A%7B%22endpoints%22%3A%7B%22articles%22%3A%7B%22latest%22%3A%22%2Fapi%2Fsearch%2Fv1%2Fquery%3Fsize%3D20%26content%3Darticle%26sort%3Dnewest%26q%3D%22%7D%2C%22branded%22%3A%7B%22article%22%3A%22%2Fapi%2Fbranded%2Fv1%2Farticle%2F%3AbrandCode%2F%3Aslug%22%7D%2C%22content%22%3A%7B%22channel%22%3A%22%2Fapi%2Fcontent%2Fv1%2Fchannel%2F%22%2C%22component%22%3A%22%2Fapi%2Fcontent%2Fv1%2Fcomponent%22%2C%22contextual%22%3A%22%2Fapi%2Fdataintelligence%2Fv1%2Fcontent%2Frecommended%22%2C%22featuredIn%22%3A%22%2Fapi%2Fcontent%2Fv1%2Ffeaturedin%2F%22%2C%22homepage%22%3A%22%2Fapi%2Fcontent%2Fv1%2Fhomepage%22%2C%22latestRecipes%22%3A%22%2Fapi%2Fcontent%2Fv1%2Fcomponent%2Flatestrecipes%22%2C%22memberLookup%22%3A%22%2Fapi%2Fusers%2Fv2%2Fuser%2F%22%2C%22navItems%22%3A%22%2Fapi%2Fcontent%2Fv1%2Fcomponent%2Fnavitems%22%2C%22nutritionalInfo%22%3A%22%2Fapi%2Fnutritiondata%2Fv2%2F%22%2C%22reviews%22%3A%22%2Fapi%2Frecipes%2Fv2%2F%22%2C%22reviewsFeed%22%3A%22%2Fapi%2Frecipes%2Fv3%2F%22%2C%22reviewsSuffix%22%3A%22%2Freviews%22%2C%22reviewsFeedSuffix%22%3A%22%2Freviews%2Ffeed%22%2C%22rotd%22%3A%22%2Fapi%2Fcontent%2Fv1%2Frotd%2Fcount%2F1%22%2C%22search%22%3A%22%2Fapi%2Fcontent%2Fv1%2Fsearch%22%2C%22videoSearchByKeywords%22%3A%22%2Fapi%2Fvideo%2Fv1%2Fkeywords%2F%22%7D%2C%22contributors%22%3A%7B%22details%22%3A%22%2Fapi%2Fcontributors%2Fv1%2F%3Acontributor%2Fbundle%22%2C%22feed%22%3A%22%2Fapi%2Fcontributors%2Fv1%2F%3Acontributor%2Ffeed%22%7D%2C%22cookbooks%22%3A%7B%22details%22%3A%22%2Fapi%2Fcookbooks%2Fv1%2F%3Aslug%22%2C%22index%22%3A%22%2Fapi%2Fcookbooks%2Fv1%22%7D%2C%22menus%22%3A%7B%22details%22%3A%22%2Fapi%2Fusers%2Fv1%2F%3AuserId%2Fmenus%2F%3AmenuId%22%2C%22index%22%3A%22%2Fapi%2Fusers%2Fv1%2F%3AuserId%2Fmenus%22%7D%2C%22photos%22%3A%7B%22instagramPosts%22%3A%22%2Fapi%2Finstagram%2Fv1%22%2C%22menuCover%22%3A%22%2Fapi%2Fphotos%2Fv1%2Fmenu-cover%22%2C%22searchCard%22%3A%22%2Fapi%2Fphotos%2Fv1%2Fsearch-card%22%2C%22searchUtility%22%3A%22%2Fapi%2Fphotos%2Fv1%2Fsearch-utility%22%7D%2C%22recipeBox%22%3A%7B%22search%22%3A%22%2Fapi%2Frecipe-box%2Fv1%2F%3AuserId%2Fsearch%2F%3Aterm%22%7D%2C%22recipes%22%3A%7B%22byKeywords%22%3A%22%2Fapi%2Frecipes%2Fv2%2Fkeywords%2F%22%2C%22byKeywordsForMenus%22%3A%22%2Fapi%2Fmenus%2Fv1%2Fkeywords%2F%22%2C%22latest%22%3A%22%2Fapi%2Frecipes%2Fv2%2Flatest%22%7D%2C%22search%22%3A%7B%22details%22%3A%22%2Fapi%2Frecipes%2Fv3%2Fdetail%22%2C%22facets%22%3A%22%2Fapi%2Fsearch%2Fv1%2Ffacets%22%2C%22query%22%3A%22%2Fapi%2Fsearch%2Fv1%2Fquery%22%2C%22suggest%22%3A%22%2Fapi%2Fsearch%2Fv1%2Fsuggest%22%7D%2C%22users%22%3A%7B%22authenticate%22%3A%22%2Fapi%2Fusers%2Fv2%2Fauthenticate%22%2C%22info%22%3A%22%2Fapi%2Fusers%2Fv2%2Fuser%2F%3Auuid%22%2C%22newsletterSubscriptions%22%3A%22%2Fapi%2Fnewsletters%2Fv1%2Fsubscription%2F%3Aemail%22%2C%22recipeBoxSearch%22%3A%22%2Fapi%2Fusers%2Fv1%2F%3AuserId%2Frecipe-box%2Fsearch%2F%3Aquery%22%2C%22recipes%22%3A%22%2Fapi%2Fusers%2Fv1%2F%3AuserId%2Frecipes%22%2C%22reviews%22%3A%22%2Fapi%2Fusers%2Fv1%2F%3AuserId%2Freviews%22%2C%22updateEmail%22%3A%22%2Fapi%2Fusers%2Fv2%2Fuser%2Femail%22%2C%22updatePassword%22%3A%22%2Fapi%2Fusers%2Fv2%2Fuser%2Fpassword%22%2C%22updateUsername%22%3A%22%2Fapi%2Fusers%2Fv2%2Fuser%2Fdisplayname%22%2C%22updateSettings%22%3A%22%2Fapi%2Fusers%2Fv2%2Fuser%22%7D%7D%7D%2C%22servicesHost%22%3A%22https%3A%2F%2Fservices.epicurious.com%22%2C%22simpleReach%22%3A%7B%22pid%22%3A%22570d6b7d736b79bf1d000d27%22%7D%2C%22user%22%3A%7B%22cookies%22%3A%7B%22keys%22%3A%7B%22id%22%3A%22amg_user_partner%22%2C%22username%22%3A%22amg_user_info%22%7D%2C%22names%22%3A%5B%22amg_user_ext%22%2C%22amg_user_info%22%2C%22amg_user_partner%22%2C%22amg_user_update%22%2C%22cn_uid%22%5D%7D%2C%22email%22%3A%7B%22regExp%22%3A%7B%7D%7D%2C%22password%22%3A%7B%22regExp%22%3A%7B%7D%2C%22messages%22%3A%7B%22requirements%22%3A%22A%20password%20must%20be%20at%20least%206%20characters%20long.%20It%20cannot%20begin%20or%20end%20with%20a%20space.%22%7D%7D%2C%22serviceKey%22%3A%22NtibqP3y1qSJM%2FGsy3blJgNWt%2Fo%3D%22%2C%22serviceHost%22%3A%22https%3A%2F%2Fuser-service.condenastdigital.com%22%7D%2C%22userServiceHost%22%3A%22https%3A%2F%2Fuser-service.condenastdigital.com%22%2C%22userServiceKey%22%3A%22NtibqP3y1qSJM%2FGsy3blJgNWt%2Fo%3D%22%2C%22vulcan%22%3A%7B%22host%22%3A%22https%3A%2F%2Fassets.epicurious.com%22%2C%22path%22%3A%22%2Fphotos%2F%22%7D%7D%2C%22pageName%22%3A%22search%22%2C%22siteFooter%22%3A%7B%22hasEpicuriousLinks%22%3Afalse%7D%7D%2C%22SearchStore%22%3A%7B%22activeFilters%22%3A%5B%5D%2C%22stagedFilters%22%3A%5B%5D%2C%22activeFilterGroup%22%3A%22%22%2C%22includeIngredients%22%3A%5B%5D%2C%22excludeIngredients%22%3A%5B%5D%2C%22stagedIncludeIngredients%22%3A%5B%5D%2C%22stagedExcludeIngredients%22%3A%5B%5D%2C%22api%22%3A%7B%22detailsUri%22%3A%22https%3A%2F%2Fservices.epicurious.com%2Fapi%2Frecipes%2Fv3%2Fdetail%22%2C%22facetUri%22%3A%22https%3A%2F%2Fservices.epicurious.com%2Fapi%2Fsearch%2Fv1%2Ffacets%22%2C%22searchUri%22%3A%22https%3A%2F%2Fservices.epicurious.com%2Fapi%2Fsearch%2Fv1%2Fquery%22%2C%22suggestUri%22%3A%22https%3A%2F%2Fservices.epicurious.com%2Fapi%2Fsearch%2Fv1%2Fsuggest%22%7D%2C%22compareRecipes%22%3A%5B%5D%2C%22numFound%22%3A18%2C%22moreResults%22%3Afalse%2C%22newSearch%22%3Afalse%2C%22page%22%3A%7B%22generatedAt%22%3A%222018-02-26T12%3A43%3A15.272Z%22%2C%22count%22%3A1%2C%22number%22%3A1%2C%22size%22%3A18%2C%22totalCount%22%3A18%2C%22previousNumber%22%3A1%7D%2C%22exclusiveFilters%22%3A%7B%7D%2C%22facets%22%3A%7B%22technique%22%3A%7B%22grill-barbecue%22%3A%7B%22id%22%3A%22grill-barbecue%22%2C%22name%22%3A%22Barbecue%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22bake%22%3A%7B%22id%22%3A%22bake%22%2C%22name%22%3A%22Bake%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22roast%22%3A%7B%22id%22%3A%22roast%22%2C%22name%22%3A%22Roast%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22braise%22%3A%7B%22id%22%3A%22braise%22%2C%22name%22%3A%22Braise%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22saute%22%3A%7B%22id%22%3A%22saute%22%2C%22name%22%3A%22Saut%C3%A9%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22no-cook%22%3A%7B%22id%22%3A%22no-cook%22%2C%22name%22%3A%22No-Cook%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22stir-fry%22%3A%7B%22id%22%3A%22stir-fry%22%2C%22name%22%3A%22Stir-Fry%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22pan-fry%22%3A%7B%22id%22%3A%22pan-fry%22%2C%22name%22%3A%22Pan-Fry%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22marinate%22%3A%7B%22id%22%3A%22marinate%22%2C%22name%22%3A%22Marinate%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22broil%22%3A%7B%22id%22%3A%22broil%22%2C%22name%22%3A%22Broil%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22stew%22%3A%7B%22id%22%3A%22stew%22%2C%22name%22%3A%22Stew%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22fry%22%3A%7B%22id%22%3A%22fry%22%2C%22name%22%3A%22Fry%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22chill%22%3A%7B%22id%22%3A%22chill%22%2C%22name%22%3A%22Chill%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22steam%22%3A%7B%22id%22%3A%22steam%22%2C%22name%22%3A%22Steam%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22simmer%22%3A%7B%22id%22%3A%22simmer%22%2C%22name%22%3A%22Simmer%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22poach%22%3A%7B%22id%22%3A%22poach%22%2C%22name%22%3A%22Poach%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22boil%22%3A%7B%22id%22%3A%22boil%22%2C%22name%22%3A%22Boil%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22brine%22%3A%7B%22id%22%3A%22brine%22%2C%22name%22%3A%22Brine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22freeze-chill%22%3A%7B%22id%22%3A%22freeze-chill%22%2C%22name%22%3A%22Freeze%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22deep-fry%22%3A%7B%22id%22%3A%22deep-fry%22%2C%22name%22%3A%22Deep%20Fry%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22advance-prep-required%22%3A%7B%22id%22%3A%22advance-prep-required%22%2C%22name%22%3A%22Advance%20Prep%20Req'd%22%2C%22category%22%3A%22technique%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%7D%2C%22meal%22%3A%7B%22dinner%22%3A%7B%22id%22%3A%22dinner%22%2C%22name%22%3A%22Dinner%22%2C%22category%22%3A%22meal%22%2C%22orderCount%22%3A0%2C%22count%22%3A5%7D%2C%22dessert%22%3A%7B%22id%22%3A%22dessert%22%2C%22name%22%3A%22Dessert%22%2C%22category%22%3A%22meal%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22appetizer%22%3A%7B%22id%22%3A%22appetizer%22%2C%22name%22%3A%22Appetizer%22%2C%22category%22%3A%22meal%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22side%22%3A%7B%22id%22%3A%22side%22%2C%22name%22%3A%22Side%22%2C%22category%22%3A%22meal%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22breakfast%22%3A%7B%22id%22%3A%22breakfast%22%2C%22name%22%3A%22Breakfast%22%2C%22category%22%3A%22meal%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22lunch%22%3A%7B%22id%22%3A%22lunch%22%2C%22name%22%3A%22Lunch%22%2C%22category%22%3A%22meal%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22brunch%22%3A%7B%22id%22%3A%22brunch%22%2C%22name%22%3A%22Brunch%22%2C%22category%22%3A%22meal%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22buffet%22%3A%7B%22id%22%3A%22buffet%22%2C%22name%22%3A%22Buffet%22%2C%22category%22%3A%22meal%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%7D%2C%22type%22%3A%7B%22soup-stew%22%3A%7B%22id%22%3A%22soup-stew%22%2C%22name%22%3A%22Soup%2FStew%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22salad%22%3A%7B%22id%22%3A%22salad%22%2C%22name%22%3A%22Salad%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22cookie%22%3A%7B%22id%22%3A%22cookie%22%2C%22name%22%3A%22Cookie%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22stuffing-dressing%22%3A%7B%22id%22%3A%22stuffing-dressing%22%2C%22name%22%3A%22Stuffing%2FDressing%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22cake%22%3A%7B%22id%22%3A%22cake%22%2C%22name%22%3A%22Cake%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22bread%22%3A%7B%22id%22%3A%22bread%22%2C%22name%22%3A%22Bread%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22edible-gift%22%3A%7B%22id%22%3A%22edible-gift%22%2C%22name%22%3A%22Edible%20Gift%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22sandwich%22%3A%7B%22id%22%3A%22sandwich%22%2C%22name%22%3A%22Sandwich%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22sauce%22%3A%7B%22id%22%3A%22sauce%22%2C%22name%22%3A%22Sauce%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22cocktail%22%3A%7B%22id%22%3A%22cocktail%22%2C%22name%22%3A%22Cocktail%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22candy%22%3A%7B%22id%22%3A%22candy%22%2C%22name%22%3A%22Candy%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22condiment-spread%22%3A%7B%22id%22%3A%22condiment-spread%22%2C%22name%22%3A%22Condiment%2FSpread%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22casserole-gratin%22%3A%7B%22id%22%3A%22casserole-gratin%22%2C%22name%22%3A%22Casserole%2FGratin%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22frozen-dessert%22%3A%7B%22id%22%3A%22frozen-dessert%22%2C%22name%22%3A%22Frozen%20Dessert%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22salad-dressing%22%3A%7B%22id%22%3A%22salad-dressing%22%2C%22name%22%3A%22Salad%20Dressing%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22pastry%22%3A%7B%22id%22%3A%22pastry%22%2C%22name%22%3A%22Pastry%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22ice-cream%22%3A%7B%22id%22%3A%22ice-cream%22%2C%22name%22%3A%22Ice%20Cream%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22sangria%22%3A%7B%22id%22%3A%22sangria%22%2C%22name%22%3A%22Sangria%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22margarita%22%3A%7B%22id%22%3A%22margarita%22%2C%22name%22%3A%22Margarita%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22martini%22%3A%7B%22id%22%3A%22martini%22%2C%22name%22%3A%22Martini%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22iced-tea%22%3A%7B%22id%22%3A%22iced-tea%22%2C%22name%22%3A%22Iced%20Tea%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22aperitif%22%3A%7B%22id%22%3A%22aperitif%22%2C%22name%22%3A%22Aperitif%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22digestif%22%3A%7B%22id%22%3A%22digestif%22%2C%22name%22%3A%22Digestif%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22biscuit%22%3A%7B%22id%22%3A%22biscuit%22%2C%22name%22%3A%22Biscuit%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22brownie%22%3A%7B%22id%22%3A%22brownie%22%2C%22name%22%3A%22Brownie%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22burrito%22%3A%7B%22id%22%3A%22burrito%22%2C%22name%22%3A%22Burrito%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22cheesecake%22%3A%7B%22id%22%3A%22cheesecake%22%2C%22name%22%3A%22Cheesecake%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22chili%22%3A%7B%22id%22%3A%22chili%22%2C%22name%22%3A%22Chili%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22chowder%22%3A%7B%22id%22%3A%22chowder%22%2C%22name%22%3A%22Chowder%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22cobbler-crumble%22%3A%7B%22id%22%3A%22cobbler-crumble%22%2C%22name%22%3A%22Cobbler%2FCrumble%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22crepe%22%3A%7B%22id%22%3A%22crepe%22%2C%22name%22%3A%22Cr%C3%AApe%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22cranberry-sauce%22%3A%7B%22id%22%3A%22cranberry-sauce%22%2C%22name%22%3A%22Cranberry%20Sauce%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22cupcake%22%3A%7B%22id%22%3A%22cupcake%22%2C%22name%22%3A%22Cupcake%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22custard%22%3A%7B%22id%22%3A%22custard%22%2C%22name%22%3A%22Custard%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22dip%22%3A%7B%22id%22%3A%22dip%22%2C%22name%22%3A%22Dip%22%2C%22category%22%3A%22type%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22flat-bread%22%3A%7B%22id%22%3A%22flat-bread%22%2C%22name%22%3A%22Flat%20Bread%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22frittata%22%3A%7B%22id%22%3A%22frittata%22%2C%22name%22%3A%22Frittata%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22fritter%22%3A%7B%22id%22%3A%22fritter%22%2C%22name%22%3A%22Fritter%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22guacamole%22%3A%7B%22id%22%3A%22guacamole%22%2C%22name%22%3A%22Guacamole%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22hamburger%22%3A%7B%22id%22%3A%22hamburger%22%2C%22name%22%3A%22Hamburger%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%7D%2C%22special-consideration%22%3A%7B%22vegetarian%22%3A%7B%22id%22%3A%22vegetarian%22%2C%22name%22%3A%22Vegetarian%22%2C%22category%22%3A%22special-consideration%22%2C%22orderCount%22%3A0%2C%22count%22%3A4%7D%2C%22quick-and-easy%22%3A%7B%22id%22%3A%22quick-and-easy%22%2C%22name%22%3A%22Quick%20%26%20Easy%22%2C%22category%22%3A%22special-consideration%22%2C%22orderCount%22%3A0%2C%22count%22%3A3%7D%2C%22kid-friendly%22%3A%7B%22id%22%3A%22kid-friendly%22%2C%22name%22%3A%22Kid%20Friendly%22%2C%22category%22%3A%22special-consideration%22%2C%22orderCount%22%3A0%2C%22count%22%3A2%7D%2C%22healthy%22%3A%7B%22id%22%3A%22healthy%22%2C%22name%22%3A%22Healthy%22%2C%22category%22%3A%22special-consideration%22%2C%22orderCount%22%3A0%2C%22count%22%3A4%7D%2C%22wheat-gluten-free%22%3A%7B%22id%22%3A%22wheat-gluten-free%22%2C%22name%22%3A%22Wheat%2FGluten-Free%22%2C%22category%22%3A%22special-consideration%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22vegan%22%3A%7B%22id%22%3A%22vegan%22%2C%22name%22%3A%22Vegan%22%2C%22category%22%3A%22special-consideration%22%2C%22orderCount%22%3A0%2C%22count%22%3A3%7D%2C%22low-fat%22%3A%7B%22id%22%3A%22low-fat%22%2C%22name%22%3A%22Low%20Fat%22%2C%22orderCount%22%3A0%2C%22count%22%3A2%7D%2C%22low-no-sugar%22%3A%7B%22id%22%3A%22low-no-sugar%22%2C%22name%22%3A%22Low%2FNo%20Sugar%22%2C%22category%22%3A%22special-consideration%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22low-sodium%22%3A%7B%22id%22%3A%22low-sodium%22%2C%22name%22%3A%22Low%20Sodium%22%2C%22category%22%3A%22special-consideration%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22high-fiber%22%3A%7B%22id%22%3A%22high-fiber%22%2C%22name%22%3A%22High%20Fiber%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22kosher%22%3A%7B%22id%22%3A%22kosher%22%2C%22name%22%3A%22Kosher%22%2C%22category%22%3A%22special-consideration%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22low-cholesterol%22%3A%7B%22id%22%3A%22low-cholesterol%22%2C%22name%22%3A%22Low%20Cholesterol%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22raw%22%3A%7B%22id%22%3A%22raw%22%2C%22name%22%3A%22Raw%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22kosher-for-passover%22%3A%7B%22id%22%3A%22kosher-for-passover%22%2C%22name%22%3A%22Kosher%20for%20Passover%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22organic%22%3A%7B%22id%22%3A%22organic%22%2C%22name%22%3A%22Organic%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%7D%2C%22ingredient%22%3A%7B%22chicken%22%3A%7B%22id%22%3A%22chicken%22%2C%22name%22%3A%22Chicken%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22beef%22%3A%7B%22id%22%3A%22beef%22%2C%22name%22%3A%22Beef%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22fish%22%3A%7B%22id%22%3A%22fish%22%2C%22name%22%3A%22Fish%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22pork%22%3A%7B%22id%22%3A%22pork%22%2C%22name%22%3A%22Pork%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22vegetable%22%3A%7B%22id%22%3A%22vegetable%22%2C%22name%22%3A%22Vegetable%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22turkey%22%3A%7B%22id%22%3A%22turkey%22%2C%22name%22%3A%22Turkey%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22potato%22%3A%7B%22id%22%3A%22potato%22%2C%22name%22%3A%22Potato%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22pasta%22%3A%7B%22id%22%3A%22pasta%22%2C%22name%22%3A%22Pasta%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22seafood%22%3A%7B%22id%22%3A%22seafood%22%2C%22name%22%3A%22Seafood%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22egg%22%3A%7B%22id%22%3A%22egg%22%2C%22name%22%3A%22Egg%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22lamb%22%3A%7B%22id%22%3A%22lamb%22%2C%22name%22%3A%22Lamb%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22chocolate%22%3A%7B%22id%22%3A%22chocolate%22%2C%22name%22%3A%22Chocolate%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22cranberry%22%3A%7B%22id%22%3A%22cranberry%22%2C%22name%22%3A%22Cranberry%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22salmon%22%3A%7B%22id%22%3A%22salmon%22%2C%22name%22%3A%22Salmon%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22bean%22%3A%7B%22id%22%3A%22bean%22%2C%22name%22%3A%22Bean%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22shellfish%22%3A%7B%22id%22%3A%22shellfish%22%2C%22name%22%3A%22Shellfish%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22fruit%22%3A%7B%22id%22%3A%22fruit%22%2C%22name%22%3A%22Fruit%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22duck%22%3A%7B%22id%22%3A%22duck%22%2C%22name%22%3A%22Duck%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22shrimp%22%3A%7B%22id%22%3A%22shrimp%22%2C%22name%22%3A%22Shrimp%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22mushroom%22%3A%7B%22id%22%3A%22mushroom%22%2C%22name%22%3A%22Mushroom%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22tomato%22%3A%7B%22id%22%3A%22tomato%22%2C%22name%22%3A%22Tomato%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22rice%22%3A%7B%22id%22%3A%22rice%22%2C%22name%22%3A%22Rice%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22poultry%22%3A%7B%22id%22%3A%22poultry%22%2C%22name%22%3A%22Poultry%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A2%7D%2C%22scallop%22%3A%7B%22id%22%3A%22scallop%22%2C%22name%22%3A%22Scallop%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22ham%22%3A%7B%22id%22%3A%22ham%22%2C%22name%22%3A%22Ham%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22carrot%22%3A%7B%22id%22%3A%22carrot%22%2C%22name%22%3A%22Carrot%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22eggplant%22%3A%7B%22id%22%3A%22eggplant%22%2C%22name%22%3A%22Eggplant%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22leafy-green%22%3A%7B%22id%22%3A%22leafy-green%22%2C%22name%22%3A%22Leafy%20Green%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22apple%22%3A%7B%22id%22%3A%22apple%22%2C%22name%22%3A%22Apple%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22green-bean%22%3A%7B%22id%22%3A%22green-bean%22%2C%22name%22%3A%22Green%20Bean%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22broccoli%22%3A%7B%22id%22%3A%22broccoli%22%2C%22name%22%3A%22Broccoli%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22zucchini%22%3A%7B%22id%22%3A%22zucchini%22%2C%22name%22%3A%22Zucchini%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22cabbage%22%3A%7B%22id%22%3A%22cabbage%22%2C%22name%22%3A%22Cabbage%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22ground-beef%22%3A%7B%22id%22%3A%22ground-beef%22%2C%22name%22%3A%22Ground%20Beef%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22lemon%22%3A%7B%22id%22%3A%22lemon%22%2C%22name%22%3A%22Lemon%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22sweet-potato-yam%22%3A%7B%22id%22%3A%22sweet-potato-yam%22%2C%22name%22%3A%22Sweet%20Potato%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22citrus%22%3A%7B%22id%22%3A%22citrus%22%2C%22name%22%3A%22Citrus%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22kale%22%3A%7B%22id%22%3A%22kale%22%2C%22name%22%3A%22Kale%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22spinach%22%3A%7B%22id%22%3A%22spinach%22%2C%22name%22%3A%22Spinach%22%2C%22category%22%3A%22ingredient%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%7D%2C%22cuisine%22%3A%7B%22italian%22%3A%7B%22id%22%3A%22italian%22%2C%22name%22%3A%22Italian%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22mexican%22%3A%7B%22id%22%3A%22mexican%22%2C%22name%22%3A%22Mexican%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22moroccan%22%3A%7B%22id%22%3A%22moroccan%22%2C%22name%22%3A%22Moroccan%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22french%22%3A%7B%22id%22%3A%22french%22%2C%22name%22%3A%22French%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22asian%22%3A%7B%22id%22%3A%22asian%22%2C%22name%22%3A%22Asian%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A4%7D%2C%22indian%22%3A%7B%22id%22%3A%22indian%22%2C%22name%22%3A%22Indian%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A2%7D%2C%22thai%22%3A%7B%22id%22%3A%22thai%22%2C%22name%22%3A%22Thai%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A2%7D%2C%22mediterranean%22%3A%7B%22id%22%3A%22mediterranean%22%2C%22name%22%3A%22Mediterranean%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22middle-eastern%22%3A%7B%22id%22%3A%22middle-eastern%22%2C%22name%22%3A%22Middle%20Eastern%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22spanish-portuguese%22%3A%7B%22id%22%3A%22spanish-portuguese%22%2C%22name%22%3A%22Spanish%2FPortuguese%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22greek%22%3A%7B%22id%22%3A%22greek%22%2C%22name%22%3A%22Greek%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22american%22%3A%7B%22id%22%3A%22american%22%2C%22name%22%3A%22American%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22chinese%22%3A%7B%22id%22%3A%22chinese%22%2C%22name%22%3A%22Chinese%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A3%7D%2C%22cajun-creole%22%3A%7B%22id%22%3A%22cajun-creole%22%2C%22name%22%3A%22Cajun%2FCreole%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22central-south-american%22%3A%7B%22id%22%3A%22central-south-american%22%2C%22name%22%3A%22Central%2FS.%20American%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22southern%22%3A%7B%22id%22%3A%22southern%22%2C%22name%22%3A%22Southern%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22german%22%3A%7B%22id%22%3A%22german%22%2C%22name%22%3A%22German%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22african%22%3A%7B%22id%22%3A%22african%22%2C%22name%22%3A%22African%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22scandinavian%22%3A%7B%22id%22%3A%22scandinavian%22%2C%22name%22%3A%22Scandinavian%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22eastern-european-russian%22%3A%7B%22id%22%3A%22eastern-european-russian%22%2C%22name%22%3A%22Eastern%20European%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22irish%22%3A%7B%22id%22%3A%22irish%22%2C%22name%22%3A%22Irish%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22japanese%22%3A%7B%22id%22%3A%22japanese%22%2C%22name%22%3A%22Japanese%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A2%7D%2C%22southwestern%22%3A%7B%22id%22%3A%22southwestern%22%2C%22name%22%3A%22Southwestern%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22vietnamese%22%3A%7B%22id%22%3A%22vietnamese%22%2C%22name%22%3A%22Vietnamese%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22english%22%3A%7B%22id%22%3A%22english%22%2C%22name%22%3A%22English%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22jewish%22%3A%7B%22id%22%3A%22jewish%22%2C%22name%22%3A%22Jewish%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22central-american-caribbean%22%3A%7B%22id%22%3A%22central-american-caribbean%22%2C%22name%22%3A%22Caribbean%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22latin-american%22%3A%7B%22id%22%3A%22latin-american%22%2C%22name%22%3A%22Latin%20American%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22tex-mex%22%3A%7B%22id%22%3A%22tex-mex%22%2C%22name%22%3A%22Tex-Mex%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22italian-american%22%3A%7B%22id%22%3A%22italian-american%22%2C%22name%22%3A%22Italian%20American%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22turkish%22%3A%7B%22id%22%3A%22turkish%22%2C%22name%22%3A%22Turkish%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22korean%22%3A%7B%22id%22%3A%22korean%22%2C%22name%22%3A%22Korean%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22southeast-asian%22%3A%7B%22id%22%3A%22southeast-asian%22%2C%22name%22%3A%22Southeast%20Asian%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22european%22%3A%7B%22id%22%3A%22european%22%2C%22name%22%3A%22European%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22british%22%3A%7B%22id%22%3A%22british%22%2C%22name%22%3A%22British%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22cuban%22%3A%7B%22id%22%3A%22cuban%22%2C%22name%22%3A%22Cuban%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22south-american%22%3A%7B%22id%22%3A%22south-american%22%2C%22name%22%3A%22South%20American%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22nuevo-latino%22%3A%7B%22id%22%3A%22nuevo-latino%22%2C%22name%22%3A%22Nuevo%20Latino%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22californian%22%3A%7B%22id%22%3A%22californian%22%2C%22name%22%3A%22Californian%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22south-asian%22%3A%7B%22id%22%3A%22south-asian%22%2C%22name%22%3A%22South%20Asian%22%2C%22category%22%3A%22cuisine%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%7D%2C%22occasion%22%3A%7B%22summer%22%3A%7B%22id%22%3A%22summer%22%2C%22name%22%3A%22Summer%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22fourth-of-july%22%3A%7B%22id%22%3A%22fourth-of-july%22%2C%22name%22%3A%22Fourth%20of%20July%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22fall%22%3A%7B%22id%22%3A%22fall%22%2C%22name%22%3A%22Fall%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22thanksgiving%22%3A%7B%22id%22%3A%22thanksgiving%22%2C%22name%22%3A%22Thanksgiving%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22christmas%22%3A%7B%22id%22%3A%22christmas%22%2C%22name%22%3A%22Christmas%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22new-years-eve%22%3A%7B%22id%22%3A%22new-years-eve%22%2C%22name%22%3A%22New%20Year's%20Eve%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22easter%22%3A%7B%22id%22%3A%22easter%22%2C%22name%22%3A%22Easter%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22passover%22%3A%7B%22id%22%3A%22passover%22%2C%22name%22%3A%22Passover%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22backyard-bbq%22%3A%7B%22id%22%3A%22backyard-bbq%22%2C%22name%22%3A%22Backyard%20BBQ%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22picnic%22%3A%7B%22id%22%3A%22picnic%22%2C%22name%22%3A%22Picnic%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22potluck%22%3A%7B%22id%22%3A%22potluck%22%2C%22name%22%3A%22Potluck%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22tailgating%22%3A%7B%22id%22%3A%22tailgating%22%2C%22name%22%3A%22Tailgating%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22birthday%22%3A%7B%22id%22%3A%22birthday%22%2C%22name%22%3A%22Birthday%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22party%22%3A%7B%22id%22%3A%22party%22%2C%22name%22%3A%22Party%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22super-bowl%22%3A%7B%22id%22%3A%22super-bowl%22%2C%22name%22%3A%22Super%20Bowl%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22winter%22%3A%7B%22id%22%3A%22winter%22%2C%22name%22%3A%22Winter%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22cocktail-party%22%3A%7B%22id%22%3A%22cocktail-party%22%2C%22name%22%3A%22Cocktail%20Party%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22christmas-eve%22%3A%7B%22id%22%3A%22christmas-eve%22%2C%22name%22%3A%22Christmas%20Eve%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22spring%22%3A%7B%22id%22%3A%22spring%22%2C%22name%22%3A%22Spring%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A1%7D%2C%22valentines-day%22%3A%7B%22id%22%3A%22valentines-day%22%2C%22name%22%3A%22Valentine's%20Day%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22halloween%22%3A%7B%22id%22%3A%22halloween%22%2C%22name%22%3A%22Halloween%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22cinco-de-mayo%22%3A%7B%22id%22%3A%22cinco-de-mayo%22%2C%22name%22%3A%22Cinco%20de%20Mayo%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22wedding%22%3A%7B%22id%22%3A%22wedding%22%2C%22name%22%3A%22Wedding%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22mothers-day%22%3A%7B%22id%22%3A%22mothers-day%22%2C%22name%22%3A%22Mother's%20Day%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22st-patricks-day%22%3A%7B%22id%22%3A%22st-patricks-day%22%2C%22name%22%3A%22St.%20Patrick's%20Day%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22hanukkah%22%3A%7B%22id%22%3A%22hanukkah%22%2C%22name%22%3A%22Hanukkah%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22shower%22%3A%7B%22id%22%3A%22shower%22%2C%22name%22%3A%22Shower%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22mardi-gras%22%3A%7B%22id%22%3A%22mardi-gras%22%2C%22name%22%3A%22Mardi%20Gras%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22family-reunion%22%3A%7B%22id%22%3A%22family-reunion%22%2C%22name%22%3A%22Family%20Reunion%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22new-years-day%22%3A%7B%22id%22%3A%22new-years-day%22%2C%22name%22%3A%22New%20Year's%20Day%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22kentucky-derby%22%3A%7B%22id%22%3A%22kentucky-derby%22%2C%22name%22%3A%22Kentucky%20Derby%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22engagement-party%22%3A%7B%22id%22%3A%22engagement-party%22%2C%22name%22%3A%22Engagement%20Party%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22poker-game-night%22%3A%7B%22id%22%3A%22poker-game-night%22%2C%22name%22%3A%22Poker%2FGame%20Night%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22oscars%22%3A%7B%22id%22%3A%22oscars%22%2C%22name%22%3A%22Oscars%22%2C%22category%22%3A%22occasion%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22fathers-day%22%3A%7B%22id%22%3A%22fathers-day%22%2C%22name%22%3A%22Father's%20Day%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22anniversary%22%3A%7B%22id%22%3A%22anniversary%22%2C%22name%22%3A%22Anniversary%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22graduation%22%3A%7B%22id%22%3A%22graduation%22%2C%22name%22%3A%22Graduation%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22rosh-hashanah-yom-kippur%22%3A%7B%22id%22%3A%22rosh-hashanah-yom-kippur%22%2C%22name%22%3A%22Rosh%20Hashanah%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22ramadan%22%3A%7B%22id%22%3A%22ramadan%22%2C%22name%22%3A%22Ramadan%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%2C%22bastille-day%22%3A%7B%22id%22%3A%22bastille-day%22%2C%22name%22%3A%22Bastille%20Day%22%2C%22orderCount%22%3A0%2C%22count%22%3A0%7D%7D%7D%2C%22filterCount%22%3A18%2C%22historyPop%22%3Afalse%2C%22idWithQuickView%22%3A%22%22%2C%22initialFacets%22%3A%7B%22content%22%3A%7B%22article%22%3A16%2C%22recipe%22%3A2%7D%7D%2C%22inputTerms%22%3A%22%20tofu%20chili%22%2C%22searchTerms%22%3A%22%20tofu%20chili%22%2C%22refreshResults%22%3Afalse%2C%22resultGroups%22%3A%5Bnull%2C%7B%22items%22%3A%5B%7B%22id%22%3A%2254a4353c6529d92b2c017b43%22%2C%22dek%22%3A%22Dau%20hu%20xa%20ot%5CnEditor's%20note%3A%20The%20recipe%20and%20introductory%20text%20below%20are%20excerpted%20from%20Pleasures%20of%20the%20Vietnamese%20Table%20by%20Mai%20Pham%20and%20are%20part%20of%20our%20story%20on%20Lunar%20New%20Year.%5CnWhile%20traveling%20on%20a%20train%20one%20time%20to%20the%20coastal%20town%20of%20Nha%20Trang%2C%20I%20sat%20next%20to%20an%20elderly%20nun.%20Over%20the%20course%20of%20our%20bumpy%20eight-hour%20ride%2C%20she%20shared%20stories%20of%20life%20at%20the%20temple%20and%20the%20difficult%20years%20after%20the%20end%20of%20the%20war%20when%20the%20Communist%20government%20cracked%20down%20on%20religious%20factions.%20Toward%20the%20end%20of%20our%20chat%2C%20she%20pulled%20out%20a%20bag%20of%20food%20she'd%20prepared%20for%20the%20trip.%20It%20was%20tofu%20that%20had%20been%20cooked%20in%20chilies%2C%20lemongrass%20and%20la%20lot%2C%20an%20aromatic%20leaf%20also%20known%20as%20pepper%20leaf.%20When%20she%20gave%20me%20a%20taste%2C%20I%20knew%20immediately%20that%20I%20had%20to%20learn%20how%20to%20make%20it.%20This%20is%20my%20rendition%20of%20that%20fabulous%20dish.%20Make%20sure%20to%20pat%20the%20tofu%20dry%20before%20marinating%20it%20and%20use%20very%20fresh%20lemongrass.%20I%20always%20love%20serving%20this%20to%20friends%20who%20think%20tofu%20dishes%20are%20bland.%22%2C%22hed%22%3A%22Spicy%20Lemongrass%20Tofu%22%2C%22pubDate%22%3A%222006-05-16T20%3A12%3A55.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Mai%20Pham%22%7D%5D%2C%22type%22%3A%22recipe%22%2C%22url%22%3A%22%2Frecipes%2Ffood%2Fviews%2Fspicy-lemongrass-tofu-233844%22%2C%22photoData%22%3A%7B%22id%22%3A%225a0ddb046e013d11dde39630%22%2C%22filename%22%3A%22no-recipe-card-green-15112017.jpg%22%2C%22caption%22%3A%22No%20Recipe%20Card-%20GREEN%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%22%2C%22colors%22%3A%7B%22average%22%3A%22%23000000%22%7D%2C%22type%22%3A%22photo%22%7D%2C%22tag%22%3A%7B%22category%22%3A%22cuisine%22%2C%22name%22%3A%22Vietnamese%22%2C%22url%22%3A%22%22%2C%22photosBadgeAltText%22%3A%22%22%2C%22photosBadgeFileName%22%3A%22%22%2C%22photosBadgeID%22%3A%22%22%2C%22photosBadgeRelatedUri%22%3A%22%22%7D%2C%22aggregateRating%22%3A3.38%2C%22ingredients%22%3A%5B%222%20lemongrass%20stalks%2C%20outer%20layers%20peeled%2C%20bottom%20white%20part%20thinly%20sliced%20and%20finely%20chopped%20(about%201%2F4%20cup)%22%2C%221%201%2F2%20tablespoons%20soy%20sauce%22%2C%222%20teaspoons%20chopped%20Thai%20bird%20chilies%20or%20another%20fresh%20chili%22%2C%221%2F2%20teaspoon%20dried%20chili%20flakes%22%2C%221%20teaspoon%20ground%20turmeric%22%2C%222%20teaspoons%20sugar%22%2C%221%2F2%20teaspoon%20salt%22%2C%2212%20ounces%20tofu%2C%20drained%2C%20patted%20dry%20and%20cut%20into%203%2F4-inch%20cubes%22%2C%224%20tablespoons%20vegetable%20oil%22%2C%221%2F2%20yellow%20onion%2C%20cut%20into%201%2F8-inch%20slices%22%2C%222%20shallots%2C%20thinly%20sliced%22%2C%221%20teaspoon%20minced%20garlic%22%2C%224%20tablespoons%20chopped%20roasted%20peanuts%22%2C%2210%20la%20lot%2C%20or%20pepper%20leaves%2C%20shredded%2C%20or%202%2F3%20cup%20loosely%20packed%20Asian%20basil%20leaves%22%5D%2C%22prepSteps%22%3A%5B%221.%20Combine%20the%20lemongrass%2C%20soy%20sauce%2C%20chilies%2C%20chili%20flakes%2C%20turmeric%2C%20sugar%20and%20salt%20in%20a%20bowl.%20Add%20the%20tofu%20cubes%20and%20turn%20to%20coat%20them%20evenly.%20Marinate%20for%2030%20minutes.%22%2C%222.%20Heat%20half%20of%20the%20oil%20in%20a%2012-inch%20nonstick%20skillet%20over%20moderately%20high%20heat.%20Add%20the%20onion%2C%20shallot%20and%20garlic%20and%20stir%20until%20fragrant%2C%20about%201%20minute.%20Reduce%20the%20heat%20to%20low%20and%20cook%20until%20the%20onions%20are%20soft%2C%20about%203%20minutes.%20Transfer%20to%20a%20plate%20and%20keep%20warm.%22%2C%223.%20Wipe%20the%20pan%20clean%20and%20heat%20the%20remaining%20oil%20over%20moderate%20heat.%20Add%20the%20tofu%20mixture%20and%2C%20using%20chopstick%20or%20wooden%20spoons%2C%20turn%20so%20it%20cooks%20evenly%2C%20about%204%20to%205%20minutes.%20Add%20the%20onion%20mixture%20and%20cook%2C%20uncovered%2C%20for%20another%202%20to%203%20minutes.%20Add%20half%20the%20peanuts%20and%20all%20the%20pepper%20leaves.%22%2C%224.%20Remove%20from%20the%20heat%20and%20transfer%20to%20a%20serving%20plate.%20Garnish%20with%20the%20remaining%20peanuts%20and%20serve%20immediately%20with%20steamed%20rice.%22%5D%2C%22reviewsCount%22%3A17%2C%22willMakeAgainPct%22%3A88%7D%2C%7B%22id%22%3A%22586d586dad47301837c7c437%22%2C%22dek%22%3A%22Those%20packages%20of%20pink%20meat%20might%20not%20be%20as%20healthy%20as%20you%20think.%22%2C%22hed%22%3A%22Ground%20Turkey%20Breast%20Must%20Be%20Stopped%22%2C%22pubDate%22%3A%222017-01-09T22%3A30%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Adina%20Steiman%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fground-turkey-breast-versus-ground-dark-meat-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2254aca7e419925f464b3ae4a3%22%2C%22filename%22%3A%2251218010_thai-turkey-burgers_1x1.jpg%22%2C%22caption%22%3A%22Thai%20Turkey%20Burgers%20with%20Crispy%20Kale%22%2C%22credit%22%3A%22Will%20Heap%22%2C%22promoTitle%22%3A%22Thai%20Turkey%20Burgers%20with%20Crispy%20Kale%22%2C%22title%22%3A%22Thai%20Turkey%20Burgers%20with%20Crispy%20Kale%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%22560195d208929a1609a0874f%22%2C%22dek%22%3A%22Takeout%20is%20convenient%20and%20all.%20But%20in%20the%20time%20it%20takes%20to%20wait%20for%20delivery%2C%20you%20can%20make%20lighter%2C%20brighter%2C%20fresher%20pad%20Thai%20at%20home.%22%2C%22hed%22%3A%22How%20to%20Make%20Better-Than-Takeout%20Pad%20Thai%20in%2022%20Minutes%22%2C%22pubDate%22%3A%222015-09-23T14%3A25%3A11.547Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22%20Mindy%20Fox%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fhow-to-make-pad-thai-in-22-minutes-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2256019379c301a9f945eeec1d%22%2C%22filename%22%3A%22EP_09212015_padthai_HERO.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Food%20Styling%20by%20Mindy%20Fox%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Pad%20Thai%20HERO%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2255005a88fe3762782ee2abe1%22%2C%22dek%22%3A%22Sriracha's%20great%20and%20all%2C%20but%20we've%20moved%20on.%22%2C%22hed%22%3A%22Forget%20Sriracha%3A%20New%20Ways%20to%20Spice%20Up%20Everything%22%2C%22pubDate%22%3A%222015-03-11T17%3A05%3A05.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Matt%20Duckor%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fingredients%2Ffour-hot-sauces-beyond-sriracha-article%22%2C%22photoData%22%3A%7B%22id%22%3A%22550059adfe3762782ee2aa4b%22%2C%22filename%22%3A%22EP-03102015-hotsauce-tops-6x4.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22EP-03102015-hotsauce-tops-6x4.jpg%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2254c979ddc252103035d36c0b%22%2C%22dek%22%3A%22Upgrade%20your%20instant%20ramen%20with%20help%20from%20Chef%20Bill%20Kim%20of%20UrbanBelly%20in%20Chicago.%22%2C%22hed%22%3A%22It's%20Time%20To%20Up%20Your%20Instant%20Ramen%20Game%22%2C%22pubDate%22%3A%222015-01-29T00%3A13%3A01.545Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Sheela%20Prakash%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fits-time-to-up-your-instant-ramen-game-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2254c97a24c252103035d36c0d%22%2C%22filename%22%3A%229x4.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22%22%2C%22promoTitle%22%3A%22Dry%20Instant%20Ramen%20Noodles%22%2C%22title%22%3A%22Dry%20Instant%20Ramen%20Noodles%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2256290173e82ff50f71a50a27%22%2C%22dek%22%3A%22With%20these%20ingredients%20always%20on%20hand%2C%20there's%20no%20Thai%20(or%20Indian)%20(or%20Chinese)%20recipe%20you%20can't%20try.%22%2C%22hed%22%3A%22The%20International%20Ingredients%20You%20Need%20in%20Your%20Freezer%20Right%20Now%22%2C%22pubDate%22%3A%222015-10-23T16%3A04%3A27.286Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22The%20Epicurious%20Editors%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fingredients%2Ffrozen-ethnic-international-thai-chinese-indian-ingredients-article%22%2C%22photoData%22%3A%7B%22id%22%3A%225628f99dc538d7c61b19af94%22%2C%22filename%22%3A%22EP_10212015_EthnicFoods_hero.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Prop%20Styling%20by%20Alex%20Brannian%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Freezer%20Pantry%20Ethnic%20Foods-%20HERO%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2258c86c391c37a669031b2fdc%22%2C%22dek%22%3A%22A%20week%20of%20healthy%20meat-free%20dinners%20to%20get%20ready%20for%20spring.%22%2C%22hed%22%3A%22Easy%20Vegetarian%20Dinners%20for%20Every%20Night%20of%20the%20Week%22%2C%22pubDate%22%3A%222017-03-26T11%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Katherine%20Sacks%22%2C%22uri%22%3A%22http%3A%2F%2Fwww.epicurious.com%2Fcontributors%2Fkatherine-sacks%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Frecipes-menus%2Feasy-vegetarian-dinners-for-every-night-of-the-week-article%22%2C%22photoData%22%3A%7B%22id%22%3A%225877a9262ee1d51f664035fb%22%2C%22filename%22%3A%22Miso-Tahini-Squash-Soup-With-Brown-Rice-recipe-12012017.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Food%20Styling%20by%20Katherine%20Sacks%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Miso-Tahini%20Squash%20Soup%20With%20Brown%20Rice%20Recipe%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2256fc1bfce1a64a6456ab9bc1%22%2C%22dek%22%3A%22Bowl%20food%20doesn't%20get%20beautiful%20(and%20nutritious)%20(and%20filling)%20automatically%E2%80%94it%20takes%20a%20plan.%20And%20there's%20nobody%20better%20to%20get%20that%20plan%20from%20than%20the%20guy%20who%20literally%20wrote%20the%20book%20on%20the%20subject.%22%2C%22hed%22%3A%22How%20to%20Become%20a%20Pro%20Bowler%22%2C%22pubDate%22%3A%222016-03-30T23%3A33%3A05.306Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Lukas%20Volger%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fhow-to-make-grain-noodle-bowls-lukas-volger-article%22%2C%22photoData%22%3A%7B%22id%22%3A%22568194b8fb9544f72b678fd4%22%2C%22filename%22%3A%22Ramen-Noodle-Bowl-With-Escarole.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Peden%20%26%20Munk%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Ramen-Noodle-Bowl-With-Escarole.jpg%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%225a0330ea57e96a30a35d163c%22%2C%22dek%22%3A%22The%20novelist%20and%20independent%20book%20publisher%20relies%20on%20intuition%2C%20leftovers%2C%20and%20some%20very%20good%20cookbooks%20to%20make%20dinner%20happen.%22%2C%22hed%22%3A%22What%20the%20Writer-Publisher-Mom%20Emily%20Gould%20Cooks%20in%20a%20Week%20%20%22%2C%22pubDate%22%3A%222017-11-08T18%3A39%3A04.371Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Emily%20Gould%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Femily-gould-cooking-diary-article%22%2C%22photoData%22%3A%7B%22id%22%3A%225a021ac7d63f7339eb97b547%22%2C%22filename%22%3A%22Small-Plates-Dinner-Diary-Emily-Gould-Alt-2.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Caleb%20Adams%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Small%20Plates%20Dinner%20Diary%20Emily%20Gould%20Alt-2%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2256c3b2bc954091771eccb7a5%22%2C%22dek%22%3A%22It's%20the%20all-you-can-eat%20fondue%20party%20that%2C%20thankfully%2C%20isn't%20a%20fondue%20party%20at%20all.%22%2C%22hed%22%3A%22How%20to%20Throw%20a%20Hot%20Pot%20Party%20With%20a%20Slow%20Cooker%22%2C%22pubDate%22%3A%222016-02-22T21%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Tommy%20Werner%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fhow-to-throw-hot-pot-party-at-home-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2256c49a3baf8e89242020c65c%22%2C%22filename%22%3A%22CC5P7196.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Sarah%20Keough%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Hot%20Pot%20Photo-%20HERO%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%225877bca8b3246f1f0afae652%22%2C%22dek%22%3A%22You%20probably%20have%20it%20in%20the%20fridge%20already.%22%2C%22hed%22%3A%22The%20Unlikely%20Ingredient%20That's%20Key%20to%20Making%20This%20Hearty%20Vegetarian%20Soup%22%2C%22pubDate%22%3A%222017-01-23T23%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Katherine%20Sacks%22%2C%22uri%22%3A%22http%3A%2F%2Fwww.epicurious.com%2Fcontributors%2Fkatherine-sacks%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fhearty-vegetarian-soup-miso-tahini-recipe-vegan-article%22%2C%22photoData%22%3A%7B%22id%22%3A%225877a9263b8997bc108a447a%22%2C%22filename%22%3A%22Miso-Tahini-Squash-Soup-With-Brown-Rice-hero-12012017.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Food%20Styling%20by%20Katherine%20Sacks%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Miso-Tahini%20Squash%20Soup%20With%20Brown%20Rice%20Hero%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%225a5544e97b8677001031c6c2%22%2C%22dek%22%3A%22For%20starters%2C%20don't%20underestimate%20their%20taste%20preferences.%22%2C%22hed%22%3A%22How%20to%20Get%20Your%20Kids%20to%20Eat%E2%80%94and%20Maybe%20Even%20Like%E2%80%94Spicy%20Foods%22%2C%22pubDate%22%3A%222018-01-16T20%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Janet%20Rausa%20Fuller%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fhow-to-encourage-your-kids-to-try-spicy-foods-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2258e52bfe0e58ea1cc02885db%22%2C%22filename%22%3A%22Chicken-Enchilada-Skillet-05042017.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Food%20Styling%20by%20Anna%20Stockwell%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Chicken%20Enchilada%20Skillet%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2259af4272cd4b601170effb8f%22%2C%22dek%22%3A%22This%20is%20not%20your%20average%20low-carb%20eating%20plan.%22%2C%22hed%22%3A%22What%20the%20Heck%20Is%20the%20Low%20FODMAP%20Diet%3F%22%2C%22pubDate%22%3A%222017-09-07T17%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Janet%20Rausa%20Fuller%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fwhat-is-the-low-fodmap-diet-article%22%2C%22photoData%22%3A%7B%22id%22%3A%22599b45c2c7b38c7c1757e5e3%22%2C%22filename%22%3A%22pounded-flank-steak-with-zucchini-salsa-recipe-BA-081717.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Alex%20Lau%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Pounded%20Flank%20Steak%20with%20Zucchini%20Salsa%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2258e2b0412ebb9b1ae1fbffe2%22%2C%22dek%22%3A%22From%20dried%20lentils%20to%20cashew%20butter%2C%20here%20are%20the%2010%20things%20vegans%20should%20never%20leave%20the%20store%20without.%22%2C%22hed%22%3A%2210%20Groceries%20to%20Buy%20Every%20Single%20Week%20If%20You're%20Vegan%22%2C%22pubDate%22%3A%222017-04-14T19%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Katherine%20Sacks%22%2C%22uri%22%3A%22http%3A%2F%2Fwww.epicurious.com%2Fcontributors%2Fkatherine-sacks%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2F10-vegan-groceries-to-buy-every-single-week-recipes-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2258e2673b2ebb9b1ae1fbffd7%22%2C%22filename%22%3A%22Vegan-Grocery-List-03292017.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Food%20Styling%20by%20Katherine%20Sacks%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22The%2013%20Foods%20to%20Buy%20Every%20Week%20If%20You're%20Vegan%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2259e20a9e5fe95265dcaf065b%22%2C%22dek%22%3A%22Chicago-based%20cookbook%20author%20Anupy%20Singla%20changed%20careers%20in%20order%20to%20change%20her%20family's%20eating%20habits.%20It%20worked.%22%2C%22hed%22%3A%22Here's%20Why%20My%20Kids%20Eat%20Indian%20Food%20Almost%20Every%20Night%22%2C%22pubDate%22%3A%222017-10-24T21%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Anupy%20Singla%20as%20told%20to%20Janet%20Rausa%20Fuller%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fwhy-my-kids-eat-indian-food-almost-every-night-article%22%2C%22photoData%22%3A%7B%22id%22%3A%225877ae09662a112d66d6b5f3%22%2C%22filename%22%3A%22Cumin-Scented-Potatoes-With-Tomatoes-recipe-12012017.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Prop%20Styling%20by%20Alex%20Brannian%2C%20Food%20Styling%20by%20Anna%20Hampton%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Cumin%20Scented%20Potatoes%20With%20Tomatoes%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2257165251cdfa3dbe4601dcac%22%2C%22dek%22%3A%22Everything%20you%20need%20to%20know%20when%20buying%20pickles.%22%2C%22hed%22%3A%22How%20to%20Navigate%20the%20Pickle%20Aisle%22%2C%22pubDate%22%3A%222016-04-21T15%3A30%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Tommy%20Werner%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fingredients%2Fultimate-guide-to-buying-pickles-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2257164d920e3f6ffc32109070%22%2C%22filename%22%3A%22shutterstock_284400587.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Shutterstock%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Pickle%20Jar%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2258b9734669e8831407f8f384%22%2C%22dek%22%3A%22This%20old-school%20herbal%20remedy%20can%20be%20turned%20into%20new-school%20cocktails%2C%20salads%2C%20and%20more.%22%2C%22hed%22%3A%22How%20to%20Make%20Fire%20Cider%20at%20Home%E2%80%94and%20Cook%20With%20It%22%2C%22pubDate%22%3A%222017-03-03T18%3A23%3A23.315Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Lukas%20Volger%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Frecipes-menus%2Ffire-cider-recipe-cooking-tips-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2258af3b41d011dc426027376b%22%2C%22filename%22%3A%22Fire-Cider-hero-17022017.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Food%20Styling%20by%20Ali%20Nardi%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Fire%20Cider%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2254a428a26529d92b2c00d920%22%2C%22dek%22%3A%22If%20you%20live%20near%20a%20Chinese%20market%2C%20pick%20up%20barbecued%20or%20smoked%20duck%20there.%20Otherwise%2C%20smoked%20chicken%20or%20turkey%20from%20the%20supermarket%20(or%20leftover%20roast%20chicken)%20would%20be%20terrific%20tossed%20with%20the%20noodles.%20To%20make%20it%20a%20meal%2C%20add%20a%20platter%20of%20chilled%20silken%20tofu.%20Drizzle%20the%20tofu%20with%20soy%20sauce%20and%20chili%20sauce%2C%20then%20top%20with%20chopped%20green%20onions.%20Coconut%20ice%20cream%20with%20fresh%20berries%20and%20lychees%20would%20make%20a%20nice%20dessert.%22%2C%22hed%22%3A%22Chinese%20Egg%20Noodles%20with%20Smoked%20Duck%20and%20Snow%20Peas%22%2C%22pubDate%22%3A%222009-07-02T04%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Jill%20Dupleix%22%7D%5D%2C%22type%22%3A%22recipe%22%2C%22url%22%3A%22%2Frecipes%2Ffood%2Fviews%2Fchinese-egg-noodles-with-smoked-duck-and-snow-peas-354302%22%2C%22photoData%22%3A%7B%22id%22%3A%2254cd1c181f13bb9b2edfc482%22%2C%22filename%22%3A%22ChineseEggNoodleswithSmokedDuckandSnowPeas_6x4.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Misha%20Gravenor%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Chinese%20Egg%20Noodles%20with%20Smoked%20Duck%20and%20Snow%20Peas%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%2C%22tag%22%3A%7B%22category%22%3A%22cuisine%22%2C%22name%22%3A%22Asian%22%2C%22url%22%3A%22%22%2C%22photosBadgeAltText%22%3A%22%22%2C%22photosBadgeFileName%22%3A%22%22%2C%22photosBadgeID%22%3A%22%22%2C%22photosBadgeRelatedUri%22%3A%22%22%7D%2C%22aggregateRating%22%3A2.25%2C%22ingredients%22%3A%5B%222%20tablespoons%20soy%20sauce%22%2C%222%20tablespoons%20olive%20oil%22%2C%221%20tablespoon%20hoisin%20sauce*%22%2C%221%20tablespoon%20unseasoned%20rice%20vinegar%22%2C%223%201%2F2%20ounces%20snow%20peas%20or%20sugar%20snap%20peas%2C%20thinly%20sliced%20lengthwise%20(about%201%201%2F2%20cups)%22%2C%221%20red%20bell%20pepper%2C%20cut%20into%20matchstick-size%20strips%22%2C%221%2017-ounce%20(425-gram)%20package%20thin%20fresh%20Chinese%20egg%20noodles**%22%2C%221%20pound%20smoked%20duck%2C%20turkey%2C%20or%20chicken%2C%20cut%20into%203x1%2F4-inch%20strips%22%2C%221%2F2%20cup%20fresh%20cilantro%20leaves%22%2C%222%20tablespoons%20thinly%20sliced%20green%20onion%22%2C%221%20tablespoon%20thinly%20sliced%20jalape%C3%B1o%20chile%20with%20seeds%20(preferably%20red)%22%5D%2C%22prepSteps%22%3A%5B%22Whisk%20soy%20sauce%2C%20olive%20oil%2C%20hoisin%20sauce%2C%20and%20rice%20vinegar%20in%20large%20bowl%20to%20blend.%20Let%20dressing%20stand%20at%20room%20temperature%20while%20preparing%20noodles.%22%2C%22Place%20sliced%20snow%20peas%20and%20bell%20pepper%20strips%20in%20medium%20bowl.%20Bring%20kettle%20or%20large%20saucepan%20of%20water%20to%20boil.%20Pour%20boiling%20water%20over%20snow%20pea%20mixture%20and%20let%20soak%2010%20seconds.%20Drain%20well.%20Rinse%20mixture%20with%20cold%20water%20and%20drain%20again.%20Place%20egg%20noodles%20in%20large%20bowl.%20Bring%20kettle%20or%20large%20saucepan%20of%20water%20to%20boil%2C%20then%20pour%20boiling%20water%20over%20noodles%3B%20stir%20noodles%20to%20separate.%20Let%20noodles%20soak%2010%20seconds.%20Drain%20well.%20Rinse%20noodles%20with%20cold%20water%3B%20drain%20again.%22%2C%22Add%20snow%20pea%20mixture%2C%20egg%20noodles%2C%20smoked%20duck%2C%20cilantro%20leaves%2C%20green%20onion%2C%20and%20jalape%C3%B1o%20to%20bowl%20with%20dressing.%20Toss%20well%20to%20coat.%20Divide%20noodle%20mixture%20among%20plates%20and%20serve.%22%2C%22*%20Hoisin%20sauce%20is%20available%20in%20the%20Asian%20foods%20section%20of%20many%20supermarkets%20and%20at%20Asian%20markets.%22%2C%22**%20Fresh%20Chinese%20egg%20noodles%20are%20available%20in%20the%20refrigerated%20section%20of%20Asian%20markets.%22%5D%2C%22reviewsCount%22%3A4%2C%22willMakeAgainPct%22%3A50%7D%5D%2C%22page%22%3A%7B%22generatedAt%22%3A%222018-02-26T12%3A43%3A15.272Z%22%2C%22count%22%3A1%2C%22number%22%3A1%2C%22size%22%3A18%2C%22totalCount%22%3A18%2C%22previousNumber%22%3A1%7D%7D%5D%2C%22resultGroupDisplaySize%22%3A18%2C%22spellcheck%22%3A%7B%22requested%22%3A%22%20tofu%20chili%22%7D%2C%22suggestions%22%3A%5B%5D%2C%22walkthrough%22%3A%7B%7D%2C%22items%22%3A%5B%7B%22id%22%3A%2254a4353c6529d92b2c017b43%22%2C%22dek%22%3A%22Dau%20hu%20xa%20ot%5CnEditor's%20note%3A%20The%20recipe%20and%20introductory%20text%20below%20are%20excerpted%20from%20Pleasures%20of%20the%20Vietnamese%20Table%20by%20Mai%20Pham%20and%20are%20part%20of%20our%20story%20on%20Lunar%20New%20Year.%5CnWhile%20traveling%20on%20a%20train%20one%20time%20to%20the%20coastal%20town%20of%20Nha%20Trang%2C%20I%20sat%20next%20to%20an%20elderly%20nun.%20Over%20the%20course%20of%20our%20bumpy%20eight-hour%20ride%2C%20she%20shared%20stories%20of%20life%20at%20the%20temple%20and%20the%20difficult%20years%20after%20the%20end%20of%20the%20war%20when%20the%20Communist%20government%20cracked%20down%20on%20religious%20factions.%20Toward%20the%20end%20of%20our%20chat%2C%20she%20pulled%20out%20a%20bag%20of%20food%20she'd%20prepared%20for%20the%20trip.%20It%20was%20tofu%20that%20had%20been%20cooked%20in%20chilies%2C%20lemongrass%20and%20la%20lot%2C%20an%20aromatic%20leaf%20also%20known%20as%20pepper%20leaf.%20When%20she%20gave%20me%20a%20taste%2C%20I%20knew%20immediately%20that%20I%20had%20to%20learn%20how%20to%20make%20it.%20This%20is%20my%20rendition%20of%20that%20fabulous%20dish.%20Make%20sure%20to%20pat%20the%20tofu%20dry%20before%20marinating%20it%20and%20use%20very%20fresh%20lemongrass.%20I%20always%20love%20serving%20this%20to%20friends%20who%20think%20tofu%20dishes%20are%20bland.%22%2C%22hed%22%3A%22Spicy%20Lemongrass%20Tofu%22%2C%22pubDate%22%3A%222006-05-16T20%3A12%3A55.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Mai%20Pham%22%7D%5D%2C%22type%22%3A%22recipe%22%2C%22url%22%3A%22%2Frecipes%2Ffood%2Fviews%2Fspicy-lemongrass-tofu-233844%22%2C%22photoData%22%3A%7B%22id%22%3A%225a0ddb046e013d11dde39630%22%2C%22filename%22%3A%22no-recipe-card-green-15112017.jpg%22%2C%22caption%22%3A%22No%20Recipe%20Card-%20GREEN%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%22%2C%22colors%22%3A%7B%22average%22%3A%22%23000000%22%7D%2C%22type%22%3A%22photo%22%7D%2C%22tag%22%3A%7B%22category%22%3A%22cuisine%22%2C%22name%22%3A%22Vietnamese%22%2C%22url%22%3A%22%22%2C%22photosBadgeAltText%22%3A%22%22%2C%22photosBadgeFileName%22%3A%22%22%2C%22photosBadgeID%22%3A%22%22%2C%22photosBadgeRelatedUri%22%3A%22%22%7D%2C%22aggregateRating%22%3A3.38%2C%22ingredients%22%3A%5B%222%20lemongrass%20stalks%2C%20outer%20layers%20peeled%2C%20bottom%20white%20part%20thinly%20sliced%20and%20finely%20chopped%20(about%201%2F4%20cup)%22%2C%221%201%2F2%20tablespoons%20soy%20sauce%22%2C%222%20teaspoons%20chopped%20Thai%20bird%20chilies%20or%20another%20fresh%20chili%22%2C%221%2F2%20teaspoon%20dried%20chili%20flakes%22%2C%221%20teaspoon%20ground%20turmeric%22%2C%222%20teaspoons%20sugar%22%2C%221%2F2%20teaspoon%20salt%22%2C%2212%20ounces%20tofu%2C%20drained%2C%20patted%20dry%20and%20cut%20into%203%2F4-inch%20cubes%22%2C%224%20tablespoons%20vegetable%20oil%22%2C%221%2F2%20yellow%20onion%2C%20cut%20into%201%2F8-inch%20slices%22%2C%222%20shallots%2C%20thinly%20sliced%22%2C%221%20teaspoon%20minced%20garlic%22%2C%224%20tablespoons%20chopped%20roasted%20peanuts%22%2C%2210%20la%20lot%2C%20or%20pepper%20leaves%2C%20shredded%2C%20or%202%2F3%20cup%20loosely%20packed%20Asian%20basil%20leaves%22%5D%2C%22prepSteps%22%3A%5B%221.%20Combine%20the%20lemongrass%2C%20soy%20sauce%2C%20chilies%2C%20chili%20flakes%2C%20turmeric%2C%20sugar%20and%20salt%20in%20a%20bowl.%20Add%20the%20tofu%20cubes%20and%20turn%20to%20coat%20them%20evenly.%20Marinate%20for%2030%20minutes.%22%2C%222.%20Heat%20half%20of%20the%20oil%20in%20a%2012-inch%20nonstick%20skillet%20over%20moderately%20high%20heat.%20Add%20the%20onion%2C%20shallot%20and%20garlic%20and%20stir%20until%20fragrant%2C%20about%201%20minute.%20Reduce%20the%20heat%20to%20low%20and%20cook%20until%20the%20onions%20are%20soft%2C%20about%203%20minutes.%20Transfer%20to%20a%20plate%20and%20keep%20warm.%22%2C%223.%20Wipe%20the%20pan%20clean%20and%20heat%20the%20remaining%20oil%20over%20moderate%20heat.%20Add%20the%20tofu%20mixture%20and%2C%20using%20chopstick%20or%20wooden%20spoons%2C%20turn%20so%20it%20cooks%20evenly%2C%20about%204%20to%205%20minutes.%20Add%20the%20onion%20mixture%20and%20cook%2C%20uncovered%2C%20for%20another%202%20to%203%20minutes.%20Add%20half%20the%20peanuts%20and%20all%20the%20pepper%20leaves.%22%2C%224.%20Remove%20from%20the%20heat%20and%20transfer%20to%20a%20serving%20plate.%20Garnish%20with%20the%20remaining%20peanuts%20and%20serve%20immediately%20with%20steamed%20rice.%22%5D%2C%22reviewsCount%22%3A17%2C%22willMakeAgainPct%22%3A88%7D%2C%7B%22id%22%3A%22586d586dad47301837c7c437%22%2C%22dek%22%3A%22Those%20packages%20of%20pink%20meat%20might%20not%20be%20as%20healthy%20as%20you%20think.%22%2C%22hed%22%3A%22Ground%20Turkey%20Breast%20Must%20Be%20Stopped%22%2C%22pubDate%22%3A%222017-01-09T22%3A30%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Adina%20Steiman%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fground-turkey-breast-versus-ground-dark-meat-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2254aca7e419925f464b3ae4a3%22%2C%22filename%22%3A%2251218010_thai-turkey-burgers_1x1.jpg%22%2C%22caption%22%3A%22Thai%20Turkey%20Burgers%20with%20Crispy%20Kale%22%2C%22credit%22%3A%22Will%20Heap%22%2C%22promoTitle%22%3A%22Thai%20Turkey%20Burgers%20with%20Crispy%20Kale%22%2C%22title%22%3A%22Thai%20Turkey%20Burgers%20with%20Crispy%20Kale%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%22560195d208929a1609a0874f%22%2C%22dek%22%3A%22Takeout%20is%20convenient%20and%20all.%20But%20in%20the%20time%20it%20takes%20to%20wait%20for%20delivery%2C%20you%20can%20make%20lighter%2C%20brighter%2C%20fresher%20pad%20Thai%20at%20home.%22%2C%22hed%22%3A%22How%20to%20Make%20Better-Than-Takeout%20Pad%20Thai%20in%2022%20Minutes%22%2C%22pubDate%22%3A%222015-09-23T14%3A25%3A11.547Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22%20Mindy%20Fox%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fhow-to-make-pad-thai-in-22-minutes-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2256019379c301a9f945eeec1d%22%2C%22filename%22%3A%22EP_09212015_padthai_HERO.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Food%20Styling%20by%20Mindy%20Fox%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Pad%20Thai%20HERO%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2255005a88fe3762782ee2abe1%22%2C%22dek%22%3A%22Sriracha's%20great%20and%20all%2C%20but%20we've%20moved%20on.%22%2C%22hed%22%3A%22Forget%20Sriracha%3A%20New%20Ways%20to%20Spice%20Up%20Everything%22%2C%22pubDate%22%3A%222015-03-11T17%3A05%3A05.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Matt%20Duckor%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fingredients%2Ffour-hot-sauces-beyond-sriracha-article%22%2C%22photoData%22%3A%7B%22id%22%3A%22550059adfe3762782ee2aa4b%22%2C%22filename%22%3A%22EP-03102015-hotsauce-tops-6x4.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22EP-03102015-hotsauce-tops-6x4.jpg%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2254c979ddc252103035d36c0b%22%2C%22dek%22%3A%22Upgrade%20your%20instant%20ramen%20with%20help%20from%20Chef%20Bill%20Kim%20of%20UrbanBelly%20in%20Chicago.%22%2C%22hed%22%3A%22It's%20Time%20To%20Up%20Your%20Instant%20Ramen%20Game%22%2C%22pubDate%22%3A%222015-01-29T00%3A13%3A01.545Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Sheela%20Prakash%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fits-time-to-up-your-instant-ramen-game-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2254c97a24c252103035d36c0d%22%2C%22filename%22%3A%229x4.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22%22%2C%22promoTitle%22%3A%22Dry%20Instant%20Ramen%20Noodles%22%2C%22title%22%3A%22Dry%20Instant%20Ramen%20Noodles%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2256290173e82ff50f71a50a27%22%2C%22dek%22%3A%22With%20these%20ingredients%20always%20on%20hand%2C%20there's%20no%20Thai%20(or%20Indian)%20(or%20Chinese)%20recipe%20you%20can't%20try.%22%2C%22hed%22%3A%22The%20International%20Ingredients%20You%20Need%20in%20Your%20Freezer%20Right%20Now%22%2C%22pubDate%22%3A%222015-10-23T16%3A04%3A27.286Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22The%20Epicurious%20Editors%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fingredients%2Ffrozen-ethnic-international-thai-chinese-indian-ingredients-article%22%2C%22photoData%22%3A%7B%22id%22%3A%225628f99dc538d7c61b19af94%22%2C%22filename%22%3A%22EP_10212015_EthnicFoods_hero.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Prop%20Styling%20by%20Alex%20Brannian%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Freezer%20Pantry%20Ethnic%20Foods-%20HERO%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2258c86c391c37a669031b2fdc%22%2C%22dek%22%3A%22A%20week%20of%20healthy%20meat-free%20dinners%20to%20get%20ready%20for%20spring.%22%2C%22hed%22%3A%22Easy%20Vegetarian%20Dinners%20for%20Every%20Night%20of%20the%20Week%22%2C%22pubDate%22%3A%222017-03-26T11%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Katherine%20Sacks%22%2C%22uri%22%3A%22http%3A%2F%2Fwww.epicurious.com%2Fcontributors%2Fkatherine-sacks%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Frecipes-menus%2Feasy-vegetarian-dinners-for-every-night-of-the-week-article%22%2C%22photoData%22%3A%7B%22id%22%3A%225877a9262ee1d51f664035fb%22%2C%22filename%22%3A%22Miso-Tahini-Squash-Soup-With-Brown-Rice-recipe-12012017.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Food%20Styling%20by%20Katherine%20Sacks%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Miso-Tahini%20Squash%20Soup%20With%20Brown%20Rice%20Recipe%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2256fc1bfce1a64a6456ab9bc1%22%2C%22dek%22%3A%22Bowl%20food%20doesn't%20get%20beautiful%20(and%20nutritious)%20(and%20filling)%20automatically%E2%80%94it%20takes%20a%20plan.%20And%20there's%20nobody%20better%20to%20get%20that%20plan%20from%20than%20the%20guy%20who%20literally%20wrote%20the%20book%20on%20the%20subject.%22%2C%22hed%22%3A%22How%20to%20Become%20a%20Pro%20Bowler%22%2C%22pubDate%22%3A%222016-03-30T23%3A33%3A05.306Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Lukas%20Volger%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fhow-to-make-grain-noodle-bowls-lukas-volger-article%22%2C%22photoData%22%3A%7B%22id%22%3A%22568194b8fb9544f72b678fd4%22%2C%22filename%22%3A%22Ramen-Noodle-Bowl-With-Escarole.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Peden%20%26%20Munk%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Ramen-Noodle-Bowl-With-Escarole.jpg%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%225a0330ea57e96a30a35d163c%22%2C%22dek%22%3A%22The%20novelist%20and%20independent%20book%20publisher%20relies%20on%20intuition%2C%20leftovers%2C%20and%20some%20very%20good%20cookbooks%20to%20make%20dinner%20happen.%22%2C%22hed%22%3A%22What%20the%20Writer-Publisher-Mom%20Emily%20Gould%20Cooks%20in%20a%20Week%20%20%22%2C%22pubDate%22%3A%222017-11-08T18%3A39%3A04.371Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Emily%20Gould%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Femily-gould-cooking-diary-article%22%2C%22photoData%22%3A%7B%22id%22%3A%225a021ac7d63f7339eb97b547%22%2C%22filename%22%3A%22Small-Plates-Dinner-Diary-Emily-Gould-Alt-2.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Caleb%20Adams%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Small%20Plates%20Dinner%20Diary%20Emily%20Gould%20Alt-2%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2256c3b2bc954091771eccb7a5%22%2C%22dek%22%3A%22It's%20the%20all-you-can-eat%20fondue%20party%20that%2C%20thankfully%2C%20isn't%20a%20fondue%20party%20at%20all.%22%2C%22hed%22%3A%22How%20to%20Throw%20a%20Hot%20Pot%20Party%20With%20a%20Slow%20Cooker%22%2C%22pubDate%22%3A%222016-02-22T21%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Tommy%20Werner%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fhow-to-throw-hot-pot-party-at-home-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2256c49a3baf8e89242020c65c%22%2C%22filename%22%3A%22CC5P7196.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Sarah%20Keough%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Hot%20Pot%20Photo-%20HERO%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%225877bca8b3246f1f0afae652%22%2C%22dek%22%3A%22You%20probably%20have%20it%20in%20the%20fridge%20already.%22%2C%22hed%22%3A%22The%20Unlikely%20Ingredient%20That's%20Key%20to%20Making%20This%20Hearty%20Vegetarian%20Soup%22%2C%22pubDate%22%3A%222017-01-23T23%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Katherine%20Sacks%22%2C%22uri%22%3A%22http%3A%2F%2Fwww.epicurious.com%2Fcontributors%2Fkatherine-sacks%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fhearty-vegetarian-soup-miso-tahini-recipe-vegan-article%22%2C%22photoData%22%3A%7B%22id%22%3A%225877a9263b8997bc108a447a%22%2C%22filename%22%3A%22Miso-Tahini-Squash-Soup-With-Brown-Rice-hero-12012017.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Food%20Styling%20by%20Katherine%20Sacks%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Miso-Tahini%20Squash%20Soup%20With%20Brown%20Rice%20Hero%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%225a5544e97b8677001031c6c2%22%2C%22dek%22%3A%22For%20starters%2C%20don't%20underestimate%20their%20taste%20preferences.%22%2C%22hed%22%3A%22How%20to%20Get%20Your%20Kids%20to%20Eat%E2%80%94and%20Maybe%20Even%20Like%E2%80%94Spicy%20Foods%22%2C%22pubDate%22%3A%222018-01-16T20%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Janet%20Rausa%20Fuller%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fhow-to-encourage-your-kids-to-try-spicy-foods-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2258e52bfe0e58ea1cc02885db%22%2C%22filename%22%3A%22Chicken-Enchilada-Skillet-05042017.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Food%20Styling%20by%20Anna%20Stockwell%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Chicken%20Enchilada%20Skillet%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2259af4272cd4b601170effb8f%22%2C%22dek%22%3A%22This%20is%20not%20your%20average%20low-carb%20eating%20plan.%22%2C%22hed%22%3A%22What%20the%20Heck%20Is%20the%20Low%20FODMAP%20Diet%3F%22%2C%22pubDate%22%3A%222017-09-07T17%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Janet%20Rausa%20Fuller%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fwhat-is-the-low-fodmap-diet-article%22%2C%22photoData%22%3A%7B%22id%22%3A%22599b45c2c7b38c7c1757e5e3%22%2C%22filename%22%3A%22pounded-flank-steak-with-zucchini-salsa-recipe-BA-081717.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Alex%20Lau%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Pounded%20Flank%20Steak%20with%20Zucchini%20Salsa%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2258e2b0412ebb9b1ae1fbffe2%22%2C%22dek%22%3A%22From%20dried%20lentils%20to%20cashew%20butter%2C%20here%20are%20the%2010%20things%20vegans%20should%20never%20leave%20the%20store%20without.%22%2C%22hed%22%3A%2210%20Groceries%20to%20Buy%20Every%20Single%20Week%20If%20You're%20Vegan%22%2C%22pubDate%22%3A%222017-04-14T19%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Katherine%20Sacks%22%2C%22uri%22%3A%22http%3A%2F%2Fwww.epicurious.com%2Fcontributors%2Fkatherine-sacks%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2F10-vegan-groceries-to-buy-every-single-week-recipes-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2258e2673b2ebb9b1ae1fbffd7%22%2C%22filename%22%3A%22Vegan-Grocery-List-03292017.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Food%20Styling%20by%20Katherine%20Sacks%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22The%2013%20Foods%20to%20Buy%20Every%20Week%20If%20You're%20Vegan%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2259e20a9e5fe95265dcaf065b%22%2C%22dek%22%3A%22Chicago-based%20cookbook%20author%20Anupy%20Singla%20changed%20careers%20in%20order%20to%20change%20her%20family's%20eating%20habits.%20It%20worked.%22%2C%22hed%22%3A%22Here's%20Why%20My%20Kids%20Eat%20Indian%20Food%20Almost%20Every%20Night%22%2C%22pubDate%22%3A%222017-10-24T21%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Anupy%20Singla%20as%20told%20to%20Janet%20Rausa%20Fuller%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fexpert-advice%2Fwhy-my-kids-eat-indian-food-almost-every-night-article%22%2C%22photoData%22%3A%7B%22id%22%3A%225877ae09662a112d66d6b5f3%22%2C%22filename%22%3A%22Cumin-Scented-Potatoes-With-Tomatoes-recipe-12012017.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Prop%20Styling%20by%20Alex%20Brannian%2C%20Food%20Styling%20by%20Anna%20Hampton%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Cumin%20Scented%20Potatoes%20With%20Tomatoes%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2257165251cdfa3dbe4601dcac%22%2C%22dek%22%3A%22Everything%20you%20need%20to%20know%20when%20buying%20pickles.%22%2C%22hed%22%3A%22How%20to%20Navigate%20the%20Pickle%20Aisle%22%2C%22pubDate%22%3A%222016-04-21T15%3A30%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Tommy%20Werner%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Fingredients%2Fultimate-guide-to-buying-pickles-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2257164d920e3f6ffc32109070%22%2C%22filename%22%3A%22shutterstock_284400587.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Shutterstock%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Pickle%20Jar%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2258b9734669e8831407f8f384%22%2C%22dek%22%3A%22This%20old-school%20herbal%20remedy%20can%20be%20turned%20into%20new-school%20cocktails%2C%20salads%2C%20and%20more.%22%2C%22hed%22%3A%22How%20to%20Make%20Fire%20Cider%20at%20Home%E2%80%94and%20Cook%20With%20It%22%2C%22pubDate%22%3A%222017-03-03T18%3A23%3A23.315Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Lukas%20Volger%22%7D%5D%2C%22type%22%3A%22article%22%2C%22url%22%3A%22%2Frecipes-menus%2Ffire-cider-recipe-cooking-tips-article%22%2C%22photoData%22%3A%7B%22id%22%3A%2258af3b41d011dc426027376b%22%2C%22filename%22%3A%22Fire-Cider-hero-17022017.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Chelsea%20Kyle%2C%20Food%20Styling%20by%20Ali%20Nardi%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Fire%20Cider%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%7D%2C%7B%22id%22%3A%2254a428a26529d92b2c00d920%22%2C%22dek%22%3A%22If%20you%20live%20near%20a%20Chinese%20market%2C%20pick%20up%20barbecued%20or%20smoked%20duck%20there.%20Otherwise%2C%20smoked%20chicken%20or%20turkey%20from%20the%20supermarket%20(or%20leftover%20roast%20chicken)%20would%20be%20terrific%20tossed%20with%20the%20noodles.%20To%20make%20it%20a%20meal%2C%20add%20a%20platter%20of%20chilled%20silken%20tofu.%20Drizzle%20the%20tofu%20with%20soy%20sauce%20and%20chili%20sauce%2C%20then%20top%20with%20chopped%20green%20onions.%20Coconut%20ice%20cream%20with%20fresh%20berries%20and%20lychees%20would%20make%20a%20nice%20dessert.%22%2C%22hed%22%3A%22Chinese%20Egg%20Noodles%20with%20Smoked%20Duck%20and%20Snow%20Peas%22%2C%22pubDate%22%3A%222009-07-02T04%3A00%3A00.000Z%22%2C%22author%22%3A%5B%7B%22name%22%3A%22Jill%20Dupleix%22%7D%5D%2C%22type%22%3A%22recipe%22%2C%22url%22%3A%22%2Frecipes%2Ffood%2Fviews%2Fchinese-egg-noodles-with-smoked-duck-and-snow-peas-354302%22%2C%22photoData%22%3A%7B%22id%22%3A%2254cd1c181f13bb9b2edfc482%22%2C%22filename%22%3A%22ChineseEggNoodleswithSmokedDuckandSnowPeas_6x4.jpg%22%2C%22caption%22%3A%22%22%2C%22credit%22%3A%22Photo%20by%20Misha%20Gravenor%22%2C%22promoTitle%22%3A%22%22%2C%22title%22%3A%22Chinese%20Egg%20Noodles%20with%20Smoked%20Duck%20and%20Snow%20Peas%22%2C%22orientation%22%3A%22landscape%22%2C%22restrictCropping%22%3Afalse%7D%2C%22tag%22%3A%7B%22category%22%3A%22cuisine%22%2C%22name%22%3A%22Asian%22%2C%22url%22%3A%22%22%2C%22photosBadgeAltText%22%3A%22%22%2C%22photosBadgeFileName%22%3A%22%22%2C%22photosBadgeID%22%3A%22%22%2C%22photosBadgeRelatedUri%22%3A%22%22%7D%2C%22aggregateRating%22%3A2.25%2C%22ingredients%22%3A%5B%222%20tablespoons%20soy%20sauce%22%2C%222%20tablespoons%20olive%20oil%22%2C%221%20tablespoon%20hoisin%20sauce*%22%2C%221%20tablespoon%20unseasoned%20rice%20vinegar%22%2C%223%201%2F2%20ounces%20snow%20peas%20or%20sugar%20snap%20peas%2C%20thinly%20sliced%20lengthwise%20(about%201%201%2F2%20cups)%22%2C%221%20red%20bell%20pepper%2C%20cut%20into%20matchstick-size%20strips%22%2C%221%2017-ounce%20(425-gram)%20package%20thin%20fresh%20Chinese%20egg%20noodles**%22%2C%221%20pound%20smoked%20duck%2C%20turkey%2C%20or%20chicken%2C%20cut%20into%203x1%2F4-inch%20strips%22%2C%221%2F2%20cup%20fresh%20cilantro%20leaves%22%2C%222%20tablespoons%20thinly%20sliced%20green%20onion%22%2C%221%20tablespoon%20thinly%20sliced%20jalape%C3%B1o%20chile%20with%20seeds%20(preferably%20red)%22%5D%2C%22prepSteps%22%3A%5B%22Whisk%20soy%20sauce%2C%20olive%20oil%2C%20hoisin%20sauce%2C%20and%20rice%20vinegar%20in%20large%20bowl%20to%20blend.%20Let%20dressing%20stand%20at%20room%20temperature%20while%20preparing%20noodles.%22%2C%22Place%20sliced%20snow%20peas%20and%20bell%20pepper%20strips%20in%20medium%20bowl.%20Bring%20kettle%20or%20large%20saucepan%20of%20water%20to%20boil.%20Pour%20boiling%20water%20over%20snow%20pea%20mixture%20and%20let%20soak%2010%20seconds.%20Drain%20well.%20Rinse%20mixture%20with%20cold%20water%20and%20drain%20again.%20Place%20egg%20noodles%20in%20large%20bowl.%20Bring%20kettle%20or%20large%20saucepan%20of%20water%20to%20boil%2C%20then%20pour%20boiling%20water%20over%20noodles%3B%20stir%20noodles%20to%20separate.%20Let%20noodles%20soak%2010%20seconds.%20Drain%20well.%20Rinse%20noodles%20with%20cold%20water%3B%20drain%20again.%22%2C%22Add%20snow%20pea%20mixture%2C%20egg%20noodles%2C%20smoked%20duck%2C%20cilantro%20leaves%2C%20green%20onion%2C%20and%20jalape%C3%B1o%20to%20bowl%20with%20dressing.%20Toss%20well%20to%20coat.%20Divide%20noodle%20mixture%20among%20plates%20and%20serve.%22%2C%22*%20Hoisin%20sauce%20is%20available%20in%20the%20Asian%20foods%20section%20of%20many%20supermarkets%20and%20at%20Asian%20markets.%22%2C%22**%20Fresh%20Chinese%20egg%20noodles%20are%20available%20in%20the%20refrigerated%20section%20of%20Asian%20markets.%22%5D%2C%22reviewsCount%22%3A4%2C%22willMakeAgainPct%22%3A50%7D%5D%7D%7D%7D%2C%22options%22%3A%7B%22optimizePromiseCallback%22%3Afalse%7D%2C%22plugins%22%3A%7B%7D%7D%2C%22plugins%22%3A%7B%7D%7D\";\n </script>\n <script async=\"\" src=\"/static/js/dist/search__3.11.1.js\">\n </script>\n <script>\n // Drop meta data about ads on page.\n _app.adInit = {\"channel\":\"search\",\"contentType\":\"search\",\"env\":\"production\",\"slug\":\"%20tofu%20chili\",\"subChannel\":\"\",\"targeting\":[\"search\",\"%252520tofu%252520chili\",\"food\",\"content\"]}\n </script>\n \n <div id=\"parsely-root\" style=\"display: none\">\n <a data-parsely-site=\"epicurious.com\" href=\"https://parsely.com\" id=\"parsely-cfg\">\n Powered by the Parse.ly Publisher Platform (P3).\n </a>\n </div>\n <script type=\"text/javascript\">\n (function(w, d, s) {\n function go() {\n var js,\n fjs = d.getElementsByTagName(s)[0],\n load = function(url, id, opts) {\n if (d.getElementById(id)) {return;}\n js = d.createElement(s);\n js.src = url;\n js.id = id;\n if (opts && opts.divId) {\n document.getElementById(opts.divId).appendChild(js);\n } else {\n fjs.parentNode.insertBefore(js, fjs);\n }\n };\n load('https://d1z2jf7jlzjs58.cloudfront.net/p.js', 'parsely', { divId : 'parsely-root'});\n }\n go();\n }(window, document, 'script'));\n </script>\n \n <script type=\"text/javascript\">\n if (document.querySelectorAll('.embed-instagram').length > 0) {\n EPI.onCompleteActions.push(function () {\n var script = document.createElement('script');\n script.async = true;\n script.defer = true;\n script.src = \"https://platform.instagram.com/en_US/embeds.js\";\n document.getElementsByTagName('head')[0].appendChild(script);\n });\n }\n\n if (document.querySelectorAll('.embed-twitter').length > 0) {\n EPI.onCompleteActions.push(function () {\n var script = document.createElement('script');\n script.async = true;\n script.defer = true;\n script.src = \"https://platform.twitter.com/widgets.js\";\n document.getElementsByTagName('head')[0].appendChild(script);\n });\n }\n\n if (document.querySelectorAll('body.recipe-detail, body.article, body.gallery-page').length) {\n EPI.onCompleteActions.push(function () {\n var script = document.createElement('script');\n script.async = true;\n script.defer = true;\n script.src = \"https://wms.assoc-amazon.com/20070822/US/js/auto-tagger.js?tag=epicuriousfig-20&locale=US&overwrite=Y\";\n document.getElementsByTagName('head')[0].appendChild(script);\n });\n }\n\n if (document.querySelectorAll('.fb-video').length > 0) {\n window.fbAsyncInit = function () {\n FB.init({\n appId : '1636080783276430',\n xfbml : true,\n version : 'v2.10'\n });\n };\n\n EPI.onCompleteActions.push(function () {\n (function (d, s, id) {\n var js, fjs = d.getElementsByTagName(s)[0];\n if (d.getElementById(id)) return;\n js = d.createElement(s); js.id = id;\n js.src = \"https://connect.facebook.net/en_US/sdk.js\";\n fjs.parentNode.insertBefore(js, fjs);\n }(document, 'script', 'facebook-jssdk'));\n });\n }\n </script>\n <script>\n (function (root) {\n/* -- Data -- */\nroot.__INITIAL_STATE__ || (root.__INITIAL_STATE__ = {});\nroot.__INITIAL_STATE__.config = {\"cookieDomain\":\".epicurious.com\",\"endpoints\":{\"contextual\":\"\\u002Fapi\\u002Fdataintelligence\\u002Fv1\\u002Fcontent\\u002Frecommended\",\"rotd\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Frotd\\u002Fcount\\u002F1\"},\"env\":\"PROD\",\"facebook\":{\"appId\":\"1636080783276430\"},\"https\":\"true\",\"media\":{\"bp\":{\"xs\":0,\"s\":600,\"m\":768,\"l\":1024,\"xl\":1360,\"w\":1710}},\"newsletters\":{\"current\":[{\"description\":\"Become a better cook instantly with this weekly report of our ten most helpful tips, tricks, and kitchen secrets. Don't miss it!\",\"id\":\"5\",\"name\":\"The Top Ten\"},{\"description\":\"Love recipes, but hate searching? We do the work for you. You'll get our favorite seasonal recipe plus collections of our exclusive editors' picks.\",\"id\":\"5117\",\"name\":\"Cook This Now\"},{\"description\":\"Get a daily dose of the hottest recipes from Epicurious, Bon Appétit, and other great sites.\",\"id\":\"195169\",\"name\":\"Trending Recipes\"},{\"id\":\"248781\",\"name\":\"Announcements\"},{\"id\":\"248789\",\"name\":\"Cook 90\"},{\"id\":\"248818\",\"name\":\"Diabetes Friendly\"},{\"id\":\"248842\",\"name\":\"Small Plates\"}]},\"server\":\"https:\\u002F\\u002Fwww.epicurious.com\",\"services\":{\"endpoints\":{\"articles\":{\"latest\":\"\\u002Fapi\\u002Fsearch\\u002Fv1\\u002Fquery?size=20&content=article&sort=newest&q=\"},\"branded\":{\"article\":\"\\u002Fapi\\u002Fbranded\\u002Fv1\\u002Farticle\\u002F:brandCode\\u002F:slug\"},\"content\":{\"channel\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Fchannel\\u002F\",\"component\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Fcomponent\",\"contextual\":\"\\u002Fapi\\u002Fdataintelligence\\u002Fv1\\u002Fcontent\\u002Frecommended\",\"featuredIn\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Ffeaturedin\\u002F\",\"homepage\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Fhomepage\",\"latestRecipes\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Fcomponent\\u002Flatestrecipes\",\"memberLookup\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fuser\\u002F\",\"navItems\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Fcomponent\\u002Fnavitems\",\"nutritionalInfo\":\"\\u002Fapi\\u002Fnutritiondata\\u002Fv2\\u002F\",\"reviews\":\"\\u002Fapi\\u002Frecipes\\u002Fv2\\u002F\",\"reviewsFeed\":\"\\u002Fapi\\u002Frecipes\\u002Fv3\\u002F\",\"reviewsSuffix\":\"\\u002Freviews\",\"reviewsFeedSuffix\":\"\\u002Freviews\\u002Ffeed\",\"rotd\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Frotd\\u002Fcount\\u002F1\",\"search\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Fsearch\",\"videoSearchByKeywords\":\"\\u002Fapi\\u002Fvideo\\u002Fv1\\u002Fkeywords\\u002F\"},\"contributors\":{\"details\":\"\\u002Fapi\\u002Fcontributors\\u002Fv1\\u002F:contributor\\u002Fbundle\",\"feed\":\"\\u002Fapi\\u002Fcontributors\\u002Fv1\\u002F:contributor\\u002Ffeed\"},\"cookbooks\":{\"details\":\"\\u002Fapi\\u002Fcookbooks\\u002Fv1\\u002F:slug\",\"index\":\"\\u002Fapi\\u002Fcookbooks\\u002Fv1\"},\"menus\":{\"details\":\"\\u002Fapi\\u002Fusers\\u002Fv1\\u002F:userId\\u002Fmenus\\u002F:menuId\",\"index\":\"\\u002Fapi\\u002Fusers\\u002Fv1\\u002F:userId\\u002Fmenus\"},\"photos\":{\"instagramPosts\":\"\\u002Fapi\\u002Finstagram\\u002Fv1\",\"menuCover\":\"\\u002Fapi\\u002Fphotos\\u002Fv1\\u002Fmenu-cover\",\"searchCard\":\"\\u002Fapi\\u002Fphotos\\u002Fv1\\u002Fsearch-card\",\"searchUtility\":\"\\u002Fapi\\u002Fphotos\\u002Fv1\\u002Fsearch-utility\"},\"recipeBox\":{\"search\":\"\\u002Fapi\\u002Frecipe-box\\u002Fv1\\u002F:userId\\u002Fsearch\\u002F:term\"},\"recipes\":{\"byKeywords\":\"\\u002Fapi\\u002Frecipes\\u002Fv2\\u002Fkeywords\\u002F\",\"byKeywordsForMenus\":\"\\u002Fapi\\u002Fmenus\\u002Fv1\\u002Fkeywords\\u002F\",\"latest\":\"\\u002Fapi\\u002Frecipes\\u002Fv2\\u002Flatest\"},\"search\":{\"details\":\"\\u002Fapi\\u002Frecipes\\u002Fv3\\u002Fdetail\",\"facets\":\"\\u002Fapi\\u002Fsearch\\u002Fv1\\u002Ffacets\",\"query\":\"\\u002Fapi\\u002Fsearch\\u002Fv1\\u002Fquery\",\"suggest\":\"\\u002Fapi\\u002Fsearch\\u002Fv1\\u002Fsuggest\"},\"users\":{\"authenticate\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fauthenticate\",\"info\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fuser\\u002F:uuid\",\"newsletterSubscriptions\":\"\\u002Fapi\\u002Fnewsletters\\u002Fv1\\u002Fsubscription\\u002F:email\",\"recipeBoxSearch\":\"\\u002Fapi\\u002Fusers\\u002Fv1\\u002F:userId\\u002Frecipe-box\\u002Fsearch\\u002F:query\",\"recipes\":\"\\u002Fapi\\u002Fusers\\u002Fv1\\u002F:userId\\u002Frecipes\",\"reviews\":\"\\u002Fapi\\u002Fusers\\u002Fv1\\u002F:userId\\u002Freviews\",\"updateEmail\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fuser\\u002Femail\",\"updatePassword\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fuser\\u002Fpassword\",\"updateUsername\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fuser\\u002Fdisplayname\",\"updateSettings\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fuser\"}}},\"servicesHost\":\"https:\\u002F\\u002Fservices.epicurious.com\",\"simpleReach\":{\"pid\":\"570d6b7d736b79bf1d000d27\"},\"user\":{\"cookies\":{\"keys\":{\"id\":\"amg_user_partner\",\"username\":\"amg_user_info\"},\"names\":[\"amg_user_ext\",\"amg_user_info\",\"amg_user_partner\",\"amg_user_update\",\"cn_uid\"]},\"email\":{\"regExp\":/^(([^<>()[\\]\\.,;:\\s@\\\"]+(\\.[^<>()[\\]\\.,;:\\s@\\\"]+)*)|(\\\".+\\\"))@(([^<>()[\\]\\.,;:\\s@\\\"]+\\.)+[^<>()[\\]\\.,;:\\s@\\\"]{2,})$/i},\"password\":{\"regExp\":/^[a-zA-Z0-9!\"\\#$%&'()*+,\\-.\\/:;<=>?@\\[\\\\\\]^_`{|}~][a-zA-Z0-9 !\"\\#$%&'()*+,\\-.\\/:;<=>?@\\[\\\\\\]^_`{|}~]{4,253}[a-zA-Z0-9!\"\\#$%&'()*+,\\-.\\/:;<=>?@\\[\\\\\\]^_`{|}~]$/,\"messages\":{\"requirements\":\"A password must be at least 6 characters long. It cannot begin or end with a space.\"}},\"serviceKey\":\"NtibqP3y1qSJM\\u002FGsy3blJgNWt\\u002Fo=\",\"serviceHost\":\"https:\\u002F\\u002Fuser-service.condenastdigital.com\"},\"userServiceHost\":\"https:\\u002F\\u002Fuser-service.condenastdigital.com\",\"userServiceKey\":\"NtibqP3y1qSJM\\u002FGsy3blJgNWt\\u002Fo=\",\"vulcan\":{\"host\":\"https:\\u002F\\u002Fassets.epicurious.com\",\"path\":\"\\u002Fphotos\\u002F\"}};\nroot.__INITIAL_STATE__.siteHeader = {\"config\":{\"cookieDomain\":\".epicurious.com\",\"endpoints\":{\"contextual\":\"\\u002Fapi\\u002Fdataintelligence\\u002Fv1\\u002Fcontent\\u002Frecommended\",\"rotd\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Frotd\\u002Fcount\\u002F1\"},\"env\":\"PROD\",\"facebook\":{\"appId\":\"1636080783276430\"},\"https\":\"true\",\"media\":{\"bp\":{\"xs\":0,\"s\":600,\"m\":768,\"l\":1024,\"xl\":1360,\"w\":1710}},\"newsletters\":{\"current\":[{\"description\":\"Become a better cook instantly with this weekly report of our ten most helpful tips, tricks, and kitchen secrets. Don't miss it!\",\"id\":\"5\",\"name\":\"The Top Ten\"},{\"description\":\"Love recipes, but hate searching? We do the work for you. You'll get our favorite seasonal recipe plus collections of our exclusive editors' picks.\",\"id\":\"5117\",\"name\":\"Cook This Now\"},{\"description\":\"Get a daily dose of the hottest recipes from Epicurious, Bon Appétit, and other great sites.\",\"id\":\"195169\",\"name\":\"Trending Recipes\"},{\"id\":\"248781\",\"name\":\"Announcements\"},{\"id\":\"248789\",\"name\":\"Cook 90\"},{\"id\":\"248818\",\"name\":\"Diabetes Friendly\"},{\"id\":\"248842\",\"name\":\"Small Plates\"}]},\"server\":\"https:\\u002F\\u002Fwww.epicurious.com\",\"services\":{\"endpoints\":{\"articles\":{\"latest\":\"\\u002Fapi\\u002Fsearch\\u002Fv1\\u002Fquery?size=20&content=article&sort=newest&q=\"},\"branded\":{\"article\":\"\\u002Fapi\\u002Fbranded\\u002Fv1\\u002Farticle\\u002F:brandCode\\u002F:slug\"},\"content\":{\"channel\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Fchannel\\u002F\",\"component\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Fcomponent\",\"contextual\":\"\\u002Fapi\\u002Fdataintelligence\\u002Fv1\\u002Fcontent\\u002Frecommended\",\"featuredIn\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Ffeaturedin\\u002F\",\"homepage\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Fhomepage\",\"latestRecipes\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Fcomponent\\u002Flatestrecipes\",\"memberLookup\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fuser\\u002F\",\"navItems\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Fcomponent\\u002Fnavitems\",\"nutritionalInfo\":\"\\u002Fapi\\u002Fnutritiondata\\u002Fv2\\u002F\",\"reviews\":\"\\u002Fapi\\u002Frecipes\\u002Fv2\\u002F\",\"reviewsFeed\":\"\\u002Fapi\\u002Frecipes\\u002Fv3\\u002F\",\"reviewsSuffix\":\"\\u002Freviews\",\"reviewsFeedSuffix\":\"\\u002Freviews\\u002Ffeed\",\"rotd\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Frotd\\u002Fcount\\u002F1\",\"search\":\"\\u002Fapi\\u002Fcontent\\u002Fv1\\u002Fsearch\",\"videoSearchByKeywords\":\"\\u002Fapi\\u002Fvideo\\u002Fv1\\u002Fkeywords\\u002F\"},\"contributors\":{\"details\":\"\\u002Fapi\\u002Fcontributors\\u002Fv1\\u002F:contributor\\u002Fbundle\",\"feed\":\"\\u002Fapi\\u002Fcontributors\\u002Fv1\\u002F:contributor\\u002Ffeed\"},\"cookbooks\":{\"details\":\"\\u002Fapi\\u002Fcookbooks\\u002Fv1\\u002F:slug\",\"index\":\"\\u002Fapi\\u002Fcookbooks\\u002Fv1\"},\"menus\":{\"details\":\"\\u002Fapi\\u002Fusers\\u002Fv1\\u002F:userId\\u002Fmenus\\u002F:menuId\",\"index\":\"\\u002Fapi\\u002Fusers\\u002Fv1\\u002F:userId\\u002Fmenus\"},\"photos\":{\"instagramPosts\":\"\\u002Fapi\\u002Finstagram\\u002Fv1\",\"menuCover\":\"\\u002Fapi\\u002Fphotos\\u002Fv1\\u002Fmenu-cover\",\"searchCard\":\"\\u002Fapi\\u002Fphotos\\u002Fv1\\u002Fsearch-card\",\"searchUtility\":\"\\u002Fapi\\u002Fphotos\\u002Fv1\\u002Fsearch-utility\"},\"recipeBox\":{\"search\":\"\\u002Fapi\\u002Frecipe-box\\u002Fv1\\u002F:userId\\u002Fsearch\\u002F:term\"},\"recipes\":{\"byKeywords\":\"\\u002Fapi\\u002Frecipes\\u002Fv2\\u002Fkeywords\\u002F\",\"byKeywordsForMenus\":\"\\u002Fapi\\u002Fmenus\\u002Fv1\\u002Fkeywords\\u002F\",\"latest\":\"\\u002Fapi\\u002Frecipes\\u002Fv2\\u002Flatest\"},\"search\":{\"details\":\"\\u002Fapi\\u002Frecipes\\u002Fv3\\u002Fdetail\",\"facets\":\"\\u002Fapi\\u002Fsearch\\u002Fv1\\u002Ffacets\",\"query\":\"\\u002Fapi\\u002Fsearch\\u002Fv1\\u002Fquery\",\"suggest\":\"\\u002Fapi\\u002Fsearch\\u002Fv1\\u002Fsuggest\"},\"users\":{\"authenticate\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fauthenticate\",\"info\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fuser\\u002F:uuid\",\"newsletterSubscriptions\":\"\\u002Fapi\\u002Fnewsletters\\u002Fv1\\u002Fsubscription\\u002F:email\",\"recipeBoxSearch\":\"\\u002Fapi\\u002Fusers\\u002Fv1\\u002F:userId\\u002Frecipe-box\\u002Fsearch\\u002F:query\",\"recipes\":\"\\u002Fapi\\u002Fusers\\u002Fv1\\u002F:userId\\u002Frecipes\",\"reviews\":\"\\u002Fapi\\u002Fusers\\u002Fv1\\u002F:userId\\u002Freviews\",\"updateEmail\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fuser\\u002Femail\",\"updatePassword\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fuser\\u002Fpassword\",\"updateUsername\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fuser\\u002Fdisplayname\",\"updateSettings\":\"\\u002Fapi\\u002Fusers\\u002Fv2\\u002Fuser\"}}},\"servicesHost\":\"https:\\u002F\\u002Fservices.epicurious.com\",\"simpleReach\":{\"pid\":\"570d6b7d736b79bf1d000d27\"},\"user\":{\"cookies\":{\"keys\":{\"id\":\"amg_user_partner\",\"username\":\"amg_user_info\"},\"names\":[\"amg_user_ext\",\"amg_user_info\",\"amg_user_partner\",\"amg_user_update\",\"cn_uid\"]},\"email\":{\"regExp\":/^(([^<>()[\\]\\.,;:\\s@\\\"]+(\\.[^<>()[\\]\\.,;:\\s@\\\"]+)*)|(\\\".+\\\"))@(([^<>()[\\]\\.,;:\\s@\\\"]+\\.)+[^<>()[\\]\\.,;:\\s@\\\"]{2,})$/i},\"password\":{\"regExp\":/^[a-zA-Z0-9!\"\\#$%&'()*+,\\-.\\/:;<=>?@\\[\\\\\\]^_`{|}~][a-zA-Z0-9 !\"\\#$%&'()*+,\\-.\\/:;<=>?@\\[\\\\\\]^_`{|}~]{4,253}[a-zA-Z0-9!\"\\#$%&'()*+,\\-.\\/:;<=>?@\\[\\\\\\]^_`{|}~]$/,\"messages\":{\"requirements\":\"A password must be at least 6 characters long. It cannot begin or end with a space.\"}},\"serviceKey\":\"NtibqP3y1qSJM\\u002FGsy3blJgNWt\\u002Fo=\",\"serviceHost\":\"https:\\u002F\\u002Fuser-service.condenastdigital.com\"},\"userServiceHost\":\"https:\\u002F\\u002Fuser-service.condenastdigital.com\",\"userServiceKey\":\"NtibqP3y1qSJM\\u002FGsy3blJgNWt\\u002Fo=\",\"vulcan\":{\"host\":\"https:\\u002F\\u002Fassets.epicurious.com\",\"path\":\"\\u002Fphotos\\u002F\"}}};\nroot.__INITIAL_STATE__.contentType = \"search\";\n}(this));\n\n window.copilot = window.copilot || window.__INITIAL_STATE__.copilotData;\n </script>\n <link as=\"style\" href=\"/static/css/patch.css\" onload=\"this.rel='stylesheet'\" rel=\"preload\"/>\n <noscript>\n <link href=\"/static/css/patch.css\" rel=\"stylesheet\"/>\n </noscript>\n <span data-react-checksum=\"-664854917\" data-reactid=\"1\" data-reactroot=\"\" id=\"boomerang-beacon\">\n \n </span>\n </body>\n</html>\n\n" ] ], [ [ "<a id='step2'></a>\n<h3>BS4 functions</h3>", "_____no_output_____" ], [ "<a id='model1'></a>\n<h4>find_all</h4> finds all instances of a specified tag returns a **result_set (a list)**", "_____no_output_____" ] ], [ [ "all_a_tags = results_page.find_all('a')\nprint(type(all_a_tags))", "<class 'bs4.element.ResultSet'>\n" ], [ "all_a_tags", "_____no_output_____" ] ], [ [ "<a id='model2'></a>\n<h4>find</h4> finds the <u>first</u> instance of a specified tag and returns a **bs4 element**", "_____no_output_____" ] ], [ [ "div_tag = results_page.find('div')\nprint(div_tag)", "<div class=\"header-wrapper\" data-reactid=\"2\"><div class=\"header\" data-reactid=\"3\" role=\"banner\"><h2 data-reactid=\"4\" itemtype=\"https://schema.org/Organization\"><a data-reactid=\"5\" href=\"/\" itemprop=\"url\" title=\"Epicurious\">Epicurious</a></h2><div class=\"search-form-container\" data-reactid=\"6\"><form action=\"/search/\" autocomplete=\"off\" data-reactid=\"7\" method=\"get\" role=\"search\"><fieldset data-reactid=\"8\"><button class=\"submit\" data-reactid=\"9\" type=\"submit\">search</button><input autocomplete=\"off\" data-reactid=\"10\" maxlength=\"120\" name=\"terms\" placeholder=\"Find a Recipe\" type=\"text\" value=\" tofu chili\"/><button class=\"filter mobile\" data-reactid=\"11\">filters</button><button class=\"filter tablet\" data-reactid=\"12\">filter results</button></fieldset></form><div class=\"ingredient-filters\" data-reactid=\"13\"><h3 data-reactid=\"14\">Include/Exclude Ingredients</h3><form class=\"include-ingredients\" data-reactid=\"15\"><fieldset data-reactid=\"16\"><input aria-label=\"Include ingredients\" data-reactid=\"17\" placeholder=\"Include ingredients:\" type=\"text\"/><button data-reactid=\"18\">include</button></fieldset></form><form class=\"exclude-ingredients\" data-reactid=\"19\"><fieldset data-reactid=\"20\"><input aria-label=\"Exclude ingredients\" data-reactid=\"21\" placeholder=\"Exclude ingredients:\" type=\"text\"/><button data-reactid=\"22\">exclude</button></fieldset></form></div><button class=\"filter ingredient-filter\" data-reactid=\"23\">include/exclude ingredients</button><div class=\"search-tags\" data-reactid=\"24\"><button class=\"clear-all\" data-reactid=\"25\">Clear all</button></div></div></div><div class=\"filters\" data-reactid=\"26\"><div data-reactid=\"27\"><div class=\"filter-action-panel\" data-reactid=\"29\"><p data-reactid=\"30\">18 matching results</p><button data-reactid=\"33\">Apply</button><button data-reactid=\"34\">Cancel</button></div></div></div></div>\n" ], [ "type(div_tag)\n", "_____no_output_____" ], [ "type(results_page)", "_____no_output_____" ], [ "print(div_tag.prettify())", "<div class=\"header-wrapper\" data-reactid=\"2\">\n <div class=\"header\" data-reactid=\"3\" role=\"banner\">\n <h2 data-reactid=\"4\" itemtype=\"https://schema.org/Organization\">\n <a data-reactid=\"5\" href=\"/\" itemprop=\"url\" title=\"Epicurious\">\n Epicurious\n </a>\n </h2>\n <div class=\"search-form-container\" data-reactid=\"6\">\n <form action=\"/search/\" autocomplete=\"off\" data-reactid=\"7\" method=\"get\" role=\"search\">\n <fieldset data-reactid=\"8\">\n <button class=\"submit\" data-reactid=\"9\" type=\"submit\">\n search\n </button>\n <input autocomplete=\"off\" data-reactid=\"10\" maxlength=\"120\" name=\"terms\" placeholder=\"Find a Recipe\" type=\"text\" value=\" tofu chili\"/>\n <button class=\"filter mobile\" data-reactid=\"11\">\n filters\n </button>\n <button class=\"filter tablet\" data-reactid=\"12\">\n filter results\n </button>\n </fieldset>\n </form>\n <div class=\"ingredient-filters\" data-reactid=\"13\">\n <h3 data-reactid=\"14\">\n Include/Exclude Ingredients\n </h3>\n <form class=\"include-ingredients\" data-reactid=\"15\">\n <fieldset data-reactid=\"16\">\n <input aria-label=\"Include ingredients\" data-reactid=\"17\" placeholder=\"Include ingredients:\" type=\"text\"/>\n <button data-reactid=\"18\">\n include\n </button>\n </fieldset>\n </form>\n <form class=\"exclude-ingredients\" data-reactid=\"19\">\n <fieldset data-reactid=\"20\">\n <input aria-label=\"Exclude ingredients\" data-reactid=\"21\" placeholder=\"Exclude ingredients:\" type=\"text\"/>\n <button data-reactid=\"22\">\n exclude\n </button>\n </fieldset>\n </form>\n </div>\n <button class=\"filter ingredient-filter\" data-reactid=\"23\">\n include/exclude ingredients\n </button>\n <div class=\"search-tags\" data-reactid=\"24\">\n <button class=\"clear-all\" data-reactid=\"25\">\n Clear all\n </button>\n </div>\n </div>\n </div>\n <div class=\"filters\" data-reactid=\"26\">\n <div data-reactid=\"27\">\n \n <div class=\"filter-action-panel\" data-reactid=\"29\">\n <p data-reactid=\"30\">\n \n 18\n \n \n matching results\n \n </p>\n <button data-reactid=\"33\">\n Apply\n </button>\n <button data-reactid=\"34\">\n Cancel\n </button>\n </div>\n </div>\n </div>\n</div>\n\n" ], [ "div_tag.find('div').find('div').find('div')", "_____no_output_____" ] ], [ [ "<h4>bs4 functions can be recursively applied on elements</h4>", "_____no_output_____" ] ], [ [ "div_tag.find('a')", "_____no_output_____" ] ], [ [ "<a id='model3'></a>\nBoth __find__ as well as __find_all__ can be qualified by css selectors\n<li>using selector=value\n<li>using a dictionary", "_____no_output_____" ] ], [ [ "#When using this method and looking for 'class' use 'class_' (because class is a reserved word in python)\n#Note that we get a list back because find_all returns a list\nresults_page.find_all('article',class_=\"recipe-content-card\")", "_____no_output_____" ], [ "#Since we're using a string as the key, the fact that class is a reserved word is not a problem\n#We get an element back because find returns an element\nresults_page.find('article',{'class':'recipe-content-card'})", "_____no_output_____" ] ], [ [ "<a id='model4'></a>\n__get_text()__ returns the marked up text (the content) enclosed in a tag.\n<li>returns a string", "_____no_output_____" ] ], [ [ "results_page.find('article',{'class':'recipe-content-card'}).get_text()", "_____no_output_____" ], [ "print(results_page.find('article',{'class':'recipe-content-card'}).get_text())", "recipeSpicy Lemongrass TofuDau hu xa ot\nEditor's note: The recipe and introductory text below are excerpted from Pleasures of the Vietnamese Table by Mai Pham and are part of our story on Lunar New Year.\nWhile traveling on a train one time to the coastal town of Nha Trang, I sat next to an elderly nun. Over the course of our bumpy eight-hour ride, she shared stories of life at the temple and the difficult years after the end of the war when the Communist government cracked down on religious factions. Toward the end of our chat, she pulled out a bag of food she'd prepared for the trip. It was tofu that had been cooked in chilies, lemongrass and la lot, an aromatic leaf also known as pepper leaf. When she gave me a taste, I knew immediately that I had to learn how to make it. This is my rendition of that fabulous dish. Make sure to pat the tofu dry before marinating it and use very fresh lemongrass. I always love serving this to friends who think tofu dishes are bland.Average user rating3.5/4Reviews17Percentage of reviewers who will make this recipe again88%View “Spicy Lemongrass Tofu”View RecipeQuick viewCompare Recipe\n" ] ], [ [ "<a id='model5'></a>\n__get()__ returns the __value__ of a tag attribute\n<li>returns a string", "_____no_output_____" ] ], [ [ "recipe_tag = results_page.find('article',{'class':'recipe-content-card'})\nrecipe_link = recipe_tag.find('a')\nprint(\"a tag:\",recipe_link)\nlink_url = recipe_link.get('href')\nprint(\"link url:\",link_url)\nprint(type(link_url))", "a tag: <a data-reactid=\"72\" href=\"/recipes/food/views/spicy-lemongrass-tofu-233844\">Spicy Lemongrass Tofu</a>\nlink url: /recipes/food/views/spicy-lemongrass-tofu-233844\n<class 'str'>\n" ] ], [ [ "<a id='step3'></a>\n<h1>A function that returns a list containing recipe names, recipe descriptions (if any) and recipe urls</h1>", "_____no_output_____" ] ], [ [ "def get_recipes(keywords):\n recipe_list = list()\n import requests\n from bs4 import BeautifulSoup\n url = \"http://www.epicurious.com/search/\" + keywords\n response = requests.get(url)\n if not response.status_code == 200:\n return None\n try:\n results_page = BeautifulSoup(response.content,'lxml')\n recipes = results_page.find_all('article',class_=\"recipe-content-card\")\n for recipe in recipes:\n# recipe_link = \"http://www.epicurious.com\" + recipe.find('a').get('href')\n# recipe_name = recipe.find('a').get_text()\n# try:\n# recipe_description = recipe.find('p',class_='dek').get_text()\n# except:\n# recipe_description = ''\n# recipe_list.append((recipe_name,recipe_link,recipe_description))\n recipe_list.append(recipes)\n return recipe_list\n except:\n return None", "_____no_output_____" ], [ "get_recipes(\"Tofu chili\")", "_____no_output_____" ] ], [ [ "And we can see that that's sitting inside a __paragraph__\nhere with class equals __dek__.\nSo we can always get the description\nby looking for a __paragraph tag__ with the __class = dek__.\nWe also see-- and we saw this before--\nthat the link to the next recipe,\nto the recipe detail page, is inside an __annotate tag__.\nSo there's an __a-tag__ over here.\nAnd that contains the link and it also\ncontains the name of the recipe.\n\nSo with these two things, by finding that annotate tag,\nthe first annotate tag in our recipe content card article,\nand the first paragraph tag that has a class equals dek,\nwe can get the _name, the link, and the description_.\nSo let's add these three things to our setup here.", "_____no_output_____" ] ], [ [ "def get_recipes(keywords):\n recipe_list = list()\n import requests\n from bs4 import BeautifulSoup\n url = \"http://www.epicurious.com/search/\" + keywords\n response = requests.get(url)\n if not response.status_code == 200:\n return None\n try:\n results_page = BeautifulSoup(response.content,'lxml')\n recipes = results_page.find_all('article',class_=\"recipe-content-card\")\n for recipe in recipes:\n recipe_link = \"http://www.epicurious.com\" + recipe.find('a').get('href')\n recipe_name = recipe.find('a').get_text()\n try:\n recipe_description = recipe.find('p',class_='dek').get_text()\n except:\n recipe_description = ''\n recipe_list.append((recipe_name,recipe_link,recipe_description))\n return recipe_list\n except:\n return None", "_____no_output_____" ], [ "get_recipes(\"Tofu chili\")", "_____no_output_____" ], [ "get_recipes('Nothing')", "_____no_output_____" ] ], [ [ "<a id='model6'></a>\n<h2>Let's write a new function!</h2>\n\nGiven a recipe link returns a dictionary containing the ingredients and preparation instructions", "_____no_output_____" ] ], [ [ "recipe_link = \"http://www.epicurious.com\" + '/recipes/food/views/spicy-lemongrass-tofu-233844'", "_____no_output_____" ], [ "def get_recipe_info(recipe_link):\n recipe_dict = dict()\n import requests\n from bs4 import BeautifulSoup\n try:\n response = requests.get(recipe_link)\n if not response.status_code == 200:\n return recipe_dict\n result_page = BeautifulSoup(response.content,'lxml')\n ingredient_list = list()\n prep_steps_list = list()\n for ingredient in result_page.find_all('li',class_='ingredient'):\n ingredient_list.append(ingredient.get_text())\n for prep_step in result_page.find_all('li',class_='preparation-step'):\n prep_steps_list.append(prep_step.get_text().strip())\n recipe_dict['ingredients'] = ingredient_list\n recipe_dict['preparation'] = prep_steps_list\n return recipe_dict\n except:\n return recipe_dict\n ", "_____no_output_____" ], [ "get_recipe_info(recipe_link)", "_____no_output_____" ] ], [ [ "<a id='model7'></a>\n<h2>Construct a list of dictionaries for all recipes</h2>", "_____no_output_____" ] ], [ [ "def get_all_recipes(keywords):\n results = list()\n all_recipes = get_recipes(keywords)\n for recipe in all_recipes:\n recipe_dict = get_recipe_info(recipe[1])\n recipe_dict['name'] = recipe[0]\n recipe_dict['description'] = recipe[2]\n results.append(recipe_dict)\n return(results)", "_____no_output_____" ], [ "get_all_recipes(\"Tofu chili\")", "_____no_output_____" ] ], [ [ "<a id='step4'></a>\n<h1>Logging in to a web server</h1>", "_____no_output_____" ], [ "<a id='model8'></a>\n<h2>Get username and password</h2>\n<li>Best to store in a file for reuse\n<li>You will need to set up your own login and password and place them in a file called wikidata.txt\n<li>Line one of the file should contain your username\n<li>Line two your password", "_____no_output_____" ] ], [ [ "with open('wikidata.txt') as f:\n contents = f.read().split('\\n')\n username = contents[0]\n password = contents[1]\nprint(username,password)", "VSerpak NewPeriod\n" ] ], [ [ "<a id='model9'></a>\n<h3>Construct an object that contains the data to be sent to the login page</h3>", "_____no_output_____" ] ], [ [ "\npayload = {\n 'wpName': username,\n 'wpPassword': password,\n 'wploginattempt': 'Log in',\n 'wpEditToken': \"+\\\\\",\n 'title': \"Special:UserLogin\",\n 'authAction': \"login\",\n 'force': \"\",\n 'wpForceHttps': \"1\",\n 'wpFromhttp': \"1\",\n #'wpLoginToken': ‘', #We need to read this from the page\n }", "_____no_output_____" ] ], [ [ "<a id='model10'></a>\n<h3>Get the value of the login token</h3>", "_____no_output_____" ] ], [ [ "def get_login_token(response):\n soup = BeautifulSoup(response.text, 'lxml')\n token = soup.find('input',{'name':\"wpLoginToken\"}).get('value')\n return token\n", "_____no_output_____" ] ], [ [ "<a id='model11'></a>\n<h3>Setup a session, login, and get data</h3>", "_____no_output_____" ] ], [ [ "with requests.session() as s:\n response = s.get('https://en.wikipedia.org/w/index.php?title=Special:UserLogin&returnto=Main+Page')\n payload['wpLoginToken'] = get_login_token(response)\n #Send the login request\n response_post = s.post('https://en.wikipedia.org/w/index.php?title=Special:UserLogin&action=submitlogin&type=login',\n data=payload)\n #Get another page and check if we’re still logged in\n response = s.get('https://en.wikipedia.org/wiki/Special:Watchlist')\n data = BeautifulSoup(response.content,'lxml')\n print(data.find('div',class_='mw-changeslist').get_text())", "25 February 2018\n(User creation log); 14:50 . . User account VSerpak (talk | contribs) was created ‎\n\n\n" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

ec7d16769407ebef327fd9df7391a15810a2c831

20,155

ipynb

Jupyter Notebook

PlanPrepareProcess_OT2/Plan and Prepare/Dev_Working_Notebook.ipynb

pozzo-research-group/OT2Protocols2

2386dae2b7f18a8a42fb8b4a0b8d2c6b2f3ff440

[ "MIT" ]

2

2020-06-01T16:32:43.000Z

2021-12-01T16:57:36.000Z

PlanPrepareProcess_OT2/Plan and Prepare/Dev_Working_Notebook.ipynb

pozzo-research-group/OT2Protocols2

2386dae2b7f18a8a42fb8b4a0b8d2c6b2f3ff440

[ "MIT" ]

3

2020-01-22T02:06:31.000Z

2020-07-19T18:58:51.000Z

PlanPrepareProcess_OT2/Plan and Prepare/Dev_Working_Notebook.ipynb

pozzo-research-group/OT2Protocols2

2386dae2b7f18a8a42fb8b4a0b8d2c6b2f3ff440

[ "MIT" ]

null

37.462825

570

0.548995

[ [ [ "# This notebook serves as walkthrough for planning an experiment for creation through the OT2.\n### The following modules are used and should be in the same directory as this notebook: \n* **CreateSamples** is responsible for calculating sample information, which includes component weight fractions and stock volumes\n* **OT2Commands** is responsible for setting up information to be interpretted and executed by opentrons.\n* **OT2Graphing** contains graphing tools to help visualize and explore parameter spaces.", "_____no_output_____" ] ], [ [ "import CreateSamples\nimport OT2Commands as ALH\nimport OT2Graphing as ographing\nfrom opentrons import simulate, execute, protocol_api\n\n# Would not load\nimport importlib # for reloading packages\nimport pandas as pd\nimport matplotlib.pyplot as plt", "_____no_output_____" ], [ "importlib.reload(CreateSamples)\nimportlib.reload(ALH)\nimportlib.reload(ographing)", "_____no_output_____" ] ], [ [ "## Step 1: Set up the experiment dictionary.\n* The first step to planning an experiment is to load the experiment variables and inputs from a csv file. Every variable should have an input with an acceptable datatype. At the moment this step is done by opening a CSV file in Excel, where the first column is the name of the variable and the adjacent column is the variable value. The default delimitter is (,). \n * Reading directly as csv is fine but it requires you have all data values in a string so then we can use ast.literal_eval to unpack this and the appropiate datatypes. This forces you \\to put marks ('') around each variable value when planning the experiment. NOTE: You still need to place marks around anything inside a dtype i.e components inside a list.\n * To remove this dependency we can build our own interpreters for our specfic cases such as to not use ast.literal_evals default unpacking.\n * Loading from excel can be done in a similar manner but is avoided due not having xlrd or openpyxl depdenency native to python, and the opentrons being limited in the packages we can add/update. Hence we default to a CSV.\n* **The experiment dictionary consist of keys being the variable name and the value being the variables value.**", "_____no_output_____" ] ], [ [ "path = r\"C:\\Users\\Edwin\\Desktop\\OT2Protocols\\ot2protocol\\Ouzo_OT2_Sampling\\Testing Plans\\Example_Working_Protocol.csv\"\nexperiment_csv_dict = CreateSamples.get_experiment_plan(path) ", "_____no_output_____" ] ], [ [ "## Step 1a Optional: Load custom labware dictionary (Remote Testing)\n* Provide the path to the directory holding all custom labware. This directory should have custom labware .json files you have previously made and tested, read more here: https://support.opentrons.com/en/articles/3136504-creating-custom-labware-definitions\n* The reason we provide this is when working on a device that is not connected to the OT2's Jupyter notebook there is no way to natively use custom labware. So we create a dictionary of custom labware so we can later load into our protocol to primarily simualte and test protocols for execution later once connnected to the OT2's notebook.\n * When using custom labware on the OT2's notebook it pulls from a folder labeled \"labware\", which is something built into the Opentrons hardware. It has not been tested if the custom labware dictionary will superceed this directory of custom labware if used on the OT2.", "_____no_output_____" ] ], [ [ "labware_dir_path = r\"C:\\Users\\Edwin\\Desktop\\OT2Protocols\\ot2protocol\\Ouzo_OT2_Sampling\\Custom Labware\"\ncustom_labware_dict = ALH.custom_labware_dict(labware_dir_path)", "_____no_output_____" ] ], [ [ "## Step 2: Select and Create Sampling Space\n* Create sampling space depending on the units of concentration and method of sampling. All information is pulled from the experimental dictionary made in Step 1.\n * Currently the only sampling method available are simple lattice and random based sampling. There are two potential ways to create samples in a system of n components which currently utilzie the linspaces of concentration. \n * *Remember the linspace of concentration refers to [minimum concentration, maximum concentration, concentration step size]*\n * **Case 1 (Completing case):** Specify all but one (in this case the last) component's concentrations, which with the addition of exposing the unity_filter = True, would calculate the the remaining concentraiton values using the information of the last index of all component related variables (i.e. names). This is only meant for units that require unity like volf, wtf, and molf. \n * **Case 2 (Non-completing case):** Specify all concentration linspaces, not applying any unit based filters, meant for all other non interdepedent units like molarity and mg/mL.\n* Other things to take into consideration: All units must be the same for unity based, but not for non-unity units.", "_____no_output_____" ] ], [ [ "wtf_sample_canidates = CreateSamples.generate_candidate_lattice_concentrations(experiment_csv_dict, unity_filter=True)", "_____no_output_____" ] ], [ [ "## Step 3: Calculate Volumes of Stocks\n* From the concentration values calculated in Step 2, we use those along with stock concentration information to calculate the volume of required for each sample.\n\n* This is where things get less \"*general*\" each case depending on the number of stocks, common components (i.e. component A in both stock A and B) and other requirements will typically require its own function. Luckily given the commmonality of using data frames this should be quite simple. \n* Currently the only function to calculate volumes in centered around the Ouzo emulsion systems. This system consist of 3 stock with the solvent being ethanol and two pure stocks of ethanol and water. \n\n* Ideally the way volumes should calculated is simply by calculating \"*essential information*\" given the concentration of component and the systems overall mass or volume. Using this essential information and the concentration unit of the stock, it should call the appropiate function to calcualte the volume. Many issue could arise such as having a molarity and providing a mass so would need to make sure these cases are sorted and reported back\n", "_____no_output_____" ] ], [ [ "volume_sample_canidates = CreateSamples.calculate_ouzo_volumes_from_wtf(wtf_sample_canidates, experiment_csv_dict)", "_____no_output_____" ] ], [ [ "## Step 4: Create Complete Component Cocentration and Volume Dataframe and Apply Filters", "_____no_output_____" ] ], [ [ "complete_df = CreateSamples.combine_df(wtf_sample_canidates, volume_sample_canidates) # unfiltered\ncomplete_df = pd.concat([complete_df]*48, ignore_index=True)\n# Step 3: Apply filters through df based logic, currently 4 filters exist (volume, total, general min and general max)\n\n# First filter for pipette volume constraints, optional Volume Restriction to select certain components for filter application (\"stock\" must be in column name)\ncomplete_df_f1 = CreateSamples.pipette_volume_restriction_df(complete_df, 30, 1000, experiment_csv_dict['Volume Restriction']) # last argument is optional\n\n# Second filter for overall total volume_restriction, call max destination well volume (\"Total Sample Volume\" must be in column name)\n# max_dest_well_volume = ALH.find_max_dest_volume_labware(experiment_csv_dict, custom_labware_dict)\ncomplete_df_f2 = CreateSamples.total_volume_restriction_df(complete_df_f1,1400)\n\n#Thrid filter for any general max or min filtering you would like\nfinal_complete_df = complete_df_f2#CreateSamples.general_max_restriction(complete_df_f2, 360, 'pfh-ethanol-stock uL')", "_____no_output_____" ] ], [ [ "## Step 4a (Optional): Visual", "_____no_output_____" ] ], [ [ "# ographing.xy_scatter_df_compare(complete_df, final_complete_df, 'ethanol molf', 'pfh molf')\n# ographing.xy_scatter_df(final_complete_df, 'ethanol wtf', 'pfh molf')", "_____no_output_____" ] ], [ [ "## Step 5: Finalize and Call Seperate Concentration and Volume Dataframes", "_____no_output_____" ] ], [ [ "final_wtf_df = CreateSamples.isolate_common_column(final_complete_df, 'wtf')\nfinal_volume_df = CreateSamples.isolate_common_column(final_complete_df, 'stock')\nfinal_wtf_df", "_____no_output_____" ] ], [ [ "## Step 6 (Optional): Calculate Stock Prep Information", "_____no_output_____" ] ], [ [ "chem_database_path = r\"C:\\Users\\Edwin\\Desktop\\OT2Protocols\\ot2protocol\\Ouzo_OT2_Sampling\\Chemical Database.xlsx\"\nstock_prep_df = CreateSamples.calculate_stock_prep_df(experiment_csv_dict, final_volume_df, chem_database_path)\n# pd.set_option('display.float_format', lambda x: '%.2e' % x)\nstock_prep_df", "_____no_output_____" ] ], [ [ "## Step 7: Set up Ranges for Stocks", "_____no_output_____" ] ], [ [ "protocol = simulate.get_protocol_api('2.0', extra_labware=custom_labware_dict)\nmax_vol = 20000 \nstock_ranges = ALH.stock_well_ranges(final_volume_df, max_vol) # set up volumes orders\nstock_ranges", "_____no_output_____" ], [ "# Step 8: Simulate/Execute", "_____no_output_____" ], [ "protocol = simulate.get_protocol_api('2.0', extra_labware=custom_labware_dict)\nloaded_dict = ALH.loading_labware(protocol, experiment_csv_dict) # the protocol above has been modified globally!\ninfo = ALH.pipette_stock_volumes(protocol, loaded_dict, final_volume_df, stock_ranges)", "_____no_output_____" ], [ "## Step 9: Uploaded to Google Drive", "_____no_output_____" ], [ "CreateSamples.create_csv(r\"C:\\Users\\Edwin\\Desktop\\test\", info['info concat'], final_wtf_df.values, experiment_csv_dict)\ndf = pd.read_csv(r\"C:\\Users\\Edwin\\Desktop\\test\")\n", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code" ] ]

ec7d1afd45e264523fe4d3b428b2a9a318836f9e

2,544

ipynb

Jupyter Notebook

bullshitgenerator-english.ipynb

hao-hao-hao/BullshitGenerator-English

9dd5473c52c5e7c60ded92fd1e300648dcea2dad

[ "MIT" ]

null

bullshitgenerator-english.ipynb

hao-hao-hao/BullshitGenerator-English

9dd5473c52c5e7c60ded92fd1e300648dcea2dad

[ "MIT" ]

null

bullshitgenerator-english.ipynb

hao-hao-hao/BullshitGenerator-English

9dd5473c52c5e7c60ded92fd1e300648dcea2dad

[ "MIT" ]

null

1,272

2,543

0.60967

[ [ [ "# from https://github.com/JIUYANGZH/BullshitGenerator-English\nimport pandas as pd\nimport random\n\ndf = pd.read_csv(\"https://raw.githubusercontent.com/hao-hao-hao/BullshitGenerator-English/master/bullshitgenerator_English/bullshit_generator.csv\")\n\ndef get_lists(df,phrase):\n return df[df[phrase].notnull()][phrase].tolist()\n\n[bullshits,prefix_1,addings,examples,contrasts,prefix_2,suffix,author,saying] = [get_lists(df,i) for i in df.columns]\n\n\ndef sayings():\n xx = ''\n index= random.choice(range(len(saying)))\n if random.random() > 0.3:\n xx = author[index] + ' ' + random.choice(prefix_1) + saying[index] + ' ' + random.choice(suffix).capitalize()\n else:\n xx = random.choice(prefix_2) + author[index] + ', ' + saying[index] + ' ' + random.choice(suffix).capitalize()\n return xx\n\ndef paragraph():\n xx = \". \"\n xx += \"\\r\\n\"\n xx += \" \"\n return xx\n\ndef generator(theme,length):\n tmp = ' '\n while (len(tmp) < length):\n para = random.randint(0,100)\n if para < 5 and tmp[-2] != ',':\n tmp += paragraph()\n elif para < 20 :\n tmp += random.choice(examples)\n tmp += sayings()\n elif 20 <= para <= 65:\n tmp += random.choice(addings)\n tmp += random.choice(bullshits)\n else:\n tmp += random.choice(contrasts)\n tmp += random.choice(bullshits)\n tmp = tmp.replace(\"xx\",theme)\n return tmp\n\ndef clean(a):\n a = a.replace(' ',' ').replace('. .','.').replace('? .','?').replace(', .',',').replace('..','.')\n lst = a.split(' ')\n for i in range(len(lst) - 1):\n if lst[i].endswith(',') or lst[i].endswith(':'):\n lst[i+1] = lst[i+1].lower()\n return ' '.join(lst)\n", "_____no_output_____" ], [ "tmp = generator('PUT_YOUR_WORD_HERE',10000)\ntmp = clean(tmp)\nprint(tmp)", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code" ] ]

ec7d221c20d31eaa17696060584af04a6146d6ca

737,986

ipynb

Jupyter Notebook

Coursework/Machine Learning/Assignment 1/Practical.ipynb

Sh3B0/courses

584d8d7b84ab2c58e3dd5b9f716191db64a5c2f6

[ "MIT" ]

null

Coursework/Machine Learning/Assignment 1/Practical.ipynb

Sh3B0/courses

584d8d7b84ab2c58e3dd5b9f716191db64a5c2f6

[ "MIT" ]

null

Coursework/Machine Learning/Assignment 1/Practical.ipynb

Sh3B0/courses

584d8d7b84ab2c58e3dd5b9f716191db64a5c2f6

[ "MIT" ]

null

252.216678

436,620

0.896013

[ [ [ "# **Ahmed Nouralla - B19-CS-01 - [email protected]**\n# **Machine Learning (F21) - Assignment 1**", "_____no_output_____" ] ], [ [ "import pandas as pd\nimport numpy as np\nimport matplotlib.pyplot as plt", "_____no_output_____" ] ], [ [ "## **Practical Task 1**", "_____no_output_____" ], [ "### **Data Exploration**", "_____no_output_____" ] ], [ [ "df = pd.read_csv('./task1_dataset.csv')\nground = pd.read_csv('./task1_dataset_full.csv')\ndf.describe()", "_____no_output_____" ], [ "# Renaming columns, just for convenience\ndf = df.rename(columns={\"Unnamed: 0\": \"index\",\n \"0\": \"datetime\", \"1\": \"feature1\",\n \"2\": \"feature2\", \"3\": \"feature3\"})\nground = ground.rename(columns={\"Unnamed: 0\": \"index\"})\ndf.head()", "_____no_output_____" ], [ "df.dtypes", "_____no_output_____" ], [ "df.isna().sum() / df.count()", "_____no_output_____" ] ], [ [ "**Observations**:\n1. Dataset contains 4 columns, the first one is just an index, the second one contains datetime (a categorical feature), and the last three contain numerical data.\n2. Almost 40% of the data in each of the 3 numerical columns in missing.\n3. Numerical data is independent from each other (given).", "_____no_output_____" ], [ "### **Categorical Features Encoding**", "_____no_output_____" ], [ "- The following code snippet will encode values in `datetime` column into numbers from [0, 1999] such that earlier dates have smaller corresponding encoded values.\n\n- It will also sort samples by increasing datetime value for convenience.", "_____no_output_____" ] ], [ [ "df[\"datetime\"] = df[\"datetime\"].astype('category')\ndf[\"datetime\"] = df[\"datetime\"].cat.codes\ndf.sort_values(by=['datetime'], inplace=True)\n\nground[\"datetime\"] = ground[\"datetime\"].astype('category')\nground[\"datetime\"] = ground[\"datetime\"].cat.codes\nground.sort_values(by=['datetime'], inplace=True)\n\ndf.head()", "_____no_output_____" ] ], [ [ "### **Data Visualization**", "_____no_output_____" ] ], [ [ "plt.figure(figsize=(18, 5))\n\nfor i in range(1, 4):\n plt.subplot(1, 3, i)\n plt.scatter(df[\"datetime\"], df[f\"feature{i}\"], s=5, label='samples', color='g')\n plt.xlabel(\"datetime\")\n plt.ylabel(f\"feature_{i}\")\n plt.grid()\n plt.legend()\n\nplt.show()", "_____no_output_____" ] ], [ [ "**Observations**\n- It's obvious from the plots above that the data is following a certain pattern that we can utilize to fill the missing values.\n- My initial guess is that, training a linear regression model will help predicting values for the first feature, while higher degree models will be required for the second and third features.", "_____no_output_____" ], [ "### **Data Imputation Using Polynomial Regression**\nWe need to:\n1. Train several Polynomial Regression models (with multiple degrees from [1-10]) on the present values from each of the 3 features.\n2. Predict the missing values using the models.\n3. Comapre the predictions with the actual values from the `ground` dataset\n4. Choose the model that minimizes MSE between prediction and test (ground truth).", "_____no_output_____" ] ], [ [ "# Separating training and test data for each of the three features\nx_train = {}\ny_train = {}\nx_test = {}\ny_test = {}\n\nfor i in range(1, 4):\n x_train[i] = df[df[f\"feature{i}\"].notnull()][f\"datetime\"]\n y_train[i] = df[df[f\"feature{i}\"].notnull()][f\"feature{i}\"]\n x_test[i] = df[df[f\"feature{i}\"].isnull()][\"datetime\"]\n\n tmp = df[df[f\"feature{i}\"].isnull()].index\n test_rows = []\n for j in range(len(df.index)):\n if not (j in tmp):\n test_rows.append(-1) # filling empty places to match lengths\n else:\n test_rows.append(j)\n\n y_test[i] = ground[ground[\"index\"] == test_rows][f\"feature{i}\"]\n \n x_train[i] = x_train[i].values.reshape(-1, 1)\n y_train[i] = y_train[i].values.reshape(-1, 1)\n x_test[i] = x_test[i].values.reshape(-1, 1)", "_____no_output_____" ], [ "from sklearn.preprocessing import PolynomialFeatures\nfrom sklearn.pipeline import Pipeline\nfrom sklearn.linear_model import LinearRegression\nfrom sklearn.metrics import mean_squared_error\n\n# Using polynomial regression to train the model\ny_pred = {}\nmse = {}\ncnt = 1\n\nplt.figure(figsize=(30, 15))\n\nfor i in range(1, 4):\n for degree in range(1, 11):\n poly_feats = PolynomialFeatures(degree=degree, include_bias=False)\n lin_reg = LinearRegression()\n\n pipeline = Pipeline([(\"1\", poly_feats), (\"2\", lin_reg)])\n pipeline.fit(x_train[i], y_train[i])\n y_pred[i] = pipeline.predict(x_test[i])\n\n # imputed_df contains y_train + y_pred[i] for missing values, while ground\n # contains some y values that are not necessarily the same as y_train for a\n # certain datetime.\n # Task required plotting (MSE between imputed dataset and ground truth one).\n isnull = df[f'feature{i}'].isnull()\n imputed_df = df.copy()\n imputed_df.loc[isnull, f'feature{i}'] = y_pred[i]\n mse[(i, degree)] = mean_squared_error(ground[f'feature{i}'],\n imputed_df[f'feature{i}'])\n\n\n # However, it seems more logical to plot MSE between only test set and the\n # prediction done by the model (guesses for null values in the df)\n # mse[(i, degree)] = mean_squared_error(y_test[i], y_pred[i])\n\n plt.subplot(3, 10, cnt)\n cnt += 1\n plt.plot(x_test[i], y_pred[i], color='red')\n plt.scatter(df[\"datetime\"], df[f\"feature{i}\"], s=1)\n \nplt.show()", "_____no_output_____" ] ], [ [ "### **Model Evaluation and Conclusion**\nIt seems from the plots below that:\n- For the first feature, the data seems the have a linear nature. Numerically, **the 5th degree** polynomial performed best, but there is not so big difference, **the linear model is still favorable** for it's simplicity.\n\n- For the second feature, the data seems to follow a sinusoidal shape with high frequency, and we know that sin(x) can be described by an infinite power series, and since we cannot acheive that even with 10th degree, **the linear model** will work well in practice as an average.\n\n- Third feature also follows a sinusoidal wave, but with lower frequency than the second one. Numerically, **the 2nd degree** polynomial performed best for that case, although there is not so big difference between it and **the linear model** and we still can prefer it for simplicity.", "_____no_output_____" ] ], [ [ "# Plotting MSEs\nplt.figure(figsize=(18, 5))\n\nfor i in range(1, 4):\n plt.subplot(1, 3, i)\n x_val = np.linspace(1, 10, 10)\n y_val = [mse[(i, y)] for y in range(1, 11)]\n\n print(f\"Minimal MSE for feature{i} occurs at degree: {np.argmin(y_val)+1}\")\n \n plt.plot(x_val, y_val, color='red')\n plt.xlabel(\"degree\")\n plt.ylabel(f\"mse_{i}\")\n plt.grid()\n\nplt.show()", "Minimal MSE for feature1 occurs at degree: 5\nMinimal MSE for feature2 occurs at degree: 1\nMinimal MSE for feature3 occurs at degree: 2\n" ] ], [ [ "## **Practical Task 2**", "_____no_output_____" ], [ "### **Data Exploration**", "_____no_output_____" ] ], [ [ "# Adding attribute indices on top for easier manipulation (if not added before)\nwith open('./GermanData.csv', 'r') as f:\n if f.read(3) == 'A11':\n with open('./GermanData.csv', 'r+') as f:\n content = f.read()\n f.seek(0, 0)\n f.write(','.join([f'A{i}' for i in range(1, 21)]) + ',y\\n' + content)\n\n# Reading data from CSV file\ndf = pd.read_csv('./GermanData.csv')\ndf # Note: y=1 means that row is classified as good, y=2 -> bad.", "_____no_output_____" ], [ "types = df.dtypes\nprint(types)\nprint(\"#Categorical features: \", sum(types == 'object'))\nprint(\"#Numerical features: \", sum(types == 'float64') + sum(types == 'int64'))", "A1 object\nA2 int64\nA3 object\nA4 object\nA5 int64\nA6 object\nA7 object\nA8 object\nA9 object\nA10 object\nA11 float64\nA12 object\nA13 float64\nA14 object\nA15 object\nA16 float64\nA17 object\nA18 float64\nA19 object\nA20 object\ny float64\ndtype: object\n#Categorical features: 14\n#Numerical features: 7\n" ], [ "# Manually fixing some Logical errors in the df\ndf['A2'] = df['A2'].astype(float)\ndf['A5'] = df['A5'].astype(float)\ndf.drop(df[df['A8'] == 'A192'].index, inplace=True)\ndf['A8'] = df['A8'].astype(float)\ndf['y'] = df['y'].astype(int)", "_____no_output_____" ], [ "# Show description for numerical features\ndf.describe()", "_____no_output_____" ], [ "# Show description for categorical features\ndf.describe(exclude=np.number)", "_____no_output_____" ] ], [ [ "### **Correlation Matrix**\nIt seems that features 2 (duration in month) and 5 (credit amount) are highly correlated, this can decrease regression performance as it'll be harder to tweak one parameter without changing the other.", "_____no_output_____" ] ], [ [ "import seaborn as sns\ncorr = df.corr()\nprint(sns.heatmap(corr, cmap=\"Blues\"))", "AxesSubplot(0.125,0.125;0.62x0.755)\n" ] ], [ [ "### **Data imputation**\n- Fields representing unknown data will be replaced with NaNs\n- Then impute the missing values (using mean for numerial features and most_frequent for categorical features)", "_____no_output_____" ] ], [ [ "def nan_replace(feature, nan_val):\n if not df.get(feature) is None:\n df.loc[df[feature] == nan_val, feature] = np.nan\n\nnan_replace('A1', 'A14')\nnan_replace('A6', 'A65')\nnan_replace('A12', 'A124')", "_____no_output_____" ], [ "from sklearn.impute import SimpleImputer\n\nnum_df = df[['A2', 'A5', 'A8', 'A11', 'A13', 'A16', 'A18']]\ncat_df = df[['A1', 'A3', 'A4', 'A6', 'A7', 'A9', 'A10', 'A12', 'A14', 'A15',\n 'A17', 'A19', 'A20']]\ny = df['y']\n\ntmp = SimpleImputer(strategy='mean').fit_transform(num_df)\nnum_df = pd.DataFrame(tmp, columns=num_df.columns, index=num_df.index)\n\ntmp = SimpleImputer(strategy='most_frequent').fit_transform(cat_df)\ncat_df = pd.DataFrame(tmp, columns=cat_df.columns, index=cat_df.index)\n\ndf = pd.concat([num_df, cat_df], axis=1)\ndf = df.reindex(sorted(df.columns, key=lambda x: int(x[1:])), axis=1)\ndf = pd.concat([df, y], axis=1)\ndf", "_____no_output_____" ] ], [ [ "### **Categorical Feature Encoding**\nIt seems from the [description](https://archive.ics.uci.edu/ml/datasets/statlog+(german+credit+data)) of the categorical features that:\n- For features (1, 6, 7, 15, 17); label encoding is appropriate, since they represent data that is comparable (some values have higher preference/precedence than others)\n- For the rest of the categorical features (4, 9, 10, 12, 14, 20), using one-hot-encoding is ok, since values are not comparable and not too many categories exist. ", "_____no_output_____" ] ], [ [ "from sklearn.preprocessing import OneHotEncoder\n\ndef label_encoding(feature, cats):\n for i in range(cats + 1):\n df[feature].replace(f'{feature}{i+1}', i, inplace=True)\n\ndef one_hot_encoding(feature):\n if not df.get(feature) is None:\n encoder = OneHotEncoder(handle_unknown='ignore', sparse=False)\n encoder.fit(df[df[feature].notnull()][[feature]])\n new_feats = encoder.transform(df[df[feature].notnull()][[feature]])\n new_cols = pd.DataFrame(new_feats, dtype=int)\n new_df = pd.concat([df, new_cols], axis=1)\n\n columns = {}\n for i in range(11):\n columns[i] = f'{feature}_{i}' \n\n new_df.rename(columns=columns, inplace=True)\n new_df.drop(feature, axis=1, inplace=True) \n return new_df\n\n else:\n return df\n\nlabel_encoding('A1', 3)\nlabel_encoding('A6', 4)\nlabel_encoding('A7', 5)\nlabel_encoding('A15', 3)\nlabel_encoding('A17', 4)\nlabel_encoding('A19', 2)\n\n# One more unusal label encoding\nfor i in range(5):\n df['A3'].replace(f'A3{i}', 4-i, inplace=True)\n\ndf = one_hot_encoding('A4')\ndf = one_hot_encoding('A9')\ndf = one_hot_encoding('A10')\ndf = one_hot_encoding('A12')\ndf = one_hot_encoding('A14')\ndf = one_hot_encoding('A20')\ndf.dropna(inplace=True)\ndf", "_____no_output_____" ] ], [ [ "### **Feature Scaling**\n- We can do standard scaling for all numerical features to ensure convergence of the logistic model.", "_____no_output_____" ] ], [ [ "from sklearn.preprocessing import StandardScaler, MinMaxScaler\n\nfeats_to_scale = ['A2', 'A5', 'A8', 'A11', 'A13', 'A16', 'A18']\nscaler = StandardScaler()\nfor i in feats_to_scale:\n df[i] = scaler.fit_transform(df[i].values.reshape(-1, 1))", "_____no_output_____" ] ], [ [ "### **Data Visualization**", "_____no_output_____" ] ], [ [ "from sklearn.decomposition import PCA\n\ndim_reducer = PCA(n_components=2)\ndf_reduced = dim_reducer.fit_transform(df)\nplt.scatter(df_reduced[:, 0], df_reduced[:, 1], color='blue')\n\nplt.show()", "_____no_output_____" ] ], [ [ "### **Data Splitting**", "_____no_output_____" ] ], [ [ "from sklearn.model_selection import train_test_split\n\nx = df.drop('y', axis=1)\ny = df['y']\nx_train, x_test, y_train, y_test = train_test_split(x, y, train_size=0.7, stratify=y, random_state=0)", "_____no_output_____" ] ], [ [ "### **Applying Logistic Regression**", "_____no_output_____" ] ], [ [ "from sklearn.linear_model import LogisticRegression\nfrom sklearn.preprocessing import PolynomialFeatures\nfrom sklearn.pipeline import Pipeline\nfrom sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score\nfrom sklearn.metrics import mean_squared_error\naccuracy = []\nprecision = []\nrecall = []\nf1 = []\n\ntr_err = []\nte_err = []\n\ndegrees = [i for i in range(1, 6)]\n\nfor degree in degrees:\n poly_feats = PolynomialFeatures(degree=degree)\n log_reg = LogisticRegression(random_state=0,\n class_weight='balanced',\n solver='liblinear')\n pipeline = Pipeline([(\"1\", poly_feats), (\"2\", log_reg)])\n pipeline.fit(x_train, y_train)\n y_pred = pipeline.predict(x_test)\n te_err.append(mean_squared_error(y_pred, y_test))\n tr_err.append(mean_squared_error(pipeline.predict(x_train), y_train))\n \n accuracy.append(accuracy_score(y_test, y_pred))\n precision.append(precision_score(y_test, y_pred))\n recall.append(recall_score(y_test, y_pred))\n f1.append(f1_score(y_test, y_pred))\n\nplt.plot(degrees, tr_err, label='tr_err')\nplt.plot(degrees, te_err, label='te_err')\nplt.legend()\nplt.grid()\nplt.show()\n\nplt.plot(degrees, accuracy, color='red', label=\"accuracy\")\nplt.plot(degrees, precision, color='blue', label=\"precision\")\nplt.plot(degrees, recall, color='orange', label=\"recall\")\nplt.plot(degrees, f1, color='green', label=\"f1\")\nplt.legend()\nplt.grid()\nplt.show()", "_____no_output_____" ] ], [ [ "**Observation:** It seems that the model of the **3rd** degree balances between the bias/variance, so we will use it", "_____no_output_____" ], [ "### **Optimizing hyperparameters**", "_____no_output_____" ] ], [ [ "from sklearn.linear_model import LogisticRegression\nfrom sklearn.model_selection import GridSearchCV\n\n# lbfgs solver doesn't support l1 penalty, so we split params\nparam_grid = [\n {\n '2__penalty': ['l2'],\n '2__solver': ['lbfgs'],\n '2__C': np.logspace(-4, 4, 20),\n },\n {\n '2__penalty': ['l1', 'l2'],\n '2__solver': ['liblinear'],\n '2__C': np.logspace(-4, 4, 20),\n },\n]\n\npoly_feats = PolynomialFeatures(degree=3)\nlog_reg = LogisticRegression(random_state=0, class_weight='balanced')\npipeline = Pipeline([(\"1\", poly_feats), (\"2\", log_reg)])\n\ngrid_search = GridSearchCV(estimator=pipeline,\n param_grid=param_grid,\n scoring='f1'\n )\n\ngrid_search.fit(x_train, y_train)\n\nprint(f\"Best parameters: {grid_search.best_params_}\")", "Best parameters: {'2__C': 0.0001, '2__penalty': 'l1', '2__solver': 'liblinear'}\n" ], [ "best_model = grid_search.best_estimator_\nbest_model", "_____no_output_____" ] ], [ [ "### **Model Evaluation and Conclusion**", "_____no_output_____" ], [ "Comparing the accuracy of predictions across male and female applicants, we find that accuracy for predictions on males is higher.\n\nThis can be due to:\n- The fact that the dataset has more specific fields for males (single, married, separater)\n- Only one field (divorced/separated/married) and no samples with single females (A95) at all.", "_____no_output_____" ] ], [ [ "x_test_male = x_test[x_test['A9_0'] + x_test['A9_2'] + x_test['A9_3'] > 0]\nx_test_fem = x_test[x_test['A9_1'] > 0]\n\ny_test_male = y_test[x_test_male.index]\ny_test_fem = y_test[x_test_fem.index]\n\ny_pred_male = best_model.predict(x_test_male)\ny_pred_fem = best_model.predict(x_test_fem)\n\nprint(\"Prediction accuracy for males:\", accuracy_score(y_test_male, y_pred_male))\nprint(\"Prediction accuracy for females:\", accuracy_score(y_test_fem, y_pred_fem))", "Prediction accuracy for males: 0.7323232323232324\nPrediction accuracy for females: 0.6476190476190476\n" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code", "code", "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code" ] ]

ec7d5225b095ff96ec5796f53235b2de960b7519

19,483

ipynb

Jupyter Notebook

matrix_one/day3.ipynb

ryszard-zone/dw_matrix1

a3445e9461aa7e735bf3b78ce419537b300be8c0

[ "MIT" ]

null

matrix_one/day3.ipynb

ryszard-zone/dw_matrix1

a3445e9461aa7e735bf3b78ce419537b300be8c0

[ "MIT" ]

null

matrix_one/day3.ipynb

ryszard-zone/dw_matrix1

a3445e9461aa7e735bf3b78ce419537b300be8c0

[ "MIT" ]

null

19,483

0.784119

[ [ [ "#!pip install datadotworld\n#!pip install datadotworld[pandas]", "_____no_output_____" ], [ "#!dw configure", "_____no_output_____" ], [ "from google.colab import drive\nimport pandas as pd\nimport numpy as np\nimport datadotworld as dw", "_____no_output_____" ], [ "#drive.mount(\"/content/drive\")", "_____no_output_____" ], [ "cd \"/content/drive/My Drive/Colab Notebooks/dw_matrix1\"", "/content/drive/My Drive/Colab Notebooks/dw_matrix1\n" ], [ "!echo 'data' > .gitignore", "_____no_output_____" ], [ "!git add .gitignore", "_____no_output_____" ], [ "data = dw.load_dataset('datafiniti/mens-shoe-prices')", "_____no_output_____" ], [ "df = data.dataframes['7004_1']\ndf.shape\n#df.sample(5)\n#df.columns\n#df.prices_currency.unique()\n#df.prices_currency.value_counts()\n#df.prices_currency.value_counts(normalize=True)", "/usr/local/lib/python3.6/dist-packages/datadotworld/models/dataset.py:209: UserWarning: Unable to set data frame dtypes automatically using 7004_1 schema. Data types may need to be adjusted manually. Error: Integer column has NA values in column 10\n 'Error: {}'.format(resource_name, e))\n/usr/local/lib/python3.6/dist-packages/datadotworld/util.py:121: DtypeWarning: Columns (39,45) have mixed types. Specify dtype option on import or set low_memory=False.\n return self._loader_func()\n" ], [ "df_usd = df[ df.prices_currency == 'USD' ].copy()\ndf_usd.shape", "_____no_output_____" ], [ "df.columns", "_____no_output_____" ], [ "df_usd['prices_amountmin'] = df_usd.prices_amountmin.astype(np.float)", "_____no_output_____" ], [ "df_usd[ df_usd.prices_amountmin > 10000 ][['prices_amountmin','prices_amountmax','prices_currency']].head(20)", "_____no_output_____" ], [ "filter_max = np.percentile(df_usd['prices_amountmin'], 99) #w 99% przypadkow buty kosztuja < 895.0\nfilter_max", "_____no_output_____" ], [ "df_usd_filter = df_usd[df_usd['prices_amountmin'] < filter_max ]", "_____no_output_____" ], [ "df_usd_filter['prices_amountmin'].hist(bins=100)", "_____no_output_____" ], [ "!git add matrix_one/day3.ipynb", "_____no_output_____" ], [ "!git config --global user.email \"[email protected]\"\n!git config --global user.name \"ryszard\"", "_____no_output_____" ], [ "!git commit -m \"Read Men's Shoe Prices dataset from data.world\"", "[master 1f1410a] Read Men's Shoe Prices dataset from data.world\n 1 file changed, 1 insertion(+), 1 deletion(-)\n rewrite matrix_one/day3.ipynb (70%)\n" ], [ "!git push -u origin master", "Counting objects: 9, done.\nDelta compression using up to 2 threads.\nCompressing objects: 16% (1/6) \rCompressing objects: 33% (2/6) \rCompressing objects: 50% (3/6) \rCompressing objects: 66% (4/6) \rCompressing objects: 83% (5/6) \rCompressing objects: 100% (6/6) \rCompressing objects: 100% (6/6), done.\nWriting objects: 11% (1/9) \rWriting objects: 22% (2/9) \rWriting objects: 33% (3/9) \rWriting objects: 44% (4/9) \rWriting objects: 55% (5/9) \rWriting objects: 77% (7/9) \rWriting objects: 88% (8/9) \rWriting objects: 100% (9/9) \rWriting objects: 100% (9/9), 14.47 KiB | 871.00 KiB/s, done.\nTotal 9 (delta 2), reused 0 (delta 0)\nremote: Resolving deltas: 100% (2/2), done.\u001b[K\nTo https://github.com/ryszard-zone/dw_matrix1.git\n d13f707..1f1410a master -> master\nBranch 'master' set up to track remote branch 'master' from 'origin'.\n" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7d55e9acf962223230710766ba102a7a3c2e80

2,894

ipynb

Jupyter Notebook

mytests/amumax.ipynb

MathieuMoalic/mumax3

9bb497bb447f60d9e3ad704e041ba95ce17f447d

[ "CC-BY-3.0" ]

null

mytests/amumax.ipynb

MathieuMoalic/mumax3

9bb497bb447f60d9e3ad704e041ba95ce17f447d

[ "CC-BY-3.0" ]

null

mytests/amumax.ipynb

MathieuMoalic/mumax3

9bb497bb447f60d9e3ad704e041ba95ce17f447d

[ "CC-BY-3.0" ]

null

22.091603

120

0.489634

[ [ [ "%matplotlib widget\n%load_ext autoreload\n%autoreload 2\nfrom adl import adl,parms\nfrom llyr import op\nfrom matplotlib import pyplot as plt\nimport matplotlib as mpl\nimport numpy as np\nimport cmocean\nfrom glob import glob\nfrom tqdm.notebook import tqdm\n# import gc\nplt.style.use('workfolder')", "_____no_output_____" ], [ "m = op(\"test1.zarr\")\nm.p\nplt.figure()\n# plt.plot(m.m.attrs['Buffer'])\nplt.plot(m.table.t[:],m.table.m[2,:])\n# m.plot.snapshot('m',t=5)", "/\n └── table\n ├── B_demag (3, 1001) float64\n ├── m (3, 1001) float64\n └── t (1001,) float64\n" ], [ "def largest_prime_factor(n):\n i = 2\n while i * i <= n:\n if n % i:\n i += 1\n else:\n n //= i\n return n\n\nL = []\nfor i in range(1500):\n q = largest_prime_factor(i)\n if q > 7:\n L.append(i)\nprint(L)", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code" ] ]

ec7d58a0f4902ef4175fa34dce272b62098d26ab

52,513

ipynb

Jupyter Notebook

model/bert/notebooks/Comparing TF and PT models.ipynb

pharouhk/Querying-and-forecasting-with-Natural-Language

390cf179013b43f48e6c116cf95ebd282a031b4c

[ "MIT" ]

1

2022-01-31T12:11:26.000Z

model/bert/notebooks/Comparing TF and PT models.ipynb

pharouhk/Querying-and-forecasting-with-Natural-Language

390cf179013b43f48e6c116cf95ebd282a031b4c

[ "MIT" ]

null

model/bert/notebooks/Comparing TF and PT models.ipynb

pharouhk/Querying-and-forecasting-with-Natural-Language

390cf179013b43f48e6c116cf95ebd282a031b4c

[ "MIT" ]

null

41.381403

742

0.493725

[ [ [ "# Comparing TensorFlow (original) and PyTorch models\n\nYou can use this small notebook to check the conversion of the model's weights from the TensorFlow model to the PyTorch model. In the following, we compare the weights of the last layer on a simple example (in `input.txt`) but both models returns all the hidden layers so you can check every stage of the model.\n\nTo run this notebook, follow these instructions:\n- make sure that your Python environment has both TensorFlow and PyTorch installed,\n- download the original TensorFlow implementation,\n- download a pre-trained TensorFlow model as indicaded in the TensorFlow implementation readme,\n- run the script `convert_tf_checkpoint_to_pytorch.py` as indicated in the `README` to convert the pre-trained TensorFlow model to PyTorch.\n\nIf needed change the relative paths indicated in this notebook (at the beggining of Sections 1 and 2) to point to the relevent models and code.", "_____no_output_____" ] ], [ [ "import os\nos.chdir('../')", "_____no_output_____" ] ], [ [ "## 1/ TensorFlow code", "_____no_output_____" ] ], [ [ "original_tf_inplem_dir = \"./tensorflow_code/\"\nmodel_dir = \"../google_models/uncased_L-12_H-768_A-12/\"\n\nvocab_file = model_dir + \"vocab.txt\"\nbert_config_file = model_dir + \"bert_config.json\"\ninit_checkpoint = model_dir + \"bert_model.ckpt\"\n\ninput_file = \"./samples/input.txt\"\nmax_seq_length = 128", "_____no_output_____" ], [ "import importlib.util\nimport sys\n\nspec = importlib.util.spec_from_file_location('*', original_tf_inplem_dir + '/extract_features.py')\nmodule = importlib.util.module_from_spec(spec)\nspec.loader.exec_module(module)\nsys.modules['extract_features_tensorflow'] = module\n\nfrom extract_features_tensorflow import *", "_____no_output_____" ], [ "layer_indexes = list(range(12))\nbert_config = modeling.BertConfig.from_json_file(bert_config_file)\ntokenizer = tokenization.FullTokenizer(\n vocab_file=vocab_file, do_lower_case=True)\nexamples = read_examples(input_file)\n\nfeatures = convert_examples_to_features(\n examples=examples, seq_length=max_seq_length, tokenizer=tokenizer)\nunique_id_to_feature = {}\nfor feature in features:\n unique_id_to_feature[feature.unique_id] = feature", "INFO:tensorflow:*** Example ***\nINFO:tensorflow:unique_id: 0\nINFO:tensorflow:tokens: [CLS] who was jim henson ? [SEP] jim henson was a puppet ##eer [SEP]\nINFO:tensorflow:input_ids: 101 2040 2001 3958 27227 1029 102 3958 27227 2001 1037 13997 11510 102 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\nINFO:tensorflow:input_mask: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\nINFO:tensorflow:input_type_ids: 0 0 0 0 0 0 0 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n" ], [ "is_per_host = tf.contrib.tpu.InputPipelineConfig.PER_HOST_V2\nrun_config = tf.contrib.tpu.RunConfig(\n master=None,\n tpu_config=tf.contrib.tpu.TPUConfig(\n num_shards=1,\n per_host_input_for_training=is_per_host))\n\nmodel_fn = model_fn_builder(\n bert_config=bert_config,\n init_checkpoint=init_checkpoint,\n layer_indexes=layer_indexes,\n use_tpu=False,\n use_one_hot_embeddings=False)\n\n# If TPU is not available, this will fall back to normal Estimator on CPU\n# or GPU.\nestimator = tf.contrib.tpu.TPUEstimator(\n use_tpu=False,\n model_fn=model_fn,\n config=run_config,\n predict_batch_size=1)\n\ninput_fn = input_fn_builder(\n features=features, seq_length=max_seq_length)", "WARNING:tensorflow:Estimator's model_fn (<function model_fn_builder.<locals>.model_fn at 0x12839dbf8>) includes params argument, but params are not passed to Estimator.\nWARNING:tensorflow:Using temporary folder as model directory: /var/folders/yx/cw8n_njx3js5jksyw_qlp8p00000gn/T/tmpdbx_h23u\nINFO:tensorflow:Using config: {'_model_dir': '/var/folders/yx/cw8n_njx3js5jksyw_qlp8p00000gn/T/tmpdbx_h23u', '_tf_random_seed': None, '_save_summary_steps': 100, '_save_checkpoints_steps': None, '_save_checkpoints_secs': 600, '_session_config': allow_soft_placement: true\ngraph_options {\n rewrite_options {\n meta_optimizer_iterations: ONE\n }\n}\n, '_keep_checkpoint_max': 5, '_keep_checkpoint_every_n_hours': 10000, '_log_step_count_steps': None, '_train_distribute': None, '_device_fn': None, '_protocol': None, '_eval_distribute': None, '_experimental_distribute': None, '_service': None, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x12b3e1c18>, '_task_type': 'worker', '_task_id': 0, '_global_id_in_cluster': 0, '_master': '', '_evaluation_master': '', '_is_chief': True, '_num_ps_replicas': 0, '_num_worker_replicas': 1, '_tpu_config': TPUConfig(iterations_per_loop=2, num_shards=1, num_cores_per_replica=None, per_host_input_for_training=3, tpu_job_name=None, initial_infeed_sleep_secs=None, input_partition_dims=None), '_cluster': None}\nWARNING:tensorflow:Setting TPUConfig.num_shards==1 is an unsupported behavior. Please fix as soon as possible (leaving num_shards as None.\nINFO:tensorflow:_TPUContext: eval_on_tpu True\nWARNING:tensorflow:eval_on_tpu ignored because use_tpu is False.\n" ], [ "tensorflow_all_out = []\nfor result in estimator.predict(input_fn, yield_single_examples=True):\n unique_id = int(result[\"unique_id\"])\n feature = unique_id_to_feature[unique_id]\n output_json = collections.OrderedDict()\n output_json[\"linex_index\"] = unique_id\n tensorflow_all_out_features = []\n # for (i, token) in enumerate(feature.tokens):\n all_layers = []\n for (j, layer_index) in enumerate(layer_indexes):\n print(\"extracting layer {}\".format(j))\n layer_output = result[\"layer_output_%d\" % j]\n layers = collections.OrderedDict()\n layers[\"index\"] = layer_index\n layers[\"values\"] = layer_output\n all_layers.append(layers)\n tensorflow_out_features = collections.OrderedDict()\n tensorflow_out_features[\"layers\"] = all_layers\n tensorflow_all_out_features.append(tensorflow_out_features)\n\n output_json[\"features\"] = tensorflow_all_out_features\n tensorflow_all_out.append(output_json)", "INFO:tensorflow:Could not find trained model in model_dir: /var/folders/yx/cw8n_njx3js5jksyw_qlp8p00000gn/T/tmpdbx_h23u, running initialization to predict.\nINFO:tensorflow:Calling model_fn.\nINFO:tensorflow:Running infer on CPU\nINFO:tensorflow:Done calling model_fn.\nINFO:tensorflow:Graph was finalized.\nINFO:tensorflow:Running local_init_op.\nINFO:tensorflow:Done running local_init_op.\nextracting layer 0\nextracting layer 1\nextracting layer 2\nextracting layer 3\nextracting layer 4\nextracting layer 5\nextracting layer 6\nextracting layer 7\nextracting layer 8\nextracting layer 9\nextracting layer 10\nextracting layer 11\nINFO:tensorflow:prediction_loop marked as finished\nINFO:tensorflow:prediction_loop marked as finished\n" ], [ "print(len(tensorflow_all_out))\nprint(len(tensorflow_all_out[0]))\nprint(tensorflow_all_out[0].keys())\nprint(\"number of tokens\", len(tensorflow_all_out[0]['features']))\nprint(\"number of layers\", len(tensorflow_all_out[0]['features'][0]['layers']))\ntensorflow_all_out[0]['features'][0]['layers'][0]['values'].shape", "1\n2\nodict_keys(['linex_index', 'features'])\nnumber of tokens 1\nnumber of layers 12\n" ], [ "tensorflow_outputs = list(tensorflow_all_out[0]['features'][0]['layers'][t]['values'] for t in layer_indexes)", "_____no_output_____" ] ], [ [ "## 2/ PyTorch code", "_____no_output_____" ] ], [ [ "import extract_features\nfrom extract_features import *", "_____no_output_____" ], [ "init_checkpoint_pt = \"../google_models/uncased_L-12_H-768_A-12/pytorch_model.bin\"", "_____no_output_____" ], [ "device = torch.device(\"cpu\")\nmodel = extract_features.BertModel(bert_config)\nmodel.load_state_dict(torch.load(init_checkpoint_pt, map_location='cpu'))\nmodel.to(device)", "_____no_output_____" ], [ "all_input_ids = torch.tensor([f.input_ids for f in features], dtype=torch.long)\nall_input_mask = torch.tensor([f.input_mask for f in features], dtype=torch.long)\nall_input_type_ids = torch.tensor([f.input_type_ids for f in features], dtype=torch.long)\nall_example_index = torch.arange(all_input_ids.size(0), dtype=torch.long)\n\neval_data = TensorDataset(all_input_ids, all_input_mask, all_input_type_ids, all_example_index)\neval_sampler = SequentialSampler(eval_data)\neval_dataloader = DataLoader(eval_data, sampler=eval_sampler, batch_size=1)\n\nmodel.eval()", "_____no_output_____" ], [ "layer_indexes = list(range(12))\n\npytorch_all_out = []\nfor input_ids, input_mask, input_type_ids, example_indices in eval_dataloader:\n print(input_ids)\n print(input_mask)\n print(example_indices)\n input_ids = input_ids.to(device)\n input_mask = input_mask.to(device)\n\n all_encoder_layers, _ = model(input_ids, token_type_ids=input_type_ids, attention_mask=input_mask)\n\n for b, example_index in enumerate(example_indices):\n feature = features[example_index.item()]\n unique_id = int(feature.unique_id)\n # feature = unique_id_to_feature[unique_id]\n output_json = collections.OrderedDict()\n output_json[\"linex_index\"] = unique_id\n all_out_features = []\n # for (i, token) in enumerate(feature.tokens):\n all_layers = []\n for (j, layer_index) in enumerate(layer_indexes):\n print(\"layer\", j, layer_index)\n layer_output = all_encoder_layers[int(layer_index)].detach().cpu().numpy()\n layer_output = layer_output[b]\n layers = collections.OrderedDict()\n layers[\"index\"] = layer_index\n layer_output = layer_output\n layers[\"values\"] = layer_output if not isinstance(layer_output, (int, float)) else [layer_output]\n all_layers.append(layers)\n\n out_features = collections.OrderedDict()\n out_features[\"layers\"] = all_layers\n all_out_features.append(out_features)\n output_json[\"features\"] = all_out_features\n pytorch_all_out.append(output_json)", "tensor([[ 101, 2040, 2001, 3958, 27227, 1029, 102, 3958, 27227, 2001,\n 1037, 13997, 11510, 102, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0]])\ntensor([[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n 0, 0, 0, 0, 0, 0, 0, 0]])\ntensor([0])\nlayer 0 0\nlayer 1 1\nlayer 2 2\nlayer 3 3\nlayer 4 4\nlayer 5 5\nlayer 6 6\nlayer 7 7\nlayer 8 8\nlayer 9 9\nlayer 10 10\nlayer 11 11\n" ], [ "print(len(pytorch_all_out))\nprint(len(pytorch_all_out[0]))\nprint(pytorch_all_out[0].keys())\nprint(\"number of tokens\", len(pytorch_all_out))\nprint(\"number of layers\", len(pytorch_all_out[0]['features'][0]['layers']))\nprint(\"hidden_size\", len(pytorch_all_out[0]['features'][0]['layers'][0]['values']))\npytorch_all_out[0]['features'][0]['layers'][0]['values'].shape", "1\n2\nodict_keys(['linex_index', 'features'])\nnumber of tokens 1\nnumber of layers 12\nhidden_size 128\n" ], [ "pytorch_outputs = list(pytorch_all_out[0]['features'][0]['layers'][t]['values'] for t in layer_indexes)\nprint(pytorch_outputs[0].shape)\nprint(pytorch_outputs[1].shape)", "(128, 768)\n(128, 768)\n" ], [ "print(tensorflow_outputs[0].shape)\nprint(tensorflow_outputs[1].shape)", "(128, 768)\n(128, 768)\n" ] ], [ [ "## 3/ Comparing the standard deviation on the last layer of both models", "_____no_output_____" ] ], [ [ "import numpy as np", "_____no_output_____" ], [ "print('shape tensorflow layer, shape pytorch layer, standard deviation')\nprint('\\n'.join(list(str((np.array(tensorflow_outputs[i]).shape,\n np.array(pytorch_outputs[i]).shape, \n np.sqrt(np.mean((np.array(tensorflow_outputs[i]) - np.array(pytorch_outputs[i]))**2.0)))) for i in range(12))))", "shape tensorflow layer, shape pytorch layer, standard deviation\n((128, 768), (128, 768), 1.5258875e-07)\n((128, 768), (128, 768), 2.342731e-07)\n((128, 768), (128, 768), 2.801949e-07)\n((128, 768), (128, 768), 3.5904986e-07)\n((128, 768), (128, 768), 4.2842768e-07)\n((128, 768), (128, 768), 5.127951e-07)\n((128, 768), (128, 768), 6.14668e-07)\n((128, 768), (128, 768), 7.063922e-07)\n((128, 768), (128, 768), 7.906173e-07)\n((128, 768), (128, 768), 8.475192e-07)\n((128, 768), (128, 768), 8.975489e-07)\n((128, 768), (128, 768), 4.1671223e-07)\n" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ] ]

ec7d66b5b2c6000a7ec080980008259f56fa72de

10,247

ipynb

Jupyter Notebook

lecture_notes/Chapter_0/Ipython Notebook/1.0-arrangement.ipynb

TerenceLiu98/python_for_finance

25448b31709e25311d90d61dfb31e35e77c02af5

[ "MIT" ]

null

lecture_notes/Chapter_0/Ipython Notebook/1.0-arrangement.ipynb

TerenceLiu98/python_for_finance

25448b31709e25311d90d61dfb31e35e77c02af5

[ "MIT" ]

null

lecture_notes/Chapter_0/Ipython Notebook/1.0-arrangement.ipynb

TerenceLiu98/python_for_finance

25448b31709e25311d90d61dfb31e35e77c02af5

[ "MIT" ]

null

35.954386

354

0.619401

[ [ [ "# Course Arrangement", "_____no_output_____" ], [ "**Dr. Pengfei Zhao**\n\nFinance Mathematics Program, \n\nBNU-HKBU United International College", "_____no_output_____" ], [ "## 1. How to learn programming?", "_____no_output_____" ], [ "* For some students, this is your first touch of programming. You are lucky enough to choose `Python` as the first programming language since it is simple, powerful, widely used.\n\n* How to learn `***` well is a common question. Learning programming language has many commons with other subjects, e.g. spend time, keep practicing, etc, but it also has its own characters. According to 12 years of my own experience, I summarize the points below:\n\n * **Coding.** `Talk is cheap. Show me the code.` This is the famous quote from [Linus Torvalds](https://en.wikiquote.org/wiki/Linus_Torvalds). The best way to learn programming is to write code. This is the **most important** thing you learn any programming language. All the following tips are based on this golden rule.\n * **Read code**. From the study of industry level code writen by experts you can learn a lot, e.g. how to design the system, syntax which are not taught in class, etc. The most popular site is [GitHub](https://github.com). You can easily `clone` the project into local computer and study the code. \n * **Sprit of solving problems.** When you run code, you will encounter uncountable unexpected `bugs`. Sometimes they are subtle and you may feel frustrated that the output is not what you expect. If you have watched the movie \"The Martian\", the below quotes may inspire you.\n [![Watch the video](https://raw.github.com/GabLeRoux/WebMole/master/ressources/WebMole_Youtube_Video.png)](https://www.youtube.com/watch?v=mDYCLFE86Po)\n\n> At some point, everything's gonna go south on you and you're going to say, this is it. This is how I end. Now you can either accept that, or you can get to work. That's all it is. You just begin. You do the math. You solve one problem and you solve the next one, and then the next. And If you solve enough problems, you get come home.\n\n", "_____no_output_____" ], [ "## 2. About this course", "_____no_output_____" ], [ "### 2.1 How to get an `A`", "_____no_output_____" ], [ "| <div>Assessment Methods</div> | Weighting |\n| --- | --- |\n| Quiz | 20% |\n| Projects | 40% |\n| Programming Final Exam | 20% |\n| Written Final Exam | 20% |", "_____no_output_____" ], [ "### 2.2 About the Projects", "_____no_output_____" ], [ "* There will be 3 projects. Each of which requires you to code.\n* No team.\n* Do not expect to copy-paste or slightly modify other students' code. Will use the code plagiarism detection software to detect similar codes. High-similarity codes will all be graded 0.", "_____no_output_____" ], [ "#### 2.2.1 Project 1: Statistics of HongLouMeng characters", "_____no_output_____" ], [ " <img src=\"../Figures/name_cloud.png\" width = \"280\" height = \"150\" alt=\"图片名称\" align=center />", "_____no_output_____" ], [ "#### 2.2.2 Project 2: Construct python logo by class students names", "_____no_output_____" ], [ "* Given the Python Logo picture and enrolled student lists, use student names to construct the Python Logo.", "_____no_output_____" ], [ "Python Logo | Class 1 Logo | Class 2 Logo | \n:-------------------------:|:-------------------------:|:-------------------------:|\n<img src=\"../Figures/python1600.png\" width = \"300\" height = \"550\"/> | <img src=\"../Figures/python-string-py1.jpg\" width = \"300\" height = \"550\"/> | <img src=\"../Figures/python-string-py2.jpg\" width = \"300\" height = \"550\"/> | ", "_____no_output_____" ], [ "#### 2.2.3 Project 3: Golden cross is really Golden?", "_____no_output_____" ], [ "* Given a period of daily stock price, we apply the Golden Cross strategy to do the trading. Examine the earnings.", "_____no_output_____" ], [ " <img src=\"../Figures/tencent.jpeg\" width = \"580\" height = \"450\" alt=\"图片名称\" align=center />", "_____no_output_____" ], [ "## 3. About the Quantitative Trading (QT) Club", "_____no_output_____" ], [ "* In the quantitative trading WeChat group, articles are posted from time to time, and you have already gained some basic knowledge about QT.\n\n* There has some important points about this team:", "_____no_output_____" ], [ "### 3.1 What does QT club do?", "_____no_output_____" ], [ "* Different to various traditional student community, the QT club is highly technology and product oriented. We will spend most of the time studying various technologies relating to quantitative trading, and the goal is to build product benefitting the education and research in UIC.\n\n* The club will do tech meet up from time to time, share the technology learned by each club member. Once the technology is mature, we start to implement the system, and come up with a product which can be used by UIC students and staff. For example, the product can be a comprehensive financial database, an algorithm trading system, and so on.\n\n* Besides the meet up within semester, we will do system development during the summer holiday, with possibly paid salary (depending on the financial status).\n\n* Visit tech firms.", "_____no_output_____" ], [ "### 3.2 What will you benefit from the QT club?", "_____no_output_____" ], [ "* A community full of energetic students who are interested with financial and computer science technologies. Enjoy the time learning and sharing together.\n* To you FYP.\n* One to one supervision for a long period of time. Guidance to learn financial technologies.\n* Get hands on experience to implement visible and usable financial products, which will add great credits for you to apply for post-graduate study, making you distingushable from other students.\n* Research assistant opportunity during the summer holiday. When other students are enjoying their summer holiday at home or traveling, you will have clear objective to learn technology and impelement product which will be of impact.\n* Expose to industry. May get chances to be referred internship opportunities in high-tech firms.", "_____no_output_____" ], [ "### 3.3 What is the requirement to join the club?", "_____no_output_____" ], [ "* In this stage, I decide to set the club quota as 10. Depending on the number of students who would like to join the club, selection criteria may be needed. Currently in my mind I decide to set the following selection criterias:\n```\n(1) Math and computer science related courses GPA. This will count most of the weight, e.g. 90%.\n(2) Personal statement demonstrating your strong passion on the QT club.\n```\n\n* I know above criteria is a very rough selection criteria. It will become more detailed at the selection stage. But generally if you are good at math and programming, you will have much more chances to be selected.", "_____no_output_____" ], [ "### 3.4 What is the requirement of staying in the club?", "_____no_output_____" ], [ "* Joining the club should not be conflict with your regular study. Students should obtain at least B+ (included) semester GPA. Students whose semester GPA lower than B+ will be advised to quit the club.\n\n* Club members should join club activities regularly. If students are not interested in the club any more, it is allowed to quit immediately.\n\n* Other requirements may be added later.", "_____no_output_____" ], [ "### 3.5 What is the next step?", "_____no_output_____" ], [ "* In the 2018-19 fall semester, 3-5 students will be selected.\n* At the end of this semester, I will send an email to all the students. Students who are interested to join the club can then send email to me for registration. Once the semester GPA is released, I will inform the selection result.", "_____no_output_____" ] ] ]

[ "markdown" ]

[ [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ] ]

ec7d772c4bdcb81ffac1eef5f4009d0d7b33eea0

8,228

ipynb

Jupyter Notebook

function/lambda.ipynb

MaiaNgo/python-advanced

372d741b061548654ca77cb513085d2aeb3d767b

[ "Apache-2.0" ]

1

2022-02-16T03:14:27.000Z

function/lambda.ipynb

MaiaNgo/python-advanced

372d741b061548654ca77cb513085d2aeb3d767b

[ "Apache-2.0" ]

null

function/lambda.ipynb

MaiaNgo/python-advanced

372d741b061548654ca77cb513085d2aeb3d767b

[ "Apache-2.0" ]

null

20.725441

355

0.483106

[ [ [ "# lambda expression is a short way to create an annonymous function\n# syntax:\n# lambda [param list]: expression\nf = lambda x: x**2", "_____no_output_____" ], [ "# the lambda expression above create a function (without a name) and it is assigned to variable\n# actually it is just a function\ntype(f)", "_____no_output_____" ], [ "# we can call it as normal\nprint(f(3))", "9\n" ], [ "# or we can call it right after it is defined\n(lambda x: x**2)(4)", "_____no_output_____" ], [ "# we normally use lambda function as a param to pass into other function\ndef calc(f, *args):\n return f(*args)", "_____no_output_____" ], [ "print(calc(lambda x : x**2, 5)b", "_____no_output_____" ], [ "print(calc(lambda a, b : a + b, 1, 2))", "3\n" ], [ "# lambda function is normally very useful in sorting with the buit-in sorted() function\nhelp(sorted)", "Help on built-in function sorted in module builtins:\n\nsorted(iterable, /, *, key=None, reverse=False)\n Return a new list containing all items from the iterable in ascending order.\n \n A custom key function can be supplied to customize the sort order, and the\n reverse flag can be set to request the result in descending order.\n\n" ], [ "l = [1, 3, 4, 2, 6, 5, 7]", "_____no_output_____" ], [ "sorted(l)", "_____no_output_____" ], [ "# sometimes the default sorting logic is not what we want\nl = ['a', 'B', 'c', 'A']\nsorted(l)", "_____no_output_____" ], [ "# we want to sort in a case-insensitive manner\n# we provide a lambda function to handle the sorting logic\nsorted(l, key=lambda x: x.lower())", "_____no_output_____" ], [ "# when we use sorted() function with a dict, it will default to sorting by keys\n# because that how the default Iterable of dict works (iterating through keys)\nd = {'a': 1, 'B': 2, 'A': 3, 'c': 4}\nsorted(d)", "_____no_output_____" ], [ "# but if what we want is to sort by the value, and return the list of key that has that sorted value order\nsorted(d, key=lambda k: d.get(k))", "_____no_output_____" ], [ "# how's about we want to return key,value pair that are sorted by value\nsorted(d.items(), key=lambda i: i[1])", "_____no_output_____" ], [ "# or we can sort by key case-insensitive\nsorted(d.items(), key=lambda i: i[0].lower())", "_____no_output_____" ], [ "# fun: we can use sorted() and lambda function to randomize a list\nl = [1, 2, 3, 4, 5, 6, 7, 8]\nsorted(l)", "_____no_output_____" ], [ "import random\n\nsorted(l, key=lambda x : random.random())", "_____no_output_____" ], [ "sorted(l)", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7d8c0d34d10f9e0e2f9b52509707d90f4950e7

82,500

ipynb

Jupyter Notebook

Cluster/kmeans-kmedoids/KMeans-ellipse-21-0.3.ipynb

bcottman/photon_experiments

e2097dc809bd73482936b25b7504b1b9211512b2

[ "MIT" ]

null

Cluster/kmeans-kmedoids/KMeans-ellipse-21-0.3.ipynb

bcottman/photon_experiments

e2097dc809bd73482936b25b7504b1b9211512b2

[ "MIT" ]

null

Cluster/kmeans-kmedoids/KMeans-ellipse-21-0.3.ipynb

bcottman/photon_experiments

e2097dc809bd73482936b25b7504b1b9211512b2

[ "MIT" ]

null

54.133858

412

0.379139

[ [ [ "## parameters", "_____no_output_____" ] ], [ [ "CLUSTER_ALGO = 'KMeans'\n#C_SHAPE ='circle'\n#C_SHAPE ='CIRCLE'\nC_SHAPE ='ellipse'\n#N_CLUSTERS = [50,300, 1000]\nN_CLUSTERS = [21]\nCLUSTERS_STD = [0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3\n ,0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3\n ,0.3]\nN_P_CLUSTERS = [3, 30, 300, 3000]\n\nINNER_FOLDS = 3\nOUTER_FOLDS = 3", "_____no_output_____" ] ], [ [ "## includes", "_____no_output_____" ] ], [ [ "%matplotlib inline\n\nimport matplotlib\n\nimport matplotlib.pyplot as plt\nfrom sklearn.datasets import load_breast_cancer\nfrom sklearn.datasets import load_iris\nfrom sklearn.datasets import make_blobs\nfrom sklearn.datasets import make_moons\n", "_____no_output_____" ], [ "%load_ext autoreload\n%autoreload 2\npackages = !conda list\npackages", "_____no_output_____" ] ], [ [ "## Output registry", "_____no_output_____" ] ], [ [ "from __future__ import print_function\nimport sys, os\n\nold__file__ = !pwd\n__file__ = !cd ../../../photon ;pwd\n#__file__ = !pwd\n__file__ = __file__[0]\n__file__\nsys.path.append(__file__)\nprint(sys.path)\nos.chdir(old__file__[0])\n!pwd\nold__file__[0]", "['/docker/photon_experiments/Cluster/kmeans-kmedoids', '/opt/conda/lib/python37.zip', '/opt/conda/lib/python3.7', '/opt/conda/lib/python3.7/lib-dynload', '', '/opt/conda/lib/python3.7/site-packages', '/opt/conda/lib/python3.7/site-packages/IPython/extensions', '/home/jovyan/.ipython', '/docker/photon']\n/docker/photon_experiments/Cluster/kmeans-kmedoids\n" ], [ "\nimport seaborn as sns; sns.set() # for plot styling\nimport numpy as np\nimport pandas as pd\nfrom math import floor,ceil\nfrom sklearn.model_selection import KFold\nfrom sklearn.manifold import TSNE\nimport itertools\nimport seaborn as sns\nimport matplotlib.pyplot as plt\n%matplotlib inline\n\n#set font size of labels on matplotlib plots\nplt.rc('font', size=16)\n\n#set style of plots\nsns.set_style('white')\n\n#define a custom palette\nPALLET = ['#40111D', '#DCD5E4', '#E7CC74'\n ,'#39C8C6', '#AC5583', '#D3500C'\n ,'#FFB139', '#98ADA7', '#AD989E'\n ,'#708090','#6C8570','#3E534D'\n ,'#0B8FD3','#0B47D3','#96D30B' \n ,'#630C3A','#F1D0AF','#64788B' \n ,'#8B7764','#7A3C5D','#77648B'\n ,'#eaff39','#39ff4e','#4e39ff'\n ,'#ff4e39','#87ff39','#ff3987', ]\nN_PALLET = len(PALLET)\nsns.set_palette(PALLET)\nsns.palplot(PALLET)\n\n\nfrom clusim.clustering import Clustering, remap2match\nimport clusim.sim as sim\n\nfrom photonai.base import Hyperpipe, PipelineElement, Preprocessing, OutputSettings\nfrom photonai.optimization import FloatRange, Categorical, IntegerRange\nfrom photonai.base.photon_elements import PhotonRegistry\nfrom photonai.visual.graphics import plot_cm\nfrom photonai.photonlogger.logger import logger\n#from photonai.base.registry.registry import PhotonRegistry", "/opt/conda/lib/python3.7/site-packages/sklearn/externals/joblib/__init__.py:15: DeprecationWarning: sklearn.externals.joblib is deprecated in 0.21 and will be removed in 0.23. Please import this functionality directly from joblib, which can be installed with: pip install joblib. If this warning is raised when loading pickled models, you may need to re-serialize those models with scikit-learn 0.21+.\n warnings.warn(msg, category=DeprecationWarning)\n" ] ], [ [ "## function defintions", "_____no_output_____" ] ], [ [ "def yield_parameters_ellipse(n_p_clusters):\n# cluster_std = CLUSTERS_STD\n for n_p_cluster in n_p_clusters:\n for n_cluster in N_CLUSTERS:\n print('ncluster:', n_cluster)\n # n_cluster_std = [cluster_std for k in range(n_cluster)]\n n_samples = [n_p_cluster for k in range(n_cluster)]\n data_X, data_y = make_blobs(n_samples=n_samples,\n cluster_std=CLUSTERS_STD, random_state=0)\n transformation = [[0.6, -0.6], [-0.4, 0.8]]\n X_ellipse = np.dot(data_X, transformation)\n yield [X_ellipse, data_y, n_cluster]", "_____no_output_____" ], [ "def yield_parameters(n_p_clusters):\n# cluster_std = CLUSTERS_STD\n for n_p_cluster in n_p_clusters:\n for n_cluster in N_CLUSTERS:\n# n_cluster_std = [cluster_std for k in range(n_cluster)]\n n_samples = [n_p_cluster for k in range(n_cluster)]\n data_X, data_y = make_blobs(n_samples=n_samples,\n cluster_std=CLUSTERS_STD, random_state=0)\n yield [data_X, data_y, n_cluster]", "_____no_output_____" ], [ "def results_to_df(results):\n ll = []\n for obj in results:\n ll.append([obj.operation,\n obj.value,\n obj.metric_name])\n _results=pd.DataFrame(ll).pivot(index=2, columns=0, values=1)\n _results.columns=['Mean','STD']\n \n return(_results)", "_____no_output_____" ], [ "def cluster_plot(my_pipe, data_X, n_cluster, PALLET):\n y_pred= my_pipe.predict(data_X)\n data = pd.DataFrame(data_X[:, 0],columns=['x'])\n data['y'] = data_X[:, 1]\n data['labels'] = y_pred\n facet = sns.lmplot(data=data, x='x', y='y', hue='labels', \n aspect= 1.0, height=10,\n fit_reg=False, legend=True, legend_out=True)\n\n customPalette = PALLET #*ceil((n_cluster/N_PALLET +2))\n for i, label in enumerate( np.sort(data['labels'].unique())):\n plt.annotate(label, \n data.loc[data['labels']==label,['x','y']].mean(),\n horizontalalignment='center',\n verticalalignment='center',\n size=5, weight='bold',\n color='white',\n backgroundcolor=customPalette[i]) \n\n plt.show()\n return y_pred", "_____no_output_____" ], [ "def simple_output(string: str, number: int) -> None:\n print(string, number)\n logger.info(string, number )", "_____no_output_____" ], [ "__file__ = \"exp1.log\"\nbase_folder = os.path.dirname(os.path.abspath(''))\ncustom_elements_folder = os.path.join(base_folder, 'custom_elements')\ncustom_elements_folder", "_____no_output_____" ], [ "registry = PhotonRegistry(custom_elements_folder=custom_elements_folder)\nregistry.activate()\nregistry.PHOTON_REGISTRIES,PhotonRegistry.PHOTON_REGISTRIES", "_____no_output_____" ], [ "registry.activate()\nregistry.list_available_elements()\n# take off last name", "\nPhotonCore\nARDRegression sklearn.linear_model.ARDRegression Estimator\nAdaBoostClassifier sklearn.ensemble.AdaBoostClassifier Estimator\nAdaBoostRegressor sklearn.ensemble.AdaBoostRegressor Estimator\nBaggingClassifier sklearn.ensemble.BaggingClassifier Estimator\nBaggingRegressor sklearn.ensemble.BaggingRegressor Estimator\nBayesianGaussianMixture sklearn.mixture.BayesianGaussianMixture Estimator\nBayesianRidge sklearn.linear_model.BayesianRidge Estimator\nBernoulliNB sklearn.naive_bayes.BernoulliNB Estimator\nBernoulliRBM sklearn.neural_network.BernoulliRBM Estimator\nBinarizer sklearn.preprocessing.Binarizer Transformer\nCCA sklearn.cross_decomposition.CCA Transformer\nConfounderRemoval photonai.modelwrapper.ConfounderRemoval.ConfounderRemoval Transformer\nDecisionTreeClassifier sklearn.tree.DecisionTreeClassifier Estimator\nDecisionTreeRegressor sklearn.tree.DecisionTreeRegressor Estimator\nDictionaryLearning sklearn.decomposition.DictionaryLearning Transformer\nDummyClassifier sklearn.dummy.DummyClassifier Estimator\nDummyRegressor sklearn.dummy.DummyRegressor Estimator\nElasticNet sklearn.linear_model.ElasticNet Estimator\nExtraDecisionTreeClassifier sklearn.tree.ExtraDecisionTreeClassifier Estimator\nExtraDecisionTreeRegressor sklearn.tree.ExtraDecisionTreeRegressor Estimator\nExtraTreesClassifier sklearn.ensemble.ExtraTreesClassifier Estimator\nExtraTreesRegressor sklearn.ensemble.ExtraTreesRegressor Estimator\nFClassifSelectPercentile photonai.modelwrapper.FeatureSelection.FClassifSelectPercentile Transformer\nFRegressionFilterPValue photonai.modelwrapper.FeatureSelection.FRegressionFilterPValue Transformer\nFRegressionSelectPercentile photonai.modelwrapper.FeatureSelection.FRegressionSelectPercentile Transformer\nFactorAnalysis sklearn.decomposition.FactorAnalysis Transformer\nFastICA sklearn.decomposition.FastICA Transformer\nFeatureEncoder photonai.modelwrapper.OrdinalEncoder.FeatureEncoder Transformer\nFunctionTransformer sklearn.preprocessing.FunctionTransformer Transformer\nGaussianMixture sklearn.mixture.GaussianMixture Estimator\nGaussianNB sklearn.naive_bayes.GaussianNB Estimator\nGaussianProcessClassifier sklearn.gaussian_process.GaussianProcessClassifier Estimator\nGaussianProcessRegressor sklearn.gaussian_process.GaussianProcessRegressor Estimator\nGenericUnivariateSelect sklearn.feature_selection.GenericUnivariateSelect Transformer\nGradientBoostingClassifier sklearn.ensemble.GradientBoostingClassifier Estimator\nGradientBoostingRegressor sklearn.ensemble.GradientBoostingRegressor Estimator\nHuberRegressor sklearn.linear_model.HuberRegressor Estimator\nImbalancedDataTransformer photonai.modelwrapper.imbalanced_data_transformer.ImbalancedDataTransformer Transformer\nIncrementalPCA sklearn.decomposition.IncrementalPCA Transformer\nKNeighborsClassifier sklearn.neighbors.KNeighborsClassifier Estimator\nKNeighborsRegressor sklearn.neighbors.KNeighborsRegressor Estimator\nKerasBaseClassifier photonai.modelwrapper.keras_base_models.KerasBaseClassifier Estimator\nKerasBaseRegression photonai.modelwrapper.keras_base_models.KerasBaseRegression Estimator\nKerasDnnClassifier photonai.modelwrapper.keras_dnn_classifier.KerasDnnClassifier Estimator\nKerasDnnRegressor photonai.modelwrapper.keras_dnn_regressor.KerasDnnRegressor Estimator\nKernelCenterer sklearn.preprocessing.KernelCenterer Transformer\nKernelPCA sklearn.decomposition.KernelPCA Transformer\nKernelRidge sklearn.kernel_ridge.KernelRidge Estimator\nLabelEncoder photonai.modelwrapper.LabelEncoder.LabelEncoder Transformer\nLars sklearn.linear_model.Lars Estimator\nLasso sklearn.linear_model.Lasso Estimator\nLassoFeatureSelection photonai.modelwrapper.FeatureSelection.LassoFeatureSelection Transformer\nLassoLars sklearn.linear_model.LassoLars Estimator\nLatentDirichletAllocation sklearn.decomposition.LatentDirichletAllocation Transformer\nLinearRegression sklearn.linear_model.LinearRegression Estimator\nLinearSVC sklearn.svm.LinearSVC Estimator\nLinearSVR sklearn.svm.LinearSVR Estimator\nLogisticRegression sklearn.linear_model.LogisticRegression Estimator\nMLPClassifier sklearn.neural_network.MLPClassifier Estimator\nMLPRegressor sklearn.neural_network.MLPRegressor Estimator\nMaxAbsScaler sklearn.preprocessing.MaxAbsScaler Transformer\nMinMaxScaler sklearn.preprocessing.MinMaxScaler Transformer\nMiniBatchDictionaryLearning sklearn.decomposition.MiniBatchDictionaryLearning Transformer\nMiniBatchSparsePCA sklearn.decomposition.MiniBatchSparsePCA Transformer\nMultinomialNB sklearn.naive_bayes.MultinomialNB Estimator\nNMF sklearn.decompositcion.NMF Transformer\nNearestCentroid sklearn.neighbors.NearestCentroid Estimator\nNormalizer sklearn.preprocessing.Normalizer Transformer\nNuSVC sklearn.svm.NuSVC Estimator\nNuSVR sklearn.svm.NuSVR Estimator\nOneClassSVM sklearn.svm.OneClassSVM Estimator\nPCA sklearn.decomposition.PCA Transformer\nPLSCanonical sklearn.cross_decomposition.PLSCanonical Transformer\nPLSRegression sklearn.cross_decomposition.PLSRegression Transformer\nPLSSVD sklearn.cross_decomposition.PLSSVD Transformer\nPassiveAggressiveClassifier sklearn.linear_model.PassiveAggressiveClassifier Estimator\nPassiveAggressiveRegressor sklearn.linear_model.PassiveAggressiveRegressor Estimator\nPerceptron sklearn.linear_model.Perceptron Estimator\nPhotonMLPClassifier photonai.modelwrapper.PhotonMLPClassifier.PhotonMLPClassifier Estimator\nPhotonOneClassSVM photonai.modelwrapper.PhotonOneClassSVM.PhotonOneClassSVM Estimator\nPhotonTestXPredictor photonai.test.processing_tests.results_tests.XPredictor Estimator\nPhotonVotingClassifier photonai.modelwrapper.Voting.PhotonVotingClassifier Estimator\nPhotonVotingRegressor photonai.modelwrapper.Voting.PhotonVotingRegressor Estimator\nPolynomialFeatures sklearn.preprocessing.PolynomialFeatures Transformer\nPowerTransformer sklearn.preprocessing.PowerTransformer Transformer\nQuantileTransformer sklearn.preprocessing.QuantileTransformer Transformer\nRANSACRegressor sklearn.linear_model.RANSACRegressor Estimator\nRFE sklearn.feature_selection.RFE Transformer\nRFECV sklearn.feature_selection.RFECV Transformer\nRadiusNeighborsClassifier sklearn.neighbors.RadiusNeighborsClassifier Estimator\nRadiusNeighborsRegressor sklearn.neighbors.RadiusNeighborsRegressor Estimator\nRandomForestClassifier sklearn.ensemble.RandomForestClassifier Estimator\nRandomForestRegressor sklearn.ensemble.RandomForestRegressor Estimator\nRandomTreesEmbedding sklearn.ensemble.RandomTreesEmbedding Transformer\nRangeRestrictor photonai.modelwrapper.RangeRestrictor.RangeRestrictor Estimator\nRidge sklearn.linear_model.Ridge Estimator\nRidgeClassifier sklearn.linear_model.RidgeClassifier Estimator\nRobustScaler sklearn.preprocessing.RobustScaler Transformer\nSGDClassifier sklearn.linear_model.SGDClassifier Estimator\nSGDRegressor sklearn.linear_model.SGDRegressor Estimator\nSVC sklearn.svm.SVC Estimator\nSVR sklearn.svm.SVR Estimator\nSamplePairingClassification photonai.modelwrapper.SamplePairing.SamplePairingClassification Transformer\nSamplePairingRegression photonai.modelwrapper.SamplePairing.SamplePairingRegression Transformer\nSelectFdr sklearn.feature_selection.SelectFdr Transformer\nSelectFpr sklearn.feature_selection.SelectFpr Transformer\nSelectFromModel sklearn.feature_selection.SelectFromModel Transformer\nSelectFwe sklearn.feature_selection.SelectFwe Transformer\nSelectKBest sklearn.feature_selection.SelectKBest Transformer\nSelectPercentile sklearn.feature_selection.SelectPercentile Transformer\nSimpleImputer sklearn.impute.SimpleImputer Transformer\nSourceSplitter photonai.modelwrapper.source_splitter.SourceSplitter Transformer\nSparseCoder sklearn.decomposition.SparseCoder Transformer\nSparsePCA sklearn.decomposition.SparsePCA Transformer\nStandardScaler sklearn.preprocessing.StandardScaler Transformer\nTheilSenRegressor sklearn.linear_model.TheilSenRegressor Estimator\nTruncatedSVD sklearn.decomposition.TruncatedSVD Transformer\nVarianceThreshold sklearn.feature_selection.VarianceThreshold Transformer\ndict_learning sklearn.decomposition.dict_learning Transformer\ndict_learning_online sklearn.decomposition.dict_learning_online Transformer\nfastica sklearn.decomposition.fastica Transformer\nsparse_encode sklearn.decomposition.sparse_encode Transformer\n\nPhotonCluster\nDBSCAN sklearn.cluster.DBSCAN Estimator\nKMeans sklearn.cluster.KMeans Estimator\nKMedoids sklearn_extra.cluster.KMedoids Estimator\n\nPhotonNeuro\nBrainAtlas photonai.neuro.brain_atlas.BrainAtlas Transformer\nBrainMask photonai.neuro.brain_atlas.BrainMask Transformer\nPatchImages photonai.neuro.nifti_transformations.PatchImages Transformer\nResampleImages photonai.neuro.nifti_transformations.ResampleImages Transformer\nSmoothImages photonai.neuro.nifti_transformations.SmoothImages Transformer\n" ] ], [ [ "## KMeans blobs", "_____no_output_____" ] ], [ [ "registry.info(CLUSTER_ALGO)", "----------------------------------\nName: KMeans\nNamespace: sklearn.cluster\n----------------------------------\nPossible Hyperparameters as derived from constructor:\nn_clusters n_clusters=8 \ninit init='k-means++' \nn_init n_init=10 \nmax_iter max_iter=300 \ntol tol=0.0001 \nprecompute_distances precompute_distances='auto' \nverbose verbose=0 \nrandom_state random_state=None \ncopy_x copy_x=True \nn_jobs n_jobs=None \nalgorithm algorithm='auto' \n----------------------------------\n" ], [ "def hyper_cluster(cluster_name):\n if C_SHAPE == 'ellipse' :\n yield_cluster = yield_parameters_ellipse\n else: \n yield_cluster = yield_parameters\n \n n_p_clusters = N_P_CLUSTERS\n for data_X, data_y,n_cluster in yield_cluster(n_p_clusters):\n simple_output('CLUSTER_ALGO:', CLUSTER_ALGO)\n simple_output('C_SHAPE:',C_SHAPE)\n simple_output('n_cluster:', n_cluster)\n simple_output('CLUSTERS_STD:', CLUSTERS_STD)\n \n simple_output('INNER_FOLDS:', INNER_FOLDS)\n simple_output('OUTER_FOLDS:', OUTER_FOLDS) \n simple_output('n_points:', len(data_y))\n\n X = data_X.copy(); y = data_y.copy()\n # DESIGN YOUR PIPELINE\n settings = OutputSettings(project_folder='./tmp/')\n \n my_pipe = Hyperpipe('batching',\n optimizer='sk_opt',\n # optimizer_params={'n_configurations': 25},\n metrics=['ARI', 'MI', 'HCV', 'FM'],\n best_config_metric='ARI',\n outer_cv=KFold(n_splits=OUTER_FOLDS),\n inner_cv=KFold(n_splits=INNER_FOLDS),\n verbosity=0,\n output_settings=settings)\n\n\n my_pipe += PipelineElement(cluster_name, hyperparameters={\n 'n_clusters': IntegerRange(floor(n_cluster*.7)\n , ceil(n_cluster*1.2)),\n },random_state=777)\n\n logger.info('Cluster optimization range:', floor(n_cluster*.7), ceil(n_cluster*1.2))\n print('Cluster optimization range:', floor(n_cluster*.7), ceil(n_cluster*1.2)) \n\n # NOW TRAIN YOUR PIPELINE\n my_pipe.fit(X, y)\n\n debug = True\n #------------------------------plot\n y_pred=cluster_plot(my_pipe, X, n_cluster, PALLET)\n #--------------------------------- best\n print(pd.DataFrame(my_pipe.best_config.items()\n ,columns=['n_clusters', 'k']))\n #------------------------------\n print('train','\\n'\n ,results_to_df(my_pipe.results.metrics_train))\n print('test','\\n'\n ,results_to_df(my_pipe.results.metrics_test))\n #------------------------------ \n # turn the ground-truth labels into a clusim Clustering\n true_clustering = Clustering().from_membership_list(y) \n kmeans_clustering = Clustering().from_membership_list(y_pred) # lets see how similar the predicted k-means clustering is to the true clustering\n #------------------------------\n # using all available similar measures!\n row_format2 =\"{:>25}\" * (2)\n for simfunc in sim.available_similarity_measures:\n print(row_format2.format(simfunc, eval('sim.' + simfunc+'(true_clustering, kmeans_clustering)')))\n #------------------------------# The element-centric similarity is particularly useful for understanding\n # how a clustering method performed\n\n # Let's start with the single similarity value:\n elsim = sim.element_sim(true_clustering, kmeans_clustering)\n print(\"Element-centric similarity: {}\".format(elsim))", "_____no_output_____" ], [ "hyper_cluster(CLUSTER_ALGO)", "ncluster: 21\nCLUSTER_ALGO: KMeans\nC_SHAPE: ellipse\nn_cluster: 21\nCLUSTERS_STD: [0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3]\nINNER_FOLDS: 3\nOUTER_FOLDS: 3\nn_points: 63\nCluster optimization range: 14 26\n***************************************************************************************************************\nPHOTON ANALYSIS: batching\n***************************************************************************************************************\n\n********************************************************\nOuter Cross validation Fold 1\n********************************************************\n---------------------------------------------------------------------------------------------------------------\nBEST_CONFIG \n---------------------------------------------------------------------------------------------------------------\n{\n \"KMeans\": [\n \"n_clusters=22\"\n ]\n}\n+--------+-------------------+------------------+\n| METRIC | PERFORMANCE TRAIN | PERFORMANCE TEST |\n+--------+-------------------+------------------+\n| ARI | 0.9394 | 0.8286 |\n| MI | 0.9096 | 0.7782 |\n| HCV | 0.9889 | 0.9665 |\n| FM | 0.9414 | 0.8452 |\n+--------+-------------------+------------------+\n\n********************************************************\nOuter Cross validation Fold 2\n********************************************************\n---------------------------------------------------------------------------------------------------------------\nBEST_CONFIG \n---------------------------------------------------------------------------------------------------------------\n{\n \"KMeans\": [\n \"n_clusters=19\"\n ]\n}\n+--------+-------------------+------------------+\n| METRIC | PERFORMANCE TRAIN | PERFORMANCE TEST |\n+--------+-------------------+------------------+\n| ARI | 0.8662 | 0.7407 |\n| MI | 0.8535 | 0.6747 |\n| HCV | 0.9577 | 0.9313 |\n| FM | 0.8783 | 0.7746 |\n+--------+-------------------+------------------+\n\n********************************************************\nOuter Cross validation Fold 3\n********************************************************\n---------------------------------------------------------------------------------------------------------------\nBEST_CONFIG \n---------------------------------------------------------------------------------------------------------------\n{\n \"KMeans\": [\n \"n_clusters=21\"\n ]\n}\n+--------+-------------------+------------------+\n| METRIC | PERFORMANCE TRAIN | PERFORMANCE TEST |\n+--------+-------------------+------------------+\n| ARI | 0.8929 | 0.7501 |\n| MI | 0.8996 | 0.6965 |\n| HCV | 0.9817 | 0.9226 |\n| FM | 0.8966 | 0.7845 |\n+--------+-------------------+------------------+\n\n===============================================================================================================\nOVERALL BEST CONFIGURATION\n===============================================================================================================\n{\n \"KMeans\": [\n \"n_clusters=22\"\n ]\n}\n\nAnalysis batching done in 0:00:47.725834\nYour results are stored in ./tmp/batching_results_2020-06-10_22-36-20\n***************************************************************************************************************\nPHOTON 1.0.0b - www.photon-ai.com \n n_clusters k\n0 KMeans__n_clusters 22\ntrain \n Mean STD\n2 \nARI 0.899520 0.030223\nFM 0.905406 0.026490\nHCV 0.976110 0.013341\nMI 0.887575 0.024422\ntest \n Mean STD\n2 \nARI 0.773144 0.039380\nFM 0.801405 0.031196\nHCV 0.940133 0.018986\nMI 0.716476 0.044554\n jaccard_index 0.7857142857142857\n rand_index 0.9923195084485407\n adjrand_index 0.8760330578512392\n fowlkes_mallows_index 0.8800281613517521\n fmeasure 0.88\n purity_index 0.9365079365079365\n classification_error 0.06349206349206349\n czekanowski_index 0.88\n dice_index 0.88\n sorensen_index 0.88\n rogers_tanimoto_index 0.9847560975609756\n southwood_index 3.6666666666666665\n pearson_correlation 7.4146821899333795e-06\n corrected_chance 0.7821678434833333\n sample_expected_sim 0.03305785123966942\n nmi 0.9649080353388753\n mi 4.257551946656966\n adj_mi 0.8968861958695907\n rmi -0.2443496273954531\n vi 0.3096789683226495\n geometric_accuracy 0.9444110971837153\n overlap_quality -0.0\n onmi 0.855604621696782\n omega_index 0.876033057851239\nElement-centric similarity: 0.8888888888888888\nncluster: 21\nCLUSTER_ALGO: KMeans\nC_SHAPE: ellipse\nn_cluster: 21\nCLUSTERS_STD: [0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3, 0.3]\nINNER_FOLDS: 3\nOUTER_FOLDS: 3\nn_points: 630\nCluster optimization range: 14 26\n***************************************************************************************************************\nPHOTON ANALYSIS: batching\n***************************************************************************************************************\n\n********************************************************\nOuter Cross validation Fold 1\n********************************************************\n---------------------------------------------------------------------------------------------------------------\nBEST_CONFIG \n---------------------------------------------------------------------------------------------------------------\n{\n \"KMeans\": [\n \"n_clusters=21\"\n ]\n}\n+--------+-------------------+------------------+\n| METRIC | PERFORMANCE TRAIN | PERFORMANCE TEST |\n+--------+-------------------+------------------+\n| ARI | 0.8998 | 0.8875 |\n| MI | 0.9308 | 0.9250 |\n| HCV | 0.9435 | 0.9489 |\n| FM | 0.9045 | 0.8927 |\n+--------+-------------------+------------------+\n\n********************************************************\nOuter Cross validation Fold 2\n********************************************************\n---------------------------------------------------------------------------------------------------------------\nBEST_CONFIG \n---------------------------------------------------------------------------------------------------------------\n{\n \"KMeans\": [\n \"n_clusters=24\"\n ]\n}\n+--------+-------------------+------------------+\n| METRIC | PERFORMANCE TRAIN | PERFORMANCE TEST |\n+--------+-------------------+------------------+\n| ARI | 0.8690 | 0.8529 |\n| MI | 0.9056 | 0.8864 |\n| HCV | 0.9519 | 0.9513 |\n| FM | 0.8753 | 0.8598 |\n+--------+-------------------+------------------+\n\n********************************************************\nOuter Cross validation Fold 3\n********************************************************\n---------------------------------------------------------------------------------------------------------------\nBEST_CONFIG \n---------------------------------------------------------------------------------------------------------------\n{\n \"KMeans\": [\n \"n_clusters=21\"\n ]\n}\n+--------+-------------------+------------------+\n| METRIC | PERFORMANCE TRAIN | PERFORMANCE TEST |\n+--------+-------------------+------------------+\n| ARI | 0.9053 | 0.9114 |\n| MI | 0.9424 | 0.9310 |\n| HCV | 0.9530 | 0.9530 |\n| FM | 0.9097 | 0.9155 |\n+--------+-------------------+------------------+\n" ], [ "Cottman\n*-+\n*+-=END", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ] ]

ec7d997a99935a6c3d68f4f748ec07a443fda17d

1,359

ipynb

Jupyter Notebook

docs/_build/.jupyter_cache/executed/f0dd264295c7f662e35058309d74ceb1/base.ipynb

cancermqiao/CancerMBook

bd26c0e3e1f76f66b75aacf75b3cb8602715e803

[ "Apache-2.0" ]

null

docs/_build/.jupyter_cache/executed/f0dd264295c7f662e35058309d74ceb1/base.ipynb

cancermqiao/CancerMBook

bd26c0e3e1f76f66b75aacf75b3cb8602715e803

[ "Apache-2.0" ]

null

docs/_build/.jupyter_cache/executed/f0dd264295c7f662e35058309d74ceb1/base.ipynb

cancermqiao/CancerMBook

bd26c0e3e1f76f66b75aacf75b3cb8602715e803

[ "Apache-2.0" ]

null

20.907692

42

0.580574

[]

ec7d9f4f1e53b5de27fc5ead01c99d8b821a2132

301,173

ipynb

Jupyter Notebook

Project/SageMaker Project.ipynb

alluluA/sagemaker-dep

2e99db7f94e33785f980bc7aea817ecf25babecc

[ "MIT" ]

null

Project/SageMaker Project.ipynb

alluluA/sagemaker-dep

2e99db7f94e33785f980bc7aea817ecf25babecc

[ "MIT" ]

null

Project/SageMaker Project.ipynb

alluluA/sagemaker-dep

2e99db7f94e33785f980bc7aea817ecf25babecc

[ "MIT" ]

null

95.217515

168,168

0.787826

[ [ [ "# Creating a Sentiment Analysis Web App\n## Using PyTorch and SageMaker\n\n_Deep Learning Nanodegree Program | Deployment_\n\n---\n\nNow that we have a basic understanding of how SageMaker works we will try to use it to construct a complete project from end to end. Our goal will be to have a simple web page which a user can use to enter a movie review. The web page will then send the review off to our deployed model which will predict the sentiment of the entered review.\n\n## Instructions\n\nSome template code has already been provided for you, and you will need to implement additional functionality to successfully complete this notebook. You will not need to modify the included code beyond what is requested. Sections that begin with '**TODO**' in the header indicate that you need to complete or implement some portion within them. Instructions will be provided for each section and the specifics of the implementation are marked in the code block with a `# TODO: ...` comment. Please be sure to read the instructions carefully!\n\nIn addition to implementing code, there will be questions for you to answer which relate to the task and your implementation. Each section where you will answer a question is preceded by a '**Question:**' header. Carefully read each question and provide your answer below the '**Answer:**' header by editing the Markdown cell.\n\n> **Note**: Code and Markdown cells can be executed using the **Shift+Enter** keyboard shortcut. In addition, a cell can be edited by typically clicking it (double-click for Markdown cells) or by pressing **Enter** while it is highlighted.\n\n## General Outline\n\nRecall the general outline for SageMaker projects using a notebook instance.\n\n1. Download or otherwise retrieve the data.\n2. Process / Prepare the data.\n3. Upload the processed data to S3.\n4. Train a chosen model.\n5. Test the trained model (typically using a batch transform job).\n6. Deploy the trained model.\n7. Use the deployed model.\n\nFor this project, you will be following the steps in the general outline with some modifications. \n\nFirst, you will not be testing the model in its own step. You will still be testing the model, however, you will do it by deploying your model and then using the deployed model by sending the test data to it. One of the reasons for doing this is so that you can make sure that your deployed model is working correctly before moving forward.\n\nIn addition, you will deploy and use your trained model a second time. In the second iteration you will customize the way that your trained model is deployed by including some of your own code. In addition, your newly deployed model will be used in the sentiment analysis web app.", "_____no_output_____" ] ], [ [ "# Make sure that we use SageMaker 1.x\n!pip install sagemaker==1.72.0", "Collecting sagemaker==1.72.0\n Using cached sagemaker-1.72.0-py2.py3-none-any.whl\nRequirement already satisfied: importlib-metadata>=1.4.0 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from sagemaker==1.72.0) (4.8.2)\nRequirement already satisfied: scipy>=0.19.0 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from sagemaker==1.72.0) (1.5.3)\nRequirement already satisfied: numpy>=1.9.0 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from sagemaker==1.72.0) (1.19.5)\nRequirement already satisfied: boto3>=1.14.12 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from sagemaker==1.72.0) (1.20.4)\nRequirement already satisfied: packaging>=20.0 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from sagemaker==1.72.0) (21.2)\nRequirement already satisfied: smdebug-rulesconfig==0.1.4 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from sagemaker==1.72.0) (0.1.4)\nRequirement already satisfied: protobuf3-to-dict>=0.1.5 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from sagemaker==1.72.0) (0.1.5)\nRequirement already satisfied: protobuf>=3.1 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from sagemaker==1.72.0) (3.19.1)\nRequirement already satisfied: s3transfer<0.6.0,>=0.5.0 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from boto3>=1.14.12->sagemaker==1.72.0) (0.5.0)\nRequirement already satisfied: jmespath<1.0.0,>=0.7.1 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from boto3>=1.14.12->sagemaker==1.72.0) (0.10.0)\nRequirement already satisfied: botocore<1.24.0,>=1.23.4 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from boto3>=1.14.12->sagemaker==1.72.0) (1.23.4)\nRequirement already satisfied: python-dateutil<3.0.0,>=2.1 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from botocore<1.24.0,>=1.23.4->boto3>=1.14.12->sagemaker==1.72.0) (2.8.2)\nRequirement already satisfied: urllib3<1.27,>=1.25.4 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from botocore<1.24.0,>=1.23.4->boto3>=1.14.12->sagemaker==1.72.0) (1.26.7)\nRequirement already satisfied: typing-extensions>=3.6.4 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from importlib-metadata>=1.4.0->sagemaker==1.72.0) (3.10.0.2)\nRequirement already satisfied: zipp>=0.5 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from importlib-metadata>=1.4.0->sagemaker==1.72.0) (3.6.0)\nRequirement already satisfied: pyparsing<3,>=2.0.2 in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from packaging>=20.0->sagemaker==1.72.0) (2.4.7)\nRequirement already satisfied: six in /home/ec2-user/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages (from protobuf3-to-dict>=0.1.5->sagemaker==1.72.0) (1.16.0)\nInstalling collected packages: sagemaker\n Attempting uninstall: sagemaker\n Found existing installation: sagemaker 2.60.0\n Uninstalling sagemaker-2.60.0:\n Successfully uninstalled sagemaker-2.60.0\nSuccessfully installed sagemaker-1.72.0\n\u001b[33mWARNING: You are using pip version 21.2.4; however, version 21.3.1 is available.\nYou should consider upgrading via the '/home/ec2-user/anaconda3/envs/pytorch_p36/bin/python -m pip install --upgrade pip' command.\u001b[0m\n" ] ], [ [ "## Step 1: Downloading the data\n\nAs in the XGBoost in SageMaker notebook, we will be using the [IMDb dataset](http://ai.stanford.edu/~amaas/data/sentiment/)\n\n> Maas, Andrew L., et al. [Learning Word Vectors for Sentiment Analysis](http://ai.stanford.edu/~amaas/data/sentiment/). In _Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies_. Association for Computational Linguistics, 2011.", "_____no_output_____" ] ], [ [ "%mkdir ../data\n!wget -O ../data/aclImdb_v1.tar.gz http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz\n!tar -zxf ../data/aclImdb_v1.tar.gz -C ../data", "mkdir: cannot create directory ‘../data’: File exists\n--2021-11-20 19:35:13-- http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz\nResolving ai.stanford.edu (ai.stanford.edu)... 171.64.68.10\nConnecting to ai.stanford.edu (ai.stanford.edu)|171.64.68.10|:80... connected.\nHTTP request sent, awaiting response... 200 OK\nLength: 84125825 (80M) [application/x-gzip]\nSaving to: ‘../data/aclImdb_v1.tar.gz’\n\n../data/aclImdb_v1. 100%[===================>] 80.23M 23.5MB/s in 4.8s \n\n2021-11-20 19:35:18 (16.8 MB/s) - ‘../data/aclImdb_v1.tar.gz’ saved [84125825/84125825]\n\n" ] ], [ [ "## Step 2: Preparing and Processing the data\n\nAlso, as in the XGBoost notebook, we will be doing some initial data processing. The first few steps are the same as in the XGBoost example. To begin with, we will read in each of the reviews and combine them into a single input structure. Then, we will split the dataset into a training set and a testing set.", "_____no_output_____" ] ], [ [ "import os\nimport glob\n\ndef read_imdb_data(data_dir='../data/aclImdb'):\n data = {}\n labels = {}\n \n for data_type in ['train', 'test']:\n data[data_type] = {}\n labels[data_type] = {}\n \n for sentiment in ['pos', 'neg']:\n data[data_type][sentiment] = []\n labels[data_type][sentiment] = []\n \n path = os.path.join(data_dir, data_type, sentiment, '*.txt')\n files = glob.glob(path)\n \n for f in files:\n with open(f) as review:\n data[data_type][sentiment].append(review.read())\n # Here we represent a positive review by '1' and a negative review by '0'\n labels[data_type][sentiment].append(1 if sentiment == 'pos' else 0)\n \n assert len(data[data_type][sentiment]) == len(labels[data_type][sentiment]), \\\n \"{}/{} data size does not match labels size\".format(data_type, sentiment)\n \n return data, labels", "_____no_output_____" ], [ "data, labels = read_imdb_data()\nprint(\"IMDB reviews: train = {} pos / {} neg, test = {} pos / {} neg\".format(\n len(data['train']['pos']), len(data['train']['neg']),\n len(data['test']['pos']), len(data['test']['neg'])))", "IMDB reviews: train = 12500 pos / 12500 neg, test = 12500 pos / 12500 neg\n" ] ], [ [ "Now that we've read the raw training and testing data from the downloaded dataset, we will combine the positive and negative reviews and shuffle the resulting records.", "_____no_output_____" ] ], [ [ "from sklearn.utils import shuffle\n\ndef prepare_imdb_data(data, labels):\n \"\"\"Prepare training and test sets from IMDb movie reviews.\"\"\"\n \n #Combine positive and negative reviews and labels\n data_train = data['train']['pos'] + data['train']['neg']\n data_test = data['test']['pos'] + data['test']['neg']\n labels_train = labels['train']['pos'] + labels['train']['neg']\n labels_test = labels['test']['pos'] + labels['test']['neg']\n \n #Shuffle reviews and corresponding labels within training and test sets\n data_train, labels_train = shuffle(data_train, labels_train)\n data_test, labels_test = shuffle(data_test, labels_test)\n \n # Return a unified training data, test data, training labels, test labets\n return data_train, data_test, labels_train, labels_test", "_____no_output_____" ], [ "train_X, test_X, train_y, test_y = prepare_imdb_data(data, labels)\nprint(\"IMDb reviews (combined): train = {}, test = {}\".format(len(train_X), len(test_X)))", "IMDb reviews (combined): train = 25000, test = 25000\n" ] ], [ [ "Now that we have our training and testing sets unified and prepared, we should do a quick check and see an example of the data our model will be trained on. This is generally a good idea as it allows you to see how each of the further processing steps affects the reviews and it also ensures that the data has been loaded correctly.", "_____no_output_____" ] ], [ [ "print(train_X[100])\nprint(train_y[100])", "Four prisoners share a single cell: the domineering transvestite, Marcus (Clovis Cornillac); Marcus's idiot savant buddy, Paquerette (Dimitri Rataud), who will eat anything in sight including pocket watches, cockroaches, and his little sister; Lassalle (Philippe Laudenbach), the intelligent librarian who murdered his wife; and Carrère (Gérald Laroche), the new guy who was caught up in corporate fraud and is now focused on escaping. After a brick falls from the wall of the cell, the men discover the hidden journal written by a 'Fountain of Youth'-obsessed serial killer who occupied the cell in the 1920s. Is this journal the secret to their escape? Or is there something much more sinister behind it?<br /><br />I was a little weary about getting into this film because the only other experience I have with Eric Valette was the dreadful One Missed Call (2008), which I consider to be the worst theatrically released film I've ever seen. However, much of what was wrong with One Missed Call could probably be attributed to Klavan's awful script, because (as I remember) Valette's direction wasn't the worst part about the film (unless he chose to include the baby). Anyway, Maléfique was a good way to get my respect back. . . it's a French film (obviously something I like) and it takes place in prison (which is my second favourite horror setting after asylums). So that's two points for him before the film even starts. Luckily, Valette had me once the film ended as well. Maléfique is a rather deep, rather complex, rather compelling story of obsession and desperation. . . the desire and need to bring fantasies to reality. While it's not a terrifying film in the traditional sense, the oddity of its power makes it pretty damn frightening. The period between the climax and conclusion was some of the best film I've seen in quite some time and I would wholeheartedly recommend this to anyone who is looking for a decent psychological thriller with some pretty cool gore.<br /><br />Final verdict: 8.5/10. Quite a bit of respect earned back by Valette.<br /><br />Note: Paramount picked up the rights to make an American remake (surprise surprise). It's due out in 2009. I'm not sure why, to be honest, as this doesn't seem like something that would be a big moneymaker here in the states. But, I've been surprised before.<br /><br />Vive La France! <br /><br />-AP3-\n1\n" ] ], [ [ "The first step in processing the reviews is to make sure that any html tags that appear should be removed. In addition we wish to tokenize our input, that way words such as *entertained* and *entertaining* are considered the same with regard to sentiment analysis.", "_____no_output_____" ] ], [ [ "import nltk\nfrom nltk.corpus import stopwords\nfrom nltk.stem.porter import *\n\nimport re\nfrom bs4 import BeautifulSoup\n\ndef review_to_words(review):\n nltk.download(\"stopwords\", quiet=True)\n stemmer = PorterStemmer()\n \n text = BeautifulSoup(review, \"html.parser\").get_text() # Remove HTML tags\n text = re.sub(r\"[^a-zA-Z0-9]\", \" \", text.lower()) # Convert to lower case\n words = text.split() # Split string into words\n words = [w for w in words if w not in stopwords.words(\"english\")] # Remove stopwords\n words = [PorterStemmer().stem(w) for w in words] # stem\n \n return words", "_____no_output_____" ] ], [ [ "The `review_to_words` method defined above uses `BeautifulSoup` to remove any html tags that appear and uses the `nltk` package to tokenize the reviews. As a check to ensure we know how everything is working, try applying `review_to_words` to one of the reviews in the training set.", "_____no_output_____" ] ], [ [ "# TODO: Apply review_to_words to a review (train_X[100] or any other review)\nreview_to_words(train_X[100])", "_____no_output_____" ] ], [ [ "**Question:** Above we mentioned that `review_to_words` method removes html formatting and allows us to tokenize the words found in a review, for example, converting *entertained* and *entertaining* into *entertain* so that they are treated as though they are the same word. What else, if anything, does this method do to the input?", "_____no_output_____" ], [ "**Answer:** the method not only removes html tags and get us the review text, but it also performs tokenization which is the task of chopping sentences up into pieces, called tokens, perhaps at the same time throwing away certain characters, such as punctuation, stopping words and it changes upper-case characters inside a word to lower case.", "_____no_output_____" ], [ "The method below applies the `review_to_words` method to each of the reviews in the training and testing datasets. In addition it caches the results. This is because performing this processing step can take a long time. This way if you are unable to complete the notebook in the current session, you can come back without needing to process the data a second time.", "_____no_output_____" ] ], [ [ "import pickle\n\ncache_dir = os.path.join(\"../cache\", \"sentiment_analysis\") # where to store cache files\nos.makedirs(cache_dir, exist_ok=True) # ensure cache directory exists\n\ndef preprocess_data(data_train, data_test, labels_train, labels_test,\n cache_dir=cache_dir, cache_file=\"preprocessed_data.pkl\"):\n \"\"\"Convert each review to words; read from cache if available.\"\"\"\n\n # If cache_file is not None, try to read from it first\n cache_data = None\n if cache_file is not None:\n try:\n with open(os.path.join(cache_dir, cache_file), \"rb\") as f:\n cache_data = pickle.load(f)\n print(\"Read preprocessed data from cache file:\", cache_file)\n except:\n pass # unable to read from cache, but that's okay\n \n # If cache is missing, then do the heavy lifting\n if cache_data is None:\n # Preprocess training and test data to obtain words for each review\n #words_train = list(map(review_to_words, data_train))\n #words_test = list(map(review_to_words, data_test))\n words_train = [review_to_words(review) for review in data_train]\n words_test = [review_to_words(review) for review in data_test]\n \n # Write to cache file for future runs\n if cache_file is not None:\n cache_data = dict(words_train=words_train, words_test=words_test,\n labels_train=labels_train, labels_test=labels_test)\n with open(os.path.join(cache_dir, cache_file), \"wb\") as f:\n pickle.dump(cache_data, f)\n print(\"Wrote preprocessed data to cache file:\", cache_file)\n else:\n # Unpack data loaded from cache file\n words_train, words_test, labels_train, labels_test = (cache_data['words_train'],\n cache_data['words_test'], cache_data['labels_train'], cache_data['labels_test'])\n \n return words_train, words_test, labels_train, labels_test", "_____no_output_____" ], [ "# Preprocess data\ntrain_X, test_X, train_y, test_y = preprocess_data(train_X, test_X, train_y, test_y)", "Read preprocessed data from cache file: preprocessed_data.pkl\n" ] ], [ [ "## Transform the data\n\nIn the XGBoost notebook we transformed the data from its word representation to a bag-of-words feature representation. For the model we are going to construct in this notebook we will construct a feature representation which is very similar. To start, we will represent each word as an integer. Of course, some of the words that appear in the reviews occur very infrequently and so likely don't contain much information for the purposes of sentiment analysis. The way we will deal with this problem is that we will fix the size of our working vocabulary and we will only include the words that appear most frequently. We will then combine all of the infrequent words into a single category and, in our case, we will label it as `1`.\n\nSince we will be using a recurrent neural network, it will be convenient if the length of each review is the same. To do this, we will fix a size for our reviews and then pad short reviews with the category 'no word' (which we will label `0`) and truncate long reviews.", "_____no_output_____" ], [ "### (TODO) Create a word dictionary\n\nTo begin with, we need to construct a way to map words that appear in the reviews to integers. Here we fix the size of our vocabulary (including the 'no word' and 'infrequent' categories) to be `5000` but you may wish to change this to see how it affects the model.\n\n> **TODO:** Complete the implementation for the `build_dict()` method below. Note that even though the vocab_size is set to `5000`, we only want to construct a mapping for the most frequently appearing `4998` words. This is because we want to reserve the special labels `0` for 'no word' and `1` for 'infrequent word'.", "_____no_output_____" ] ], [ [ "import numpy as np\n\ndef build_dict(data, vocab_size = 5000):\n \"\"\"Construct and return a dictionary mapping each of the most frequently appearing words to a unique integer.\"\"\"\n \n # TODO: Determine how often each word appears in `data`. Note that `data` is a list of sentences and that a\n # sentence is a list of words.\n \n word_count = {} # A dict storing the words that appear in the reviews along with how often they occur\n for words in data:\n for word in words:\n if word in word_count:\n word_count[word] += 1\n else:\n word_count[word] = 1\n # TODO: Sort the words found in `data` so that sorted_words[0] is the most frequently appearing word and\n # sorted_words[-1] is the least frequently appearing word.\n \n sorted_words = sorted(word_count, key=word_count.get, reverse=True)\n \n word_dict = {} # This is what we are building, a dictionary that translates words into integers\n for idx, word in enumerate(sorted_words[:vocab_size - 2]): # The -2 is so that we save room for the 'no word'\n word_dict[word] = idx + 2 # 'infrequent' labels\n \n return word_dict", "_____no_output_____" ], [ "word_dict = build_dict(train_X)", "_____no_output_____" ], [ "word_dict", "_____no_output_____" ] ], [ [ "**Question:** What are the five most frequently appearing (tokenized) words in the training set? Does it makes sense that these words appear frequently in the training set?", "_____no_output_____" ], [ "**Answer:**", "_____no_output_____" ] ], [ [ "# TODO: Use this space to determine the five most frequently appearing words in the training set.\nfrom itertools import islice\n\ndef take(n, iterable):\n \"Return first n items of the iterable as a list\"\n return list(islice(iterable, n))\ntake(5, word_dict.keys())", "_____no_output_____" ] ], [ [ "### Save `word_dict`\n\nLater on when we construct an endpoint which processes a submitted review we will need to make use of the `word_dict` which we have created. As such, we will save it to a file now for future use.", "_____no_output_____" ] ], [ [ "data_dir = '../data/pytorch' # The folder we will use for storing data\nif not os.path.exists(data_dir): # Make sure that the folder exists\n os.makedirs(data_dir)", "_____no_output_____" ], [ "with open(os.path.join(data_dir, 'word_dict.pkl'), \"wb\") as f:\n pickle.dump(word_dict, f)", "_____no_output_____" ] ], [ [ "### Transform the reviews\n\nNow that we have our word dictionary which allows us to transform the words appearing in the reviews into integers, it is time to make use of it and convert our reviews to their integer sequence representation, making sure to pad or truncate to a fixed length, which in our case is `500`.", "_____no_output_____" ] ], [ [ "def convert_and_pad(word_dict, sentence, pad=500):\n NOWORD = 0 # We will use 0 to represent the 'no word' category\n INFREQ = 1 # and we use 1 to represent the infrequent words, i.e., words not appearing in word_dict\n \n working_sentence = [NOWORD] * pad\n \n for word_index, word in enumerate(sentence[:pad]):\n if word in word_dict:\n working_sentence[word_index] = word_dict[word]\n else:\n working_sentence[word_index] = INFREQ\n \n return working_sentence, min(len(sentence), pad)\n\ndef convert_and_pad_data(word_dict, data, pad=500):\n result = []\n lengths = []\n \n for sentence in data:\n converted, leng = convert_and_pad(word_dict, sentence, pad)\n result.append(converted)\n lengths.append(leng)\n \n return np.array(result), np.array(lengths)", "_____no_output_____" ], [ "train_X, train_X_len = convert_and_pad_data(word_dict, train_X)\ntest_X, test_X_len = convert_and_pad_data(word_dict, test_X)", "_____no_output_____" ] ], [ [ "As a quick check to make sure that things are working as intended, check to see what one of the reviews in the training set looks like after having been processeed. Does this look reasonable? What is the length of a review in the training set?", "_____no_output_____" ] ], [ [ "# Use this cell to examine one of the processed reviews to make sure everything is working as intended.\ntrain_X[100]", "_____no_output_____" ] ], [ [ "**Question:** In the cells above we use the `preprocess_data` and `convert_and_pad_data` methods to process both the training and testing set. Why or why not might this be a problem?", "_____no_output_____" ], [ "**Answer:** the method converts the reviews to their integer sequence representation, and it makes sure to each review length is less than 500, that is to uniform the representation of data before feeding it into the model.", "_____no_output_____" ], [ "## Step 3: Upload the data to S3\n\nAs in the XGBoost notebook, we will need to upload the training dataset to S3 in order for our training code to access it. For now we will save it locally and we will upload to S3 later on.\n\n### Save the processed training dataset locally\n\nIt is important to note the format of the data that we are saving as we will need to know it when we write the training code. In our case, each row of the dataset has the form `label`, `length`, `review[500]` where `review[500]` is a sequence of `500` integers representing the words in the review.", "_____no_output_____" ] ], [ [ "import pandas as pd\n \npd.concat([pd.DataFrame(train_y), pd.DataFrame(train_X_len), pd.DataFrame(train_X)], axis=1) \\\n .to_csv(os.path.join(data_dir, 'train.csv'), header=False, index=False)", "_____no_output_____" ] ], [ [ "### Uploading the training data\n\n\nNext, we need to upload the training data to the SageMaker default S3 bucket so that we can provide access to it while training our model.", "_____no_output_____" ] ], [ [ "import sagemaker\n\nsagemaker_session = sagemaker.Session()\n\nbucket = sagemaker_session.default_bucket()\nprefix = 'sagemaker/sentiment_rnn'\n\nrole = sagemaker.get_execution_role()", "_____no_output_____" ], [ "input_data = sagemaker_session.upload_data(path=data_dir, bucket=bucket, key_prefix=prefix)", "_____no_output_____" ] ], [ [ "**NOTE:** The cell above uploads the entire contents of our data directory. This includes the `word_dict.pkl` file. This is fortunate as we will need this later on when we create an endpoint that accepts an arbitrary review. For now, we will just take note of the fact that it resides in the data directory (and so also in the S3 training bucket) and that we will need to make sure it gets saved in the model directory.", "_____no_output_____" ], [ "## Step 4: Build and Train the PyTorch Model\n\nIn the XGBoost notebook we discussed what a model is in the SageMaker framework. In particular, a model comprises three objects\n\n - Model Artifacts,\n - Training Code, and\n - Inference Code,\n \neach of which interact with one another. In the XGBoost example we used training and inference code that was provided by Amazon. Here we will still be using containers provided by Amazon with the added benefit of being able to include our own custom code.\n\nWe will start by implementing our own neural network in PyTorch along with a training script. For the purposes of this project we have provided the necessary model object in the `model.py` file, inside of the `train` folder. You can see the provided implementation by running the cell below.", "_____no_output_____" ] ], [ [ "!pygmentize train/model.py", "\u001b[34mimport\u001b[39;49;00m \u001b[04m\u001b[36mtorch\u001b[39;49;00m\u001b[04m\u001b[36m.\u001b[39;49;00m\u001b[04m\u001b[36mnn\u001b[39;49;00m \u001b[34mas\u001b[39;49;00m \u001b[04m\u001b[36mnn\u001b[39;49;00m\r\n\r\n\u001b[34mclass\u001b[39;49;00m \u001b[04m\u001b[32mLSTMClassifier\u001b[39;49;00m(nn.Module):\r\n \u001b[33m\"\"\"\u001b[39;49;00m\r\n\u001b[33m This is the simple RNN model we will be using to perform Sentiment Analysis.\u001b[39;49;00m\r\n\u001b[33m \"\"\"\u001b[39;49;00m\r\n\r\n \u001b[34mdef\u001b[39;49;00m \u001b[32m__init__\u001b[39;49;00m(\u001b[36mself\u001b[39;49;00m, embedding_dim, hidden_dim, vocab_size):\r\n \u001b[33m\"\"\"\u001b[39;49;00m\r\n\u001b[33m Initialize the model by settingg up the various layers.\u001b[39;49;00m\r\n\u001b[33m \"\"\"\u001b[39;49;00m\r\n \u001b[36msuper\u001b[39;49;00m(LSTMClassifier, \u001b[36mself\u001b[39;49;00m).\u001b[32m__init__\u001b[39;49;00m()\r\n\r\n \u001b[36mself\u001b[39;49;00m.embedding = nn.Embedding(vocab_size, embedding_dim, padding_idx=\u001b[34m0\u001b[39;49;00m)\r\n \u001b[36mself\u001b[39;49;00m.lstm = nn.LSTM(embedding_dim, hidden_dim)\r\n \u001b[36mself\u001b[39;49;00m.dense = nn.Linear(in_features=hidden_dim, out_features=\u001b[34m1\u001b[39;49;00m)\r\n \u001b[36mself\u001b[39;49;00m.sig = nn.Sigmoid()\r\n \r\n \u001b[36mself\u001b[39;49;00m.word_dict = \u001b[34mNone\u001b[39;49;00m\r\n\r\n \u001b[34mdef\u001b[39;49;00m \u001b[32mforward\u001b[39;49;00m(\u001b[36mself\u001b[39;49;00m, x):\r\n \u001b[33m\"\"\"\u001b[39;49;00m\r\n\u001b[33m Perform a forward pass of our model on some input.\u001b[39;49;00m\r\n\u001b[33m \"\"\"\u001b[39;49;00m\r\n x = x.t()\r\n lengths = x[\u001b[34m0\u001b[39;49;00m,:]\r\n reviews = x[\u001b[34m1\u001b[39;49;00m:,:]\r\n embeds = \u001b[36mself\u001b[39;49;00m.embedding(reviews)\r\n lstm_out, _ = \u001b[36mself\u001b[39;49;00m.lstm(embeds)\r\n out = \u001b[36mself\u001b[39;49;00m.dense(lstm_out)\r\n out = out[lengths - \u001b[34m1\u001b[39;49;00m, \u001b[36mrange\u001b[39;49;00m(\u001b[36mlen\u001b[39;49;00m(lengths))]\r\n \u001b[34mreturn\u001b[39;49;00m \u001b[36mself\u001b[39;49;00m.sig(out.squeeze())\r\n" ] ], [ [ "The important takeaway from the implementation provided is that there are three parameters that we may wish to tweak to improve the performance of our model. These are the embedding dimension, the hidden dimension and the size of the vocabulary. We will likely want to make these parameters configurable in the training script so that if we wish to modify them we do not need to modify the script itself. We will see how to do this later on. To start we will write some of the training code in the notebook so that we can more easily diagnose any issues that arise.\n\nFirst we will load a small portion of the training data set to use as a sample. It would be very time consuming to try and train the model completely in the notebook as we do not have access to a gpu and the compute instance that we are using is not particularly powerful. However, we can work on a small bit of the data to get a feel for how our training script is behaving.", "_____no_output_____" ] ], [ [ "import torch\nimport torch.utils.data\n\n# Read in only the first 250 rows\ntrain_sample = pd.read_csv(os.path.join(data_dir, 'train.csv'), header=None, names=None, nrows=250)\n\n# Turn the input pandas dataframe into tensors\ntrain_sample_y = torch.from_numpy(train_sample[[0]].values).float().squeeze()\ntrain_sample_X = torch.from_numpy(train_sample.drop([0], axis=1).values).long()\n\n# Build the dataset\ntrain_sample_ds = torch.utils.data.TensorDataset(train_sample_X, train_sample_y)\n# Build the dataloader\ntrain_sample_dl = torch.utils.data.DataLoader(train_sample_ds, batch_size=50)", "_____no_output_____" ] ], [ [ "### (TODO) Writing the training method\n\nNext we need to write the training code itself. This should be very similar to training methods that you have written before to train PyTorch models. We will leave any difficult aspects such as model saving / loading and parameter loading until a little later.", "_____no_output_____" ] ], [ [ "def train(model, train_loader, epochs, optimizer, loss_fn, device):\n for epoch in range(1, epochs + 1):\n model.train()\n total_loss = 0\n for batch in train_loader: \n batch_X, batch_y = batch\n \n batch_X = batch_X.to(device)\n batch_y = batch_y.to(device)\n \n # TODO: Complete this train method to train the model provided.\n optimizer.zero_grad()\n output = model.forward(batch_X)\n loss = loss_fn(output, batch_y)\n loss.backward()\n optimizer.step()\n total_loss += loss.data.item()\n print(\"Epoch: {}, BCELoss: {}\".format(epoch, total_loss / len(train_loader)))", "_____no_output_____" ] ], [ [ "Supposing we have the training method above, we will test that it is working by writing a bit of code in the notebook that executes our training method on the small sample training set that we loaded earlier. The reason for doing this in the notebook is so that we have an opportunity to fix any errors that arise early when they are easier to diagnose.", "_____no_output_____" ] ], [ [ "import torch.optim as optim\nfrom train.model import LSTMClassifier\n\ndevice = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\nmodel = LSTMClassifier(32, 100, 5000).to(device)\noptimizer = optim.Adam(model.parameters())\nloss_fn = torch.nn.BCELoss()\n\ntrain(model, train_sample_dl, 5, optimizer, loss_fn, device)", "Epoch: 1, BCELoss: 0.6925270915031433\nEpoch: 2, BCELoss: 0.6826320648193359\nEpoch: 3, BCELoss: 0.6738275647163391\nEpoch: 4, BCELoss: 0.6640056014060974\nEpoch: 5, BCELoss: 0.6518329381942749\n" ] ], [ [ "In order to construct a PyTorch model using SageMaker we must provide SageMaker with a training script. We may optionally include a directory which will be copied to the container and from which our training code will be run. When the training container is executed it will check the uploaded directory (if there is one) for a `requirements.txt` file and install any required Python libraries, after which the training script will be run.", "_____no_output_____" ], [ "### (TODO) Training the model\n\nWhen a PyTorch model is constructed in SageMaker, an entry point must be specified. This is the Python file which will be executed when the model is trained. Inside of the `train` directory is a file called `train.py` which has been provided and which contains most of the necessary code to train our model. The only thing that is missing is the implementation of the `train()` method which you wrote earlier in this notebook.\n\n**TODO**: Copy the `train()` method written above and paste it into the `train/train.py` file where required.\n\nThe way that SageMaker passes hyperparameters to the training script is by way of arguments. These arguments can then be parsed and used in the training script. To see how this is done take a look at the provided `train/train.py` file.", "_____no_output_____" ] ], [ [ "from sagemaker.pytorch import PyTorch\n\nestimator = PyTorch(entry_point=\"train.py\",\n source_dir=\"train\",\n role=role,\n framework_version='0.4.0',\n train_instance_count=1,\n train_instance_type='ml.p2.xlarge',\n hyperparameters={\n 'epochs': 10,\n 'hidden_dim': 200,\n })", "_____no_output_____" ], [ "estimator.fit({'training': input_data})", "'create_image_uri' will be deprecated in favor of 'ImageURIProvider' class in SageMaker Python SDK v2.\n's3_input' class will be renamed to 'TrainingInput' in SageMaker Python SDK v2.\n'create_image_uri' will be deprecated in favor of 'ImageURIProvider' class in SageMaker Python SDK v2.\n" ] ], [ [ "## Step 5: Testing the model\n\nAs mentioned at the top of this notebook, we will be testing this model by first deploying it and then sending the testing data to the deployed endpoint. We will do this so that we can make sure that the deployed model is working correctly.\n\n## Step 6: Deploy the model for testing\n\nNow that we have trained our model, we would like to test it to see how it performs. Currently our model takes input of the form `review_length, review[500]` where `review[500]` is a sequence of `500` integers which describe the words present in the review, encoded using `word_dict`. Fortunately for us, SageMaker provides built-in inference code for models with simple inputs such as this.\n\nThere is one thing that we need to provide, however, and that is a function which loads the saved model. This function must be called `model_fn()` and takes as its only parameter a path to the directory where the model artifacts are stored. This function must also be present in the python file which we specified as the entry point. In our case the model loading function has been provided and so no changes need to be made.\n\n**NOTE**: When the built-in inference code is run it must import the `model_fn()` method from the `train.py` file. This is why the training code is wrapped in a main guard ( ie, `if __name__ == '__main__':` )\n\nSince we don't need to change anything in the code that was uploaded during training, we can simply deploy the current model as-is.\n\n**NOTE:** When deploying a model you are asking SageMaker to launch an compute instance that will wait for data to be sent to it. As a result, this compute instance will continue to run until *you* shut it down. This is important to know since the cost of a deployed endpoint depends on how long it has been running for.\n\nIn other words **If you are no longer using a deployed endpoint, shut it down!**\n\n**TODO:** Deploy the trained model.", "_____no_output_____" ] ], [ [ "# TODO: Deploy the trained model\npredictor = estimator.deploy(initial_instance_count=1, instance_type='ml.p2.xlarge')", "Parameter image will be renamed to image_uri in SageMaker Python SDK v2.\n'create_image_uri' will be deprecated in favor of 'ImageURIProvider' class in SageMaker Python SDK v2.\n" ], [ "predictor", "_____no_output_____" ] ], [ [ "## Step 7 - Use the model for testing\n\nOnce deployed, we can read in the test data and send it off to our deployed model to get some results. Once we collect all of the results we can determine how accurate our model is.", "_____no_output_____" ] ], [ [ "test_X = pd.concat([pd.DataFrame(test_X_len), pd.DataFrame(test_X)], axis=1)", "_____no_output_____" ], [ "# We split the data into chunks and send each chunk seperately, accumulating the results.\n\ndef predict(data, rows=512):\n split_array = np.array_split(data, int(data.shape[0] / float(rows) + 1))\n predictions = np.array([])\n for array in split_array:\n predictions = np.append(predictions, predictor.predict(array))\n \n return predictions", "_____no_output_____" ], [ "predictions = predict(test_X.values)\npredictions = [round(num) for num in predictions]", "_____no_output_____" ], [ "from sklearn.metrics import accuracy_score\naccuracy_score(test_y, predictions)", "_____no_output_____" ] ], [ [ "**Question:** How does this model compare to the XGBoost model you created earlier? Why might these two models perform differently on this dataset? Which do *you* think is better for sentiment analysis?", "_____no_output_____" ], [ "**Answer:** I don't remember the accuracy of XGBoost but as I think they perform similarly, however, RNN model is usually used in sentiment analysis problems.", "_____no_output_____" ], [ "### (TODO) More testing\n\nWe now have a trained model which has been deployed and which we can send processed reviews to and which returns the predicted sentiment. However, ultimately we would like to be able to send our model an unprocessed review. That is, we would like to send the review itself as a string. For example, suppose we wish to send the following review to our model.", "_____no_output_____" ] ], [ [ "test_review = 'The simplest pleasures in life are the best, and this film is one of them. Combining a rather basic storyline of love and adventure this movie transcends the usual weekend fair with wit and unmitigated charm.'", "_____no_output_____" ] ], [ [ "The question we now need to answer is, how do we send this review to our model?\n\nRecall in the first section of this notebook we did a bunch of data processing to the IMDb dataset. In particular, we did two specific things to the provided reviews.\n - Removed any html tags and stemmed the input\n - Encoded the review as a sequence of integers using `word_dict`\n \nIn order process the review we will need to repeat these two steps.\n\n**TODO**: Using the `review_to_words` and `convert_and_pad` methods from section one, convert `test_review` into a numpy array `test_data` suitable to send to our model. Remember that our model expects input of the form `review_length, review[500]`. So make sure you produce two variables from processing: \n- A sequence of length 500 which represents the converted review\n- The length of the review", "_____no_output_____" ] ], [ [ "# TODO: Convert test_review into a form usable by the model and save the results in test_data\ntest_data = review_to_words(test_review)\ntest_data = [np.array(convert_and_pad(word_dict, test_data)[0])]", "_____no_output_____" ] ], [ [ "Now that we have processed the review, we can send the resulting array to our model to predict the sentiment of the review.", "_____no_output_____" ] ], [ [ "predictor.predict(test_data)", "_____no_output_____" ] ], [ [ "Since the return value of our model is close to `1`, we can be certain that the review we submitted is positive.", "_____no_output_____" ], [ "### Delete the endpoint\n\nOf course, just like in the XGBoost notebook, once we've deployed an endpoint it continues to run until we tell it to shut down. Since we are done using our endpoint for now, we can delete it.", "_____no_output_____" ] ], [ [ "estimator.delete_endpoint()", "estimator.delete_endpoint() will be deprecated in SageMaker Python SDK v2. Please use the delete_endpoint() function on your predictor instead.\n" ] ], [ [ "## Step 6 (again) - Deploy the model for the web app\n\nNow that we know that our model is working, it's time to create some custom inference code so that we can send the model a review which has not been processed and have it determine the sentiment of the review.\n\nAs we saw above, by default the estimator which we created, when deployed, will use the entry script and directory which we provided when creating the model. However, since we now wish to accept a string as input and our model expects a processed review, we need to write some custom inference code.\n\nWe will store the code that we write in the `serve` directory. Provided in this directory is the `model.py` file that we used to construct our model, a `utils.py` file which contains the `review_to_words` and `convert_and_pad` pre-processing functions which we used during the initial data processing, and `predict.py`, the file which will contain our custom inference code. Note also that `requirements.txt` is present which will tell SageMaker what Python libraries are required by our custom inference code.\n\nWhen deploying a PyTorch model in SageMaker, you are expected to provide four functions which the SageMaker inference container will use.\n - `model_fn`: This function is the same function that we used in the training script and it tells SageMaker how to load our model.\n - `input_fn`: This function receives the raw serialized input that has been sent to the model's endpoint and its job is to de-serialize and make the input available for the inference code.\n - `output_fn`: This function takes the output of the inference code and its job is to serialize this output and return it to the caller of the model's endpoint.\n - `predict_fn`: The heart of the inference script, this is where the actual prediction is done and is the function which you will need to complete.\n\nFor the simple website that we are constructing during this project, the `input_fn` and `output_fn` methods are relatively straightforward. We only require being able to accept a string as input and we expect to return a single value as output. You might imagine though that in a more complex application the input or output may be image data or some other binary data which would require some effort to serialize.\n\n### (TODO) Writing inference code\n\nBefore writing our custom inference code, we will begin by taking a look at the code which has been provided.", "_____no_output_____" ] ], [ [ "!pygmentize serve/predict.py", "\u001b[34mimport\u001b[39;49;00m \u001b[04m\u001b[36margparse\u001b[39;49;00m\r\n\u001b[34mimport\u001b[39;49;00m \u001b[04m\u001b[36mjson\u001b[39;49;00m\r\n\u001b[34mimport\u001b[39;49;00m \u001b[04m\u001b[36mos\u001b[39;49;00m\r\n\u001b[34mimport\u001b[39;49;00m \u001b[04m\u001b[36mpickle\u001b[39;49;00m\r\n\u001b[34mimport\u001b[39;49;00m \u001b[04m\u001b[36msys\u001b[39;49;00m\r\n\u001b[34mimport\u001b[39;49;00m \u001b[04m\u001b[36msagemaker_containers\u001b[39;49;00m\r\n\u001b[34mimport\u001b[39;49;00m \u001b[04m\u001b[36mpandas\u001b[39;49;00m \u001b[34mas\u001b[39;49;00m \u001b[04m\u001b[36mpd\u001b[39;49;00m\r\n\u001b[34mimport\u001b[39;49;00m \u001b[04m\u001b[36mnumpy\u001b[39;49;00m \u001b[34mas\u001b[39;49;00m \u001b[04m\u001b[36mnp\u001b[39;49;00m\r\n\u001b[34mimport\u001b[39;49;00m \u001b[04m\u001b[36mtorch\u001b[39;49;00m\r\n\u001b[34mimport\u001b[39;49;00m \u001b[04m\u001b[36mtorch\u001b[39;49;00m\u001b[04m\u001b[36m.\u001b[39;49;00m\u001b[04m\u001b[36mnn\u001b[39;49;00m \u001b[34mas\u001b[39;49;00m \u001b[04m\u001b[36mnn\u001b[39;49;00m\r\n\u001b[34mimport\u001b[39;49;00m \u001b[04m\u001b[36mtorch\u001b[39;49;00m\u001b[04m\u001b[36m.\u001b[39;49;00m\u001b[04m\u001b[36moptim\u001b[39;49;00m \u001b[34mas\u001b[39;49;00m \u001b[04m\u001b[36moptim\u001b[39;49;00m\r\n\u001b[34mimport\u001b[39;49;00m \u001b[04m\u001b[36mtorch\u001b[39;49;00m\u001b[04m\u001b[36m.\u001b[39;49;00m\u001b[04m\u001b[36mutils\u001b[39;49;00m\u001b[04m\u001b[36m.\u001b[39;49;00m\u001b[04m\u001b[36mdata\u001b[39;49;00m\r\n\r\n\u001b[34mfrom\u001b[39;49;00m \u001b[04m\u001b[36mmodel\u001b[39;49;00m \u001b[34mimport\u001b[39;49;00m LSTMClassifier\r\n\r\n\u001b[34mfrom\u001b[39;49;00m \u001b[04m\u001b[36mutils\u001b[39;49;00m \u001b[34mimport\u001b[39;49;00m review_to_words, convert_and_pad\r\n\r\n\u001b[34mdef\u001b[39;49;00m \u001b[32mmodel_fn\u001b[39;49;00m(model_dir):\r\n \u001b[33m\"\"\"Load the PyTorch model from the `model_dir` directory.\"\"\"\u001b[39;49;00m\r\n \u001b[36mprint\u001b[39;49;00m(\u001b[33m\"\u001b[39;49;00m\u001b[33mLoading model.\u001b[39;49;00m\u001b[33m\"\u001b[39;49;00m)\r\n\r\n \u001b[37m# First, load the parameters used to create the model.\u001b[39;49;00m\r\n model_info = {}\r\n model_info_path = os.path.join(model_dir, \u001b[33m'\u001b[39;49;00m\u001b[33mmodel_info.pth\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m)\r\n \u001b[34mwith\u001b[39;49;00m \u001b[36mopen\u001b[39;49;00m(model_info_path, \u001b[33m'\u001b[39;49;00m\u001b[33mrb\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m) \u001b[34mas\u001b[39;49;00m f:\r\n model_info = torch.load(f)\r\n\r\n \u001b[36mprint\u001b[39;49;00m(\u001b[33m\"\u001b[39;49;00m\u001b[33mmodel_info: \u001b[39;49;00m\u001b[33m{}\u001b[39;49;00m\u001b[33m\"\u001b[39;49;00m.format(model_info))\r\n\r\n \u001b[37m# Determine the device and construct the model.\u001b[39;49;00m\r\n device = torch.device(\u001b[33m\"\u001b[39;49;00m\u001b[33mcuda\u001b[39;49;00m\u001b[33m\"\u001b[39;49;00m \u001b[34mif\u001b[39;49;00m torch.cuda.is_available() \u001b[34melse\u001b[39;49;00m \u001b[33m\"\u001b[39;49;00m\u001b[33mcpu\u001b[39;49;00m\u001b[33m\"\u001b[39;49;00m)\r\n model = LSTMClassifier(model_info[\u001b[33m'\u001b[39;49;00m\u001b[33membedding_dim\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m], model_info[\u001b[33m'\u001b[39;49;00m\u001b[33mhidden_dim\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m], model_info[\u001b[33m'\u001b[39;49;00m\u001b[33mvocab_size\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m])\r\n\r\n \u001b[37m# Load the store model parameters.\u001b[39;49;00m\r\n model_path = os.path.join(model_dir, \u001b[33m'\u001b[39;49;00m\u001b[33mmodel.pth\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m)\r\n \u001b[34mwith\u001b[39;49;00m \u001b[36mopen\u001b[39;49;00m(model_path, \u001b[33m'\u001b[39;49;00m\u001b[33mrb\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m) \u001b[34mas\u001b[39;49;00m f:\r\n model.load_state_dict(torch.load(f))\r\n\r\n \u001b[37m# Load the saved word_dict.\u001b[39;49;00m\r\n word_dict_path = os.path.join(model_dir, \u001b[33m'\u001b[39;49;00m\u001b[33mword_dict.pkl\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m)\r\n \u001b[34mwith\u001b[39;49;00m \u001b[36mopen\u001b[39;49;00m(word_dict_path, \u001b[33m'\u001b[39;49;00m\u001b[33mrb\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m) \u001b[34mas\u001b[39;49;00m f:\r\n model.word_dict = pickle.load(f)\r\n\r\n model.to(device).eval()\r\n\r\n \u001b[36mprint\u001b[39;49;00m(\u001b[33m\"\u001b[39;49;00m\u001b[33mDone loading model.\u001b[39;49;00m\u001b[33m\"\u001b[39;49;00m)\r\n \u001b[34mreturn\u001b[39;49;00m model\r\n\r\n\u001b[34mdef\u001b[39;49;00m \u001b[32minput_fn\u001b[39;49;00m(serialized_input_data, content_type):\r\n \u001b[36mprint\u001b[39;49;00m(\u001b[33m'\u001b[39;49;00m\u001b[33mDeserializing the input data.\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m)\r\n \u001b[34mif\u001b[39;49;00m content_type == \u001b[33m'\u001b[39;49;00m\u001b[33mtext/plain\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m:\r\n data = serialized_input_data.decode(\u001b[33m'\u001b[39;49;00m\u001b[33mutf-8\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m)\r\n \u001b[34mreturn\u001b[39;49;00m data\r\n \u001b[34mraise\u001b[39;49;00m \u001b[36mException\u001b[39;49;00m(\u001b[33m'\u001b[39;49;00m\u001b[33mRequested unsupported ContentType in content_type: \u001b[39;49;00m\u001b[33m'\u001b[39;49;00m + content_type)\r\n\r\n\u001b[34mdef\u001b[39;49;00m \u001b[32moutput_fn\u001b[39;49;00m(prediction_output, accept):\r\n \u001b[36mprint\u001b[39;49;00m(\u001b[33m'\u001b[39;49;00m\u001b[33mSerializing the generated output.\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m)\r\n \u001b[34mreturn\u001b[39;49;00m \u001b[36mstr\u001b[39;49;00m(prediction_output)\r\n\r\n\u001b[34mdef\u001b[39;49;00m \u001b[32mpredict_fn\u001b[39;49;00m(input_data, model):\r\n \u001b[36mprint\u001b[39;49;00m(\u001b[33m'\u001b[39;49;00m\u001b[33mInferring sentiment of input data.\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m)\r\n\r\n device = torch.device(\u001b[33m\"\u001b[39;49;00m\u001b[33mcuda\u001b[39;49;00m\u001b[33m\"\u001b[39;49;00m \u001b[34mif\u001b[39;49;00m torch.cuda.is_available() \u001b[34melse\u001b[39;49;00m \u001b[33m\"\u001b[39;49;00m\u001b[33mcpu\u001b[39;49;00m\u001b[33m\"\u001b[39;49;00m)\r\n \r\n \u001b[34mif\u001b[39;49;00m model.word_dict \u001b[35mis\u001b[39;49;00m \u001b[34mNone\u001b[39;49;00m:\r\n \u001b[34mraise\u001b[39;49;00m \u001b[36mException\u001b[39;49;00m(\u001b[33m'\u001b[39;49;00m\u001b[33mModel has not been loaded properly, no word_dict.\u001b[39;49;00m\u001b[33m'\u001b[39;49;00m)\r\n \r\n \u001b[37m# TODO: Process input_data so that it is ready to be sent to our model.\u001b[39;49;00m\r\n \u001b[37m# You should produce two variables:\u001b[39;49;00m\r\n \u001b[37m# data_X - A sequence of length 500 which represents the converted review\u001b[39;49;00m\r\n \u001b[37m# data_len - The length of the review\u001b[39;49;00m\r\n\r\n data_X = \u001b[34mNone\u001b[39;49;00m\r\n data_len = \u001b[34mNone\u001b[39;49;00m\r\n\r\n \u001b[37m# Using data_X and data_len we construct an appropriate input tensor. Remember\u001b[39;49;00m\r\n \u001b[37m# that our model expects input data of the form 'len, review[500]'.\u001b[39;49;00m\r\n data_pack = np.hstack((data_len, data_X))\r\n data_pack = data_pack.reshape(\u001b[34m1\u001b[39;49;00m, -\u001b[34m1\u001b[39;49;00m)\r\n \r\n data = torch.from_numpy(data_pack)\r\n data = data.to(device)\r\n\r\n \u001b[37m# Make sure to put the model into evaluation mode\u001b[39;49;00m\r\n model.eval()\r\n\r\n \u001b[37m# TODO: Compute the result of applying the model to the input data. The variable `result` should\u001b[39;49;00m\r\n \u001b[37m# be a numpy array which contains a single integer which is either 1 or 0\u001b[39;49;00m\r\n\r\n result = \u001b[34mNone\u001b[39;49;00m\r\n\r\n \u001b[34mreturn\u001b[39;49;00m result\r\n" ] ], [ [ "As mentioned earlier, the `model_fn` method is the same as the one provided in the training code and the `input_fn` and `output_fn` methods are very simple and your task will be to complete the `predict_fn` method. \n\n**Note**: Our model expects input data of the form 'len, review[500]'. So make sure you produce two variables from processing: \n- `data_X`: A sequence of length 500 which represents the converted review\n- `data_len`: - The length of the review\n\nMake sure that you save the completed file as `predict.py` in the `serve` directory.\n\n**TODO**: Complete the `predict_fn()` method in the `serve/predict.py` file.", "_____no_output_____" ], [ "### Deploying the model\n\nNow that the custom inference code has been written, we will create and deploy our model. To begin with, we need to construct a new PyTorchModel object which points to the model artifacts created during training and also points to the inference code that we wish to use. Then we can call the deploy method to launch the deployment container.\n\n**NOTE**: The default behaviour for a deployed PyTorch model is to assume that any input passed to the predictor is a `numpy` array. In our case we want to send a string so we need to construct a simple wrapper around the `RealTimePredictor` class to accomodate simple strings. In a more complicated situation you may want to provide a serialization object, for example if you wanted to sent image data.", "_____no_output_____" ] ], [ [ "from sagemaker.predictor import RealTimePredictor\nfrom sagemaker.pytorch import PyTorchModel\n\nclass StringPredictor(RealTimePredictor):\n def __init__(self, endpoint_name, sagemaker_session):\n super(StringPredictor, self).__init__(endpoint_name, sagemaker_session, content_type='text/plain')\n\nmodel = PyTorchModel(model_data=estimator.model_data,\n role = role,\n framework_version='0.4.0',\n entry_point='predict.py',\n source_dir='serve',\n predictor_cls=StringPredictor)\npredictor = model.deploy(initial_instance_count=1, instance_type='ml.m4.xlarge')", "Parameter image will be renamed to image_uri in SageMaker Python SDK v2.\n'create_image_uri' will be deprecated in favor of 'ImageURIProvider' class in SageMaker Python SDK v2.\n" ] ], [ [ "### Testing the model\n\nNow that we have deployed our model with the custom inference code, we should test to see if everything is working. Here we test our model by loading the first `250` positive and negative reviews and send them to the endpoint, then collect the results. The reason for only sending some of the data is that the amount of time it takes for our model to process the input and then perform inference is quite long and so testing the entire data set would be prohibitive.", "_____no_output_____" ] ], [ [ "import glob\n\ndef test_reviews(data_dir='../data/aclImdb', stop=250):\n \n results = []\n ground = []\n \n # We make sure to test both positive and negative reviews \n for sentiment in ['pos', 'neg']:\n \n path = os.path.join(data_dir, 'test', sentiment, '*.txt')\n files = glob.glob(path)\n \n files_read = 0\n \n print('Starting ', sentiment, ' files')\n \n # Iterate through the files and send them to the predictor\n for f in files:\n with open(f) as review:\n # First, we store the ground truth (was the review positive or negative)\n if sentiment == 'pos':\n ground.append(1)\n else:\n ground.append(0)\n # Read in the review and convert to 'utf-8' for transmission via HTTP\n review_input = review.read().encode('utf-8')\n # Send the review to the predictor and store the results\n results.append(float(predictor.predict(review_input)))\n \n # Sending reviews to our endpoint one at a time takes a while so we\n # only send a small number of reviews\n files_read += 1\n if files_read == stop:\n break\n \n return ground, results", "_____no_output_____" ], [ "ground, results = test_reviews()", "Starting pos files\nStarting neg files\n" ], [ "from sklearn.metrics import accuracy_score\naccuracy_score(ground, results)", "_____no_output_____" ] ], [ [ "As an additional test, we can try sending the `test_review` that we looked at earlier.", "_____no_output_____" ] ], [ [ "predictor.predict(test_review)", "_____no_output_____" ] ], [ [ "Now that we know our endpoint is working as expected, we can set up the web page that will interact with it. If you don't have time to finish the project now, make sure to skip down to the end of this notebook and shut down your endpoint. You can deploy it again when you come back.", "_____no_output_____" ], [ "## Step 7 (again): Use the model for the web app\n\n> **TODO:** This entire section and the next contain tasks for you to complete, mostly using the AWS console.\n\nSo far we have been accessing our model endpoint by constructing a predictor object which uses the endpoint and then just using the predictor object to perform inference. What if we wanted to create a web app which accessed our model? The way things are set up currently makes that not possible since in order to access a SageMaker endpoint the app would first have to authenticate with AWS using an IAM role which included access to SageMaker endpoints. However, there is an easier way! We just need to use some additional AWS services.\n\n<img src=\"Web App Diagram.svg\">\n\nThe diagram above gives an overview of how the various services will work together. On the far right is the model which we trained above and which is deployed using SageMaker. On the far left is our web app that collects a user's movie review, sends it off and expects a positive or negative sentiment in return.\n\nIn the middle is where some of the magic happens. We will construct a Lambda function, which you can think of as a straightforward Python function that can be executed whenever a specified event occurs. We will give this function permission to send and recieve data from a SageMaker endpoint.\n\nLastly, the method we will use to execute the Lambda function is a new endpoint that we will create using API Gateway. This endpoint will be a url that listens for data to be sent to it. Once it gets some data it will pass that data on to the Lambda function and then return whatever the Lambda function returns. Essentially it will act as an interface that lets our web app communicate with the Lambda function.\n\n### Setting up a Lambda function\n\nThe first thing we are going to do is set up a Lambda function. This Lambda function will be executed whenever our public API has data sent to it. When it is executed it will receive the data, perform any sort of processing that is required, send the data (the review) to the SageMaker endpoint we've created and then return the result.\n\n#### Part A: Create an IAM Role for the Lambda function\n\nSince we want the Lambda function to call a SageMaker endpoint, we need to make sure that it has permission to do so. To do this, we will construct a role that we can later give the Lambda function.\n\nUsing the AWS Console, navigate to the **IAM** page and click on **Roles**. Then, click on **Create role**. Make sure that the **AWS service** is the type of trusted entity selected and choose **Lambda** as the service that will use this role, then click **Next: Permissions**.\n\nIn the search box type `sagemaker` and select the check box next to the **AmazonSageMakerFullAccess** policy. Then, click on **Next: Review**.\n\nLastly, give this role a name. Make sure you use a name that you will remember later on, for example `LambdaSageMakerRole`. Then, click on **Create role**.\n\n#### Part B: Create a Lambda function\n\nNow it is time to actually create the Lambda function.\n\nUsing the AWS Console, navigate to the AWS Lambda page and click on **Create a function**. When you get to the next page, make sure that **Author from scratch** is selected. Now, name your Lambda function, using a name that you will remember later on, for example `sentiment_analysis_func`. Make sure that the **Python 3.6** runtime is selected and then choose the role that you created in the previous part. Then, click on **Create Function**.\n\nOn the next page you will see some information about the Lambda function you've just created. If you scroll down you should see an editor in which you can write the code that will be executed when your Lambda function is triggered. In our example, we will use the code below. \n\n```python\n# We need to use the low-level library to interact with SageMaker since the SageMaker API\n# is not available natively through Lambda.\nimport boto3\n\ndef lambda_handler(event, context):\n\n # The SageMaker runtime is what allows us to invoke the endpoint that we've created.\n runtime = boto3.Session().client('sagemaker-runtime')\n\n # Now we use the SageMaker runtime to invoke our endpoint, sending the review we were given\n response = runtime.invoke_endpoint(EndpointName = '**ENDPOINT NAME HERE**', # The name of the endpoint we created\n ContentType = 'text/plain', # The data format that is expected\n Body = event['body']) # The actual review\n\n # The response is an HTTP response whose body contains the result of our inference\n result = response['Body'].read().decode('utf-8')\n\n return {\n 'statusCode' : 200,\n 'headers' : { 'Content-Type' : 'text/plain', 'Access-Control-Allow-Origin' : '*' },\n 'body' : result\n }\n```\n\nOnce you have copy and pasted the code above into the Lambda code editor, replace the `**ENDPOINT NAME HERE**` portion with the name of the endpoint that we deployed earlier. You can determine the name of the endpoint using the code cell below.", "_____no_output_____" ] ], [ [ "predictor.endpoint", "_____no_output_____" ] ], [ [ "Once you have added the endpoint name to the Lambda function, click on **Save**. Your Lambda function is now up and running. Next we need to create a way for our web app to execute the Lambda function.\n\n### Setting up API Gateway\n\nNow that our Lambda function is set up, it is time to create a new API using API Gateway that will trigger the Lambda function we have just created.\n\nUsing AWS Console, navigate to **Amazon API Gateway** and then click on **Get started**.\n\nOn the next page, make sure that **New API** is selected and give the new api a name, for example, `sentiment_analysis_api`. Then, click on **Create API**.\n\nNow we have created an API, however it doesn't currently do anything. What we want it to do is to trigger the Lambda function that we created earlier.\n\nSelect the **Actions** dropdown menu and click **Create Method**. A new blank method will be created, select its dropdown menu and select **POST**, then click on the check mark beside it.\n\nFor the integration point, make sure that **Lambda Function** is selected and click on the **Use Lambda Proxy integration**. This option makes sure that the data that is sent to the API is then sent directly to the Lambda function with no processing. It also means that the return value must be a proper response object as it will also not be processed by API Gateway.\n\nType the name of the Lambda function you created earlier into the **Lambda Function** text entry box and then click on **Save**. Click on **OK** in the pop-up box that then appears, giving permission to API Gateway to invoke the Lambda function you created.\n\nThe last step in creating the API Gateway is to select the **Actions** dropdown and click on **Deploy API**. You will need to create a new Deployment stage and name it anything you like, for example `prod`.\n\nYou have now successfully set up a public API to access your SageMaker model. Make sure to copy or write down the URL provided to invoke your newly created public API as this will be needed in the next step. This URL can be found at the top of the page, highlighted in blue next to the text **Invoke URL**.", "_____no_output_____" ], [ "## Step 4: Deploying our web app\n\nNow that we have a publicly available API, we can start using it in a web app. For our purposes, we have provided a simple static html file which can make use of the public api you created earlier.\n\nIn the `website` folder there should be a file called `index.html`. Download the file to your computer and open that file up in a text editor of your choice. There should be a line which contains **\\*\\*REPLACE WITH PUBLIC API URL\\*\\***. Replace this string with the url that you wrote down in the last step and then save the file.\n\nNow, if you open `index.html` on your local computer, your browser will behave as a local web server and you can use the provided site to interact with your SageMaker model.\n\nIf you'd like to go further, you can host this html file anywhere you'd like, for example using github or hosting a static site on Amazon's S3. Once you have done this you can share the link with anyone you'd like and have them play with it too!\n\n> **Important Note** In order for the web app to communicate with the SageMaker endpoint, the endpoint has to actually be deployed and running. This means that you are paying for it. Make sure that the endpoint is running when you want to use the web app but that you shut it down when you don't need it, otherwise you will end up with a surprisingly large AWS bill.\n\n**TODO:** Make sure that you include the edited `index.html` file in your project submission.", "_____no_output_____" ], [ "Now that your web app is working, trying playing around with it and see how well it works.\n\n**Question**: Post a screenshot showing a sample review that you entered into your web app and the predicted sentiment. What was the predicted sentiment of your example review?", "_____no_output_____" ], [ "**Screenshot:**\n\n![Screen%20Shot%202021-11-21%20at%2010.57.30%20PM.png](attachment:Screen%20Shot%202021-11-21%20at%2010.57.30%20PM.png)\n\n**Answer:** the predicted sentiment is that the review is positive.", "_____no_output_____" ], [ "### Delete the endpoint\n\nRemember to always shut down your endpoint if you are no longer using it. You are charged for the length of time that the endpoint is running so if you forget and leave it on you could end up with an unexpectedly large bill.", "_____no_output_____" ] ], [ [ "predictor.delete_endpoint()", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code", "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ] ]

ec7da2b2a149d3b4b65178ef4bd85cc48d72e14f

13,970

ipynb

Jupyter Notebook

snobal_1d/Irwin_pipeline.ipynb

jomey/isnobal-local

38df9b7127a3ac4ff460340c673a8f2f9ef33218

[ "MIT" ]

3

2021-12-02T01:55:08.000Z

2022-01-05T21:40:55.000Z

snobal_1d/Irwin_pipeline.ipynb

jomey/isnobal-local

38df9b7127a3ac4ff460340c673a8f2f9ef33218

[ "MIT" ]

12

2020-09-24T22:32:48.000Z

2022-02-17T20:29:54.000Z

snobal_1d/Irwin_pipeline.ipynb

jomey/isnobal-local

38df9b7127a3ac4ff460340c673a8f2f9ef33218

[ "MIT" ]

1

2020-08-26T19:45:32.000Z

22.387821

149

0.519041

[ [ [ "import MesoPy\nimport json\nimport pandas as pd\nimport numpy as np\nimport missingno as msno\nimport os\nimport matplotlib.pyplot as plt\nfrom dateutil import parser", "_____no_output_____" ], [ "%load_ext autoreload\n%autoreload 1", "_____no_output_____" ], [ "%aimport station_pipeline", "_____no_output_____" ], [ "df = pd.read_csv('raw_data/Irwin_WY20.csv',\n skiprows=[0,1,2,3,4,5,7],\n na_values=['NAN','NA','N/A','NaN'])\n\ndf.info()", "_____no_output_____" ], [ "# change tz abbreviation to utc offset.\n# abbreviations are NOT unique (across contries) and therfore unable to be inferred by Pandas.\ndf['Date_Time'] = df['Date_Time'].str.replace('MDT', 'UTC-6')\ndf['Date_Time'] = df['Date_Time'].str.replace('MST', 'UTC-7')", "_____no_output_____" ], [ "# This subset df has a timestamp that includes DST changes. All others are UTC-7\n# set utc=True which incorporates dst changes, than set to utc-6, then revert to tz-unaware.\n# best practice is to use tz-aware dfs, but some functions may cause problems, so there is a trade-off\ndfir = df.iloc[:,0:10]\ndfir.index = pd.to_datetime(dfir['Date_Time'], utc=True)\ndfir.index = dfir.index.tz_convert('America/Denver')\n# removes the timezone information resulting in naive local time (now matching other dfs)\ndfir = dfir.tz_localize(None)\n# subset to start on WY\ndfir = dfir['2019-10-01':]\ndfir.info()", "_____no_output_____" ], [ "dfp = df.iloc[:,11:14]\ndfp.index = pd.to_datetime(dfp['Unnamed: 11'] + ' ' + dfp['Unnamed: 12']); dfp.info()", "_____no_output_____" ], [ "dfr = df.iloc[:,15:20]\ndfr.index = pd.to_datetime(dfr['Unnamed: 15']); dfr.info()", "_____no_output_____" ] ], [ [ "## Take a look at the data in each df, see what is missing, where we need to interpolate values, etc. ", "_____no_output_____" ] ], [ [ "dfir = dfir.dropna(how='all')\ndfir = dfir[~dfir.index.duplicated()]\n# recast to monotonic time series\n# note that snobal (IPW) expects serially complete data\ndfir = dfir.asfreq('1H')\nmsno.matrix(dfir, freq='M')", "_____no_output_____" ], [ "dfp = dfp.dropna(how='all')\ndfp.drop_duplicates(inplace=True)\ndfp = dfp.asfreq('1H')\nmsno.matrix(dfp, freq='M')", "_____no_output_____" ], [ "dfr = dfr.dropna(how='all').copy()\ndfr.drop_duplicates(inplace=True)\ndfr = dfr[['Incoming) Solar_Wm2_1_Avg','Outgoing_Solar_Wm2_3_Avg']]\n# coerce numeric, possibly strings from Excel file...\ndfr = dfr.apply(pd.to_numeric)\ndfr = dfr.asfreq('1H')\nmsno.matrix(dfr, freq='M')", "_____no_output_____" ] ], [ [ "# Processing", "_____no_output_____" ] ], [ [ "dfp.info()", "_____no_output_____" ], [ "dfir.info()", "_____no_output_____" ], [ "dfr.info()", "_____no_output_____" ], [ "# interpolate small gaps \n# note that none-floats are excluded\ndfir = dfir.interpolate(method='time', axis='index')", "_____no_output_____" ], [ "msno.matrix(dfir, freq='M')", "_____no_output_____" ], [ "dfir['net_solar'] = dfr['Incoming) Solar_Wm2_1_Avg'].subtract(dfr['Outgoing_Solar_Wm2_3_Avg'])", "_____no_output_____" ], [ "dfr[['Incoming) Solar_Wm2_1_Avg','Outgoing_Solar_Wm2_3_Avg']].plot(figsize=(20,15))", "_____no_output_____" ], [ "dfir = station_pipeline.snow_density_fraction(df=dfir, \n air_t_col='air_temp_set_1')", "_____no_output_____" ], [ "# subset to last radiation data\ndfir = dfir[:'20200630']\ndfp = dfp[:'20200630']", "_____no_output_____" ], [ "msno.matrix(dfir, freq='M')", "_____no_output_____" ], [ "dfir['air_t_K'] = np.nan\ndfir['air_t_K'] = dfir['air_temp_set_1'].apply(lambda x: x + 273.15)", "_____no_output_____" ], [ "lw_list = [station_pipeline.longwave_est_2(x, y, z) for x, y, z in zip(dfir['relative_humidity_set_1'],\n dfir['net_solar'],\n dfir['air_t_K'])]\n\nlwdf = pd.DataFrame(lw_list)\nlwdf = lwdf.set_index(dfir.index)\ndfir['lw_in_est'] = lwdf[0].copy()", "_____no_output_____" ], [ "msno.matrix(dfir, freq='M')", "_____no_output_____" ], [ "dfir.info()", "_____no_output_____" ], [ "dfir[['air_t_K','lw_in_est', 'relative_humidity_set_1']].plot(figsize=(20,10))", "_____no_output_____" ], [ "dfir = station_pipeline.vapor_pressure(df=dfir, dt='dew_point_temperature_set_1d')", "_____no_output_____" ], [ "dfir['soil_temp'] = 0", "_____no_output_____" ] ], [ [ "## Precipitation", "_____no_output_____" ] ], [ [ "dfp['precip_accum_mm'] = dfp['Precpip_Accum'] * 25.4\n#dfp['precip_accum_mm'] = dfp['precip_accum_mm'].mask(dfp['precip_accum_mm'] < 0, 0)\ndfp['precip_hourly_mm'] = dfp['precip_accum_mm'].diff(1)", "_____no_output_____" ], [ "# diff func misses first ts, so need to set as 0\ndfp['precip_hourly_mm'].loc['2019-10-01 00:00:00'] = 0\n\ndfp['precip_hourly_mm']", "_____no_output_____" ], [ "#dfp['precip_accum_mm'].loc['2019-10'].plot(figsize=(20,10))\ndfp['precip_accum_mm'].plot(figsize=(20,10))", "_____no_output_____" ], [ "dfp['precip_hourly_mm'] = dfp['precip_hourly_mm'].mask(dfp['precip_hourly_mm'] < 0, 0)\ndfp['precip_hourly_mm'].plot(figsize=(20,10))", "_____no_output_____" ], [ "dfp['fraction'] = dfir['fraction']\ndfp['density'] = dfir['density']\ndfp['p_temp'] = dfir['air_temp_set_1']", "_____no_output_____" ], [ "dfp['iter'] = range(len(dfp))", "_____no_output_____" ], [ "# 800 w/m^2 is snobal upper bound (IPW)\ndfir['net_solar'] = dfir['net_solar'].mask(dfir['net_solar'] > 800, 800)\ndfir['net_solar'] = dfir['net_solar'].mask(dfir['net_solar'] < 0, 0)", "_____no_output_____" ], [ "# set ano lower bound to 0.15 to prevent possible snobal error\ndfir['wind_speed_set_1'] = dfir['wind_speed_set_1'].mask(dfir['wind_speed_set_1'] < 0.15, 0)", "_____no_output_____" ], [ "dfp_in = dfp[['iter','precip_hourly_mm','fraction','density','p_temp']].copy()", "_____no_output_____" ], [ "dfp_in.interpolate(method='time', inplace=True)\n#dfp_in = dfp_in.interpolate(method='time', axis='index')", "_____no_output_____" ], [ "dfdat = dfir[['net_solar','lw_in_est','air_temp_set_1','vp','wind_speed_set_1','soil_temp']].copy()", "_____no_output_____" ], [ "dfdat = dfdat.interpolate(method='time', axis='index')", "_____no_output_____" ], [ "dfdat.plot(subplots=True, figsize=(20,15))", "_____no_output_____" ], [ "msno.matrix(dfdat, freq='M')", "_____no_output_____" ], [ "dfdat = dfdat.round(3)\n\ndfdat.to_csv('ipw_inputs/snobal.data.input', \n index=False,\n header=False,\n sep=' ',)", "_____no_output_____" ], [ "dfp_in.plot(subplots=True, figsize=(20,15))", "_____no_output_____" ], [ "msno.matrix(dfp_in, freq='M')", "_____no_output_____" ], [ "dfp_in.info()", "_____no_output_____" ], [ "dfdat.info()", "_____no_output_____" ], [ "#optional round floats\n#dfp_in = dfp_in.round(3)\n\ndfp_in.to_csv('ipw_inputs/snobal.ppt.input', \n index=False,\n header=False,\n sep=' ',)", "_____no_output_____" ] ], [ [ "## Write other input files", "_____no_output_____" ] ], [ [ "with open('ipw_inputs/snow.properties.input', 'w') as f:\n f.write('0 0 0 0 0 0')", "_____no_output_____" ], [ "with open('ipw_inputs/inheight.input', 'w') as f:\n f.write('0 3 3 0.001 0.15')", "_____no_output_____" ] ], [ [ "## IPW CLI", "_____no_output_____" ] ], [ [ "snobal -z 3109 -t 60 -m 0.01 -s snow.properties.input -h inheight.input -p snobal.ppt.input -i snobal.data.input -O normal -o irwin_model.v1 -c", "_____no_output_____" ] ], [ [ "dfdat['iter'] = range(len(dfdat))", "_____no_output_____" ], [ "station_pipeline.html_chart(dfdat)", "_____no_output_____" ], [ "station_pipeline.html_chart(dfp_in)", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "raw", "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "raw" ], [ "code", "code", "code" ] ]

ec7da5a1057ab31c00bb8d35b434051b158e8e49

53,438

ipynb

Jupyter Notebook

lunar_lander/solution/Deep_Q_Network_Solution.ipynb

filipgrigorov/deep-reinforcement-learning

9b70e01775551228155c940f382ec6bce509f08b

[ "MIT" ]

null

lunar_lander/solution/Deep_Q_Network_Solution.ipynb

filipgrigorov/deep-reinforcement-learning

9b70e01775551228155c940f382ec6bce509f08b

[ "MIT" ]

null

lunar_lander/solution/Deep_Q_Network_Solution.ipynb

filipgrigorov/deep-reinforcement-learning

9b70e01775551228155c940f382ec6bce509f08b

[ "MIT" ]

null

175.782895

31,116

0.891594

[ [ [ "# Deep Q-Network (DQN)\n---\nIn this notebook, you will implement a DQN agent with OpenAI Gym's LunarLander-v2 environment.\n\n### 1. Import the Necessary Packages", "_____no_output_____" ] ], [ [ "import gym\n!pip3 install box2d\nimport random\nimport torch\nimport numpy as np\nfrom collections import deque\nimport matplotlib.pyplot as plt\n%matplotlib inline\n\n!python -m pip install pyvirtualdisplay\nfrom pyvirtualdisplay import Display\ndisplay = Display(visible=0, size=(1400, 900))\ndisplay.start()\n\nis_ipython = 'inline' in plt.get_backend()\nif is_ipython:\n from IPython import display\n\nplt.ion()", "Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\nRequirement already satisfied: box2d in /home/filip/anaconda3/envs/rl/lib/python3.7/site-packages (2.3.10)\nLooking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com\nRequirement already satisfied: pyvirtualdisplay in /home/filip/anaconda3/envs/rl/lib/python3.7/site-packages (1.3.2)\nRequirement already satisfied: EasyProcess in /home/filip/anaconda3/envs/rl/lib/python3.7/site-packages (from pyvirtualdisplay) (0.3)\n" ] ], [ [ "### 2. Instantiate the Environment and Agent\n\nInitialize the environment in the code cell below.", "_____no_output_____" ] ], [ [ "env = gym.make('LunarLander-v2')\nenv.seed(0)\nprint('State shape: ', env.observation_space.shape)\nprint('Number of actions: ', env.action_space.n)", "State shape: (8,)\nNumber of actions: 4\n" ] ], [ [ "Please refer to the instructions in `Deep_Q_Network.ipynb` if you would like to write your own DQN agent. Otherwise, run the code cell below to load the solution files.", "_____no_output_____" ] ], [ [ "from dqn_agent import Agent\n\nagent = Agent(state_size=8, action_size=4, seed=0)\n\n# watch an untrained agent\nstate = env.reset()\nimg = plt.imshow(env.render(mode='rgb_array'))\nfor j in range(200):\n action = agent.act(state)\n img.set_data(env.render(mode='rgb_array')) \n plt.axis('off')\n display.display(plt.gcf())\n display.clear_output(wait=True)\n state, reward, done, _ = env.step(action)\n if done:\n break \n \nenv.close()", "_____no_output_____" ] ], [ [ "### 3. Train the Agent with DQN\n\nRun the code cell below to train the agent from scratch. You are welcome to amend the supplied values of the parameters in the function, to try to see if you can get better performance!\n\nAlternatively, you can skip to the next step below (**4. Watch a Smart Agent!**), to load the saved model weights from a pre-trained agent.", "_____no_output_____" ] ], [ [ "def dqn(n_episodes=2000, max_t=1000, eps_start=1.0, eps_end=0.01, eps_decay=0.995):\n \"\"\"Deep Q-Learning.\n \n Params\n ======\n n_episodes (int): maximum number of training episodes\n max_t (int): maximum number of timesteps per episode\n eps_start (float): starting value of epsilon, for epsilon-greedy action selection\n eps_end (float): minimum value of epsilon\n eps_decay (float): multiplicative factor (per episode) for decreasing epsilon\n \"\"\"\n scores = [] # list containing scores from each episode\n scores_window = deque(maxlen=100) # last 100 scores\n eps = eps_start # initialize epsilon\n for i_episode in range(1, n_episodes+1):\n state = env.reset()\n score = 0\n for t in range(max_t):\n action = agent.act(state, eps)\n next_state, reward, done, _ = env.step(action)\n agent.step(state, action, reward, next_state, done)\n state = next_state\n score += reward\n if done:\n break \n scores_window.append(score) # save most recent score\n scores.append(score) # save most recent score\n eps = max(eps_end, eps_decay*eps) # decrease epsilon\n print('\\rEpisode {}\\tAverage Score: {:.2f}'.format(i_episode, np.mean(scores_window)), end=\"\")\n if i_episode % 100 == 0:\n print('\\rEpisode {}\\tAverage Score: {:.2f}'.format(i_episode, np.mean(scores_window)))\n if np.mean(scores_window)>=200.0:\n print('\\nEnvironment solved in {:d} episodes!\\tAverage Score: {:.2f}'.format(i_episode-100, np.mean(scores_window)))\n torch.save(agent.qnetwork_local.state_dict(), 'checkpoint.pth')\n break\n return scores\n\nscores = dqn()\n\n# plot the scores\nfig = plt.figure()\nax = fig.add_subplot(111)\nplt.plot(np.arange(len(scores)), scores)\nplt.ylabel('Score')\nplt.xlabel('Episode #')\nplt.show()", "Episode 100\tAverage Score: -173.85\nEpisode 200\tAverage Score: -113.61\nEpisode 300\tAverage Score: -66.439\nEpisode 400\tAverage Score: -1.891\nEpisode 500\tAverage Score: 51.72\nEpisode 600\tAverage Score: 110.87\nEpisode 700\tAverage Score: 196.81\nEpisode 706\tAverage Score: 200.74\nEnvironment solved in 606 episodes!\tAverage Score: 200.74\n" ] ], [ [ "### 4. Watch a Smart Agent!\n\nIn the next code cell, you will load the trained weights from file to watch a smart agent!", "_____no_output_____" ] ], [ [ "# load the weights from file\nagent.qnetwork_local.load_state_dict(torch.load('checkpoint.pth'))\n\nfor i in range(3):\n state = env.reset()\n img = plt.imshow(env.render(mode='rgb_array'))\n for j in range(200):\n action = agent.act(state)\n img.set_data(env.render(mode='rgb_array')) \n plt.axis('off')\n display.display(plt.gcf())\n display.clear_output(wait=True)\n state, reward, done, _ = env.step(action)\n if done:\n break \n \nenv.close()", "_____no_output_____" ] ], [ [ "### 5. Explore\n\nIn this exercise, you have implemented a DQN agent and demonstrated how to use it to solve an OpenAI Gym environment. To continue your learning, you are encouraged to complete any (or all!) of the following tasks:\n- Amend the various hyperparameters and network architecture to see if you can get your agent to solve the environment faster. Once you build intuition for the hyperparameters that work well with this environment, try solving a different OpenAI Gym task with discrete actions!\n- You may like to implement some improvements such as prioritized experience replay, Double DQN, or Dueling DQN! \n- Write a blog post explaining the intuition behind the DQN algorithm and demonstrating how to use it to solve an RL environment of your choosing. ", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ] ]

ec7db8df25460ca2b35cb2421abdd12ddad86102

213,200

ipynb

Jupyter Notebook

ipython/Convex_Sample_Graph_Queries.ipynb

convex-community/graph_queries

dad4abad4bbd16c1c129f9c015bbfdea17563c02

[ "MIT" ]

2

2021-09-30T12:30:42.000Z

2022-01-14T16:52:59.000Z

ipython/Convex_Sample_Graph_Queries.ipynb

convex-community/graph_queries

dad4abad4bbd16c1c129f9c015bbfdea17563c02

[ "MIT" ]

null

ipython/Convex_Sample_Graph_Queries.ipynb

convex-community/graph_queries

dad4abad4bbd16c1c129f9c015bbfdea17563c02

[ "MIT" ]

null

106.228201

35,148

0.783443

[ [ [ "import requests\nimport pandas as pd\nimport json\nfrom matplotlib import pyplot as plt\nplt.style.use('ggplot')", "_____no_output_____" ], [ "GRT_QUERY_ENDPOINT = \"https://api.thegraph.com/subgraphs/name/convex-community/curve-pools\"", "_____no_output_____" ], [ "def grt_query(query):\n r = requests.post(GRT_QUERY_ENDPOINT, json={'query': query})\n return r.json()", "_____no_output_____" ], [ "pd.set_option('display.float_format', lambda x: '%.5f' % x)", "_____no_output_____" ] ], [ [ "## General pools info", "_____no_output_____" ] ], [ [ "query = \"\"\"\n{\n pools(\n orderBy:creationDate\n orderDirection: asc) {\n name\n id\n lpToken\n lpTokenBalance\n gauge\n crvRewardsPool\n swap\n stash\n assetType\n apr\n tvl\n creationDate\n }\n}\n\"\"\"", "_____no_output_____" ], [ "data = grt_query(query)", "_____no_output_____" ], [ "df = pd.DataFrame.from_dict(data['data']['pools'])\ndf['creationDate'] = pd.to_datetime(df['creationDate'],unit='s')\ndf['tvl'] = df['tvl'].astype(float)\ndf['apr'] = df['apr'].astype(float)\ncols = df.columns.tolist()\nfirst_columns = ['id', 'name', 'creationDate', 'apr', 'tvl']\ncols = first_columns + [_ for _ in cols if _ not in first_columns]\ndf = df.reindex(columns=cols)\ndf", "_____no_output_____" ] ], [ [ "## Historical TVL (USD)", "_____no_output_____" ] ], [ [ "query = \"\"\"\n{\n dailyPoolSnapshots(\n first: 200\n where: {poolName: \"link\"}\n orderBy:timestamp\n orderDirection: asc) \n {\n tvl\n timestamp\n }\n}\n\n\"\"\"", "_____no_output_____" ], [ "data = grt_query(query)", "_____no_output_____" ], [ "df = pd.DataFrame.from_dict(data['data']['dailyPoolSnapshots'])\ndf['timestamp'] = pd.to_datetime(df['timestamp'],unit='s')\ndf['tvl'] = df['tvl'].astype(float)\ndf", "_____no_output_____" ], [ "df.plot(x='timestamp', y='tvl', figsize=(10,6))", "_____no_output_____" ] ], [ [ "## Historical TVL (lpTokens)", "_____no_output_____" ] ], [ [ "query = \"\"\"\n{\n dailyPoolSnapshots(\n first: 200\n where: {poolName: \"seth\"}\n orderBy:timestamp\n orderDirection: asc) \n {\n lpTokenBalance\n timestamp\n }\n}\n\n\"\"\"", "_____no_output_____" ], [ "data = grt_query(query)\ndf = pd.DataFrame.from_dict(data['data']['dailyPoolSnapshots'])\ndf['timestamp'] = pd.to_datetime(df['timestamp'],unit='s')\ndf['lpTokenBalance'] = df['lpTokenBalance'].astype(float)\ndf", "_____no_output_____" ], [ "df.plot(x='timestamp', y='lpTokenBalance', figsize=(10,6))", "_____no_output_____" ] ], [ [ "## Historical APR", "_____no_output_____" ] ], [ [ "query = \"\"\"\n{\n dailyPoolSnapshots(\n first: 200\n where: {poolName: \"seth\"}\n orderBy:timestamp\n orderDirection: asc) \n {\n apr\n timestamp\n }\n}\n\n\"\"\"", "_____no_output_____" ], [ "data = grt_query(query)", "_____no_output_____" ], [ "df = pd.DataFrame.from_dict(data['data']['dailyPoolSnapshots'])\ndf['timestamp'] = pd.to_datetime(df['timestamp'],unit='s')\ndf['apr'] = df['apr'].astype(float) * 100\ndf['apr'] = df['apr'].apply(lambda x: 0 if x > 300 else x) # bug w/ several pool. apr calculus issue ?\ndf", "_____no_output_____" ], [ "df.plot(x='timestamp', y='apr', figsize=(10,6))", "_____no_output_____" ] ], [ [ "## Deposits vs Withdrawals", "_____no_output_____" ] ], [ [ "query = \"\"\"\n{\n dailyPoolSnapshots(\n where: {poolid: 3}\n orderBy:timestamp\n orderDirection: asc) \n {\n withdrawalValue\n depositValue\n timestamp\n }\n}\n\n\"\"\"", "_____no_output_____" ], [ "data = grt_query(query)", "_____no_output_____" ], [ "df = pd.DataFrame.from_dict(data['data']['dailyPoolSnapshots'])\ndf['timestamp'] = pd.to_datetime(df['timestamp'],unit='s').dt.strftime('%Y-%m-%d')\ndf['depositValue'] = df['depositValue'].astype(float)\ndf['withdrawalValue'] = df['withdrawalValue'].astype(float)\ndf", "_____no_output_____" ], [ "ax = df.plot.bar(x='timestamp', figsize=(20,6), logy=True)\nax.set_ylim(ymin=1)", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code" ] ]

ec7dc12f9de2afd8a87e5f776521d6920a9944f8

285,418

ipynb

Jupyter Notebook

GPS-RO_data_difference.ipynb

johanmeh/master

ea9b4d3b67e3fc806b6e4c824dfc79562c721e2e

[ "BSD-2-Clause" ]

null

GPS-RO_data_difference.ipynb

johanmeh/master

ea9b4d3b67e3fc806b6e4c824dfc79562c721e2e

[ "BSD-2-Clause" ]

null

GPS-RO_data_difference.ipynb

johanmeh/master

ea9b4d3b67e3fc806b6e4c824dfc79562c721e2e

[ "BSD-2-Clause" ]

null

1,740.353659

187,632

0.964866

[ [ [ "import xarray as xr\nimport numpy as np\nimport matplotlib.pyplot as plt\nimport pandas as pd", "_____no_output_____" ], [ "ds1 = xr.open_dataset('GPS-RO__CP_LR_5x5_2007-2018.nc')\nds2 = xr.open_dataset('GPS-RO__LR-CP__gridded_ALL_MISSIONS_2002-2018.nc', decode_times=False)", "_____no_output_____" ], [ "ds2['time'] = pd.date_range('2002-01-01', '2018-12-31', freq='M')\nds1_t = ds1.resample(time='M').mean()", "_____no_output_____" ], [ "ds1_av = ds1_t.mean(axis=(1,2), keep_attrs=True)\nds1_av['time'] = ds1_t.time\n", "_____no_output_____" ], [ "ds2_av = ds2.mean(axis=(1,2))\nds2_av['time'] = ds2.time", "_____no_output_____" ], [ "fig, axs = plt.subplots(2,2,figsize=(10,8))\n\nds1_av.CP_T.plot(ax=axs[0,0])\nds1_av.CP_z.plot(ax=axs[1,0])\n\nds2_av.CP_T.plot(ax=axs[0,0])\nds2_av.CP_z.plot(ax=axs[1,0])\n\n\nds1_av.LR_T.plot(ax=axs[0,1])\nds1_av.LR_z.plot(ax=axs[1,1])\n\nds2_av.LR_T.plot(ax=axs[0,1])\nds2_av.LR_z.plot(ax=axs[1,1])\n\nplt.tight_layout()", "_____no_output_____" ], [ "t_slice = slice('2007-01-01', '2018-12-31')\nd1 = ds1_av.CP_T - ds2_av.CP_T.sel(time=t_slice) \nd2 = ds1_av.CP_z - ds2_av.CP_z.sel(time=t_slice)\nd3 = ds1_av.LR_T - ds2_av.LR_T.sel(time=t_slice)\nd4 = ds1_av.LR_z - ds2_av.LR_z.sel(time=t_slice)", "_____no_output_____" ], [ "fig, axs = plt.subplots(2,2,figsize=(8,6))\n\nd1.plot(ax=axs[0,0])\nd2.plot(ax=axs[1,0])\nd3.plot(ax=axs[0,1])\nd4.plot(ax=axs[1,1])\n\nplt.tight_layout()", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7dce4a2c7b54f1e9ffc87f1f27ec74c2d3f036

21,055

ipynb

Jupyter Notebook

notebooks/landryda/model-multiplexer.ipynb

crim-ca/crims2s

0392fe320b819cf71b22522ea1d6b6e3cddf5142

[ "MIT" ]

7

2021-11-06T03:42:04.000Z

2022-03-22T00:48:24.000Z

notebooks/landryda/model-multiplexer.ipynb

crim-ca/crims2s

0392fe320b819cf71b22522ea1d6b6e3cddf5142

[ "MIT" ]

1

2021-12-03T18:54:12.000Z

notebooks/landryda/model-multiplexer.ipynb

crim-ca/crims2s

0392fe320b819cf71b22522ea1d6b6e3cddf5142

[ "MIT" ]

5

2021-11-06T02:08:19.000Z

2022-03-31T02:48:37.000Z

29.447552

170

0.531275

[ [ [ "%load_ext autoreload\n%autoreload 2", "_____no_output_____" ] ], [ [ "# Model multiplexer\n\nRework the linear model and wrap it into a model multiplexer, so that we can have one model per month-day.", "_____no_output_____" ] ], [ [ "import matplotlib.pyplot as plt\nimport numpy as np\nimport pandas as pd\nimport re\nimport seaborn as sns\nimport torch\nimport torch.distributions\nimport torch.nn as nn\nimport xarray as xr\n\nfrom crims2s.dataset import S2SDataset, TransformedDataset\nfrom crims2s.transform import CompositeTransform, AddBiweeklyDimTransform, AddMetadata, ExampleToPytorch\nfrom crims2s.util import ECMWF_FORECASTS", "_____no_output_____" ], [ "DATASET = '***BASEDIR***/mlready/2021-08-08-test/'", "_____no_output_____" ] ], [ [ "## make transform to interface dataset w/ linear model", "_____no_output_____" ] ], [ [ "def std_estimator(dataset, dim=None):\n dataset_mean = dataset.mean(dim=dim)\n \n if dim is None:\n dim_sizes = [dataset.sizes[x] for x in dataset_mean.dims]\n elif isinstance(dim, str):\n dim_sizes = dataset.sizes[dim]\n else:\n dim_sizes = [dataset.sizes[x] for x in dim]\n \n n = np.prod(dim_sizes)\n \n return xr.ufuncs.sqrt(xr.ufuncs.square(dataset - dataset_mean).sum(dim=dim) / (n - 1))", "_____no_output_____" ], [ "def model_to_distribution(model): \n model_tp_mean = model.tp.isel(lead_time=-1).mean(dim='realization').rename('tp_mu')\n model_tp_std = std_estimator(model.tp.isel(lead_time=-1), dim='realization').rename('tp_sigma')\n \n model_t2m_mean = model.t2m.mean(dim=['lead_time', 'realization']).rename('t2m_mu')\n model_t2m_std = std_estimator(model.t2m, dim=['lead_time', 'realization']).rename('t2m_sigma')\n \n return xr.merge([\n model_tp_mean, model_tp_std, model_t2m_mean, model_t2m_std\n ]).drop('lead_time').rename(biweekly_forecast='lead_time')", "_____no_output_____" ], [ "def obs_to_biweekly(obs):\n aggregate_obs_tp = obs.pr.sum(dim='lead_time', min_count=2).rename('tp')\n aggregate_obs_t2m = obs.t2m.mean(dim='lead_time')\n return xr.merge([aggregate_obs_tp, aggregate_obs_t2m])", "_____no_output_____" ], [ "def linear_model_adapter(example):\n model = model_to_distribution(example['model'])\n obs = obs_to_biweekly(example['obs'])\n \n return {\n 'model': model,\n 'obs': obs\n }", "_____no_output_____" ], [ "#raw_train_dataset = S2SDataset(DATASET, filter_str='0312.nc', include_features=False, years=list(range(2000,2017)))\n#raw_val_dataset = S2SDataset(DATASET, filter_str='0312.nc', include_features=False, years=list(range(2017,2020)))", "_____no_output_____" ], [ "filter_re = re.compile('01[0-9]{2}.nc$')", "_____no_output_____" ], [ "raw_train_dataset = S2SDataset(DATASET, include_features=False, name_filter=lambda x: filter_re.search(x), years=list(range(2000,2017)))\nraw_val_dataset = S2SDataset(DATASET, include_features=False, name_filter=lambda x: filter_re.search(x), years=list(range(2017,2020)))", "_____no_output_____" ], [ "for k in raw_train_dataset[0]['model'].data_vars:\n print(k)", "_____no_output_____" ], [ "raw_train_dataset[0]['model']['t2m']", "_____no_output_____" ], [ "transform = CompositeTransform([AddBiweeklyDimTransform(), linear_model_adapter, AddMetadata(), ExampleToPytorch()])\n\ntrain_dataset = TransformedDataset(raw_train_dataset, transform)\nval_dataset = TransformedDataset(raw_val_dataset, transform)", "_____no_output_____" ], [ "len(train_dataset)", "_____no_output_____" ], [ "train_dataloader = torch.utils.data.DataLoader(train_dataset, num_workers=4, batch_size=None, batch_sampler=None)\nval_dataloader = torch.utils.data.DataLoader(val_dataset, num_workers=4, batch_size=None, batch_sampler=None)", "_____no_output_____" ], [ "from typing import Union, Callable, Any, Hashable", "_____no_output_____" ], [ "train_dataset[0]", "_____no_output_____" ], [ "one_batch = next(iter(train_dataloader))", "_____no_output_____" ], [ "class ModelMultiplexer(nn.Module):\n \"\"\"Dispatch the training examples to multiple models depending on the example.\n For instance, we could use this to use a different model for every monthday forecast.\n \n Because it uses an arbitraty model for every sample, this module does not support batching.\n To use it, it is recommended to disable automatic batching on the dataloader.\"\"\"\n def __init__(self, key, models):\n \"\"\"Args:\n key: If a str, used as a key to fetch the model name from the example dict. \n If a callable, called on the example and should return to model name to use.\n models: A mapping from model names to model instances. They keys should correspond to what is returned when applying key on the example.\"\"\"\n super().__init__()\n \n if isinstance(key, str):\n self.key_fn = lambda x: x[key]\n else:\n self.key_fn = key\n \n self.models = nn.ModuleDict(models)\n \n def forward(self, example): \n model_name = self.key_fn(example)\n model = self.models[model_name]\n \n return model(example)\n ", "_____no_output_____" ], [ "class LinearModel(nn.Module):\n def __init__(self, *shape, fill_weights=0.0, fill_intercept=0.0):\n super().__init__()\n \n self.weights = nn.Parameter(torch.full(shape, fill_weights))\n self.intercept = nn.Parameter(torch.full(shape, fill_intercept))\n \n def forward(self, x):\n return self.intercept + self.weights * x + x", "_____no_output_____" ], [ "class TempPrecipEMOS(nn.Module):\n def __init__(self, biweekly=False):\n super().__init__()\n \n shape = (3, 121, 240) if biweekly else (121, 240)\n \n self.tp_mu_model = LinearModel(*shape)\n self.tp_sigma_model = LinearModel(*shape, fill_intercept=1.0)\n \n self.t2m_mu_model = LinearModel(*shape)\n self.t2m_sigma_model = LinearModel(*shape, fill_intercept=1.0)\n \n def forward(self, example):\n forecast_tp_mu, forecast_tp_sigma = example['model_tp_mu'], example['model_tp_sigma']\n forecast_t2m_mu, forecast_t2m_sigma = example['model_t2m_mu'], example['model_t2m_sigma']\n \n tp_mu = self.tp_mu_model(forecast_tp_mu)\n tp_sigma = self.tp_sigma_model(forecast_tp_sigma)\n tp_sigma = torch.clip(tp_sigma, min=1e-6)\n\n t2m_mu = self.t2m_mu_model(forecast_t2m_mu)\n t2m_sigma = self.t2m_sigma_model(forecast_t2m_sigma)\n t2m_sigma = torch.clip(t2m_sigma, min=1e-6)\n \n tp_dist = torch.distributions.Normal(loc=tp_mu, scale=tp_sigma)\n t2m_dist = torch.distributions.Normal(loc=t2m_mu, scale=t2m_sigma)\n \n return t2m_dist, tp_dist", "_____no_output_____" ], [ "monthdays = [f'{m:02}{d:02}' for m, d in ECMWF_FORECASTS]\nweekly_models = {monthday: TempPrecipEMOS(biweekly=True) for monthday in monthdays}", "_____no_output_____" ], [ "weekly_models", "_____no_output_____" ], [ "len(weekly_models)", "_____no_output_____" ], [ "monthdays[:5]", "_____no_output_____" ], [ "model = ModelMultiplexer('monthday', weekly_models)", "_____no_output_____" ], [ "optimizer = torch.optim.Adam(params=model.parameters(), lr=1e-3)", "_____no_output_____" ], [ "for epoch in range(20):\n train_losses = []\n train_temperature_losses = []\n train_rain_losses = []\n \n model.train()\n for example in train_dataloader:\n t2m_dist, tp_dist = model.forward(example)\n\n tp_obs = example['obs_tp']\n tp_nan_mask = tp_obs.isnan()\n tp_obs[tp_nan_mask] = 0.0\n tp_log_likelihood = tp_dist.log_prob(tp_obs)\n tp_log_likelihood[tp_nan_mask] = 0.0\n\n t2m_obs = example['obs_t2m']\n t2m_nan_mask = t2m_obs.isnan()\n t2m_obs[t2m_nan_mask] = 0.0\n t2m_log_likelihood = t2m_dist.log_prob(t2m_obs)\n t2m_log_likelihood[t2m_nan_mask] = 0.0\n\n rain_loss = -tp_log_likelihood.mean()\n temperature_loss = -t2m_log_likelihood.mean()\n loss = rain_loss + temperature_loss\n \n loss.backward()\n \n optimizer.step()\n optimizer.zero_grad()\n \n train_losses.append(float(loss.detach()))\n train_temperature_losses.append(float(temperature_loss.detach()))\n train_rain_losses.append(float(rain_loss.detach()))\n\n train_mean_loss = np.array(train_losses).mean()\n train_mean_rain_loss = np.array(train_rain_losses).mean()\n train_mean_temperature_loss = np.array(train_temperature_losses).mean()\n print(f'Epoch {epoch} train loss: {train_mean_loss}. Temperature: {train_mean_temperature_loss}. Rain: {train_mean_rain_loss}.')\n \n \n model.eval()\n with torch.no_grad():\n val_losses = []\n val_rain_losses = []\n val_t2m_losses = []\n for example in val_dataloader:\n t2m_dist, tp_dist = model(example)\n \n obs_t2m, obs_tp = example['obs_t2m'], example['obs_tp']\n \n tp_obs = example['obs_tp']\n tp_nan_mask = tp_obs.isnan()\n tp_obs[tp_nan_mask] = 0.0\n tp_log_likelihood = tp_dist.log_prob(tp_obs)\n tp_log_likelihood[tp_nan_mask] = 0.0\n\n t2m_obs = example['obs_t2m']\n t2m_nan_mask = t2m_obs.isnan()\n t2m_obs[t2m_nan_mask] = 0.0\n t2m_log_likelihood = t2m_dist.log_prob(t2m_obs)\n t2m_log_likelihood[t2m_nan_mask] = 0.0\n \n val_rain_loss = -tp_log_likelihood.mean()\n val_temperature_loss = -t2m_log_likelihood.mean()\n val_loss = val_rain_loss + val_temperature_loss\n \n val_rain_losses.append(val_rain_loss.detach())\n val_t2m_losses.append(val_temperature_loss.detach())\n val_losses.append(val_loss.detach())\n \n \n val_mean_loss = np.array(val_losses).mean()\n val_mean_rain_loss = np.array(val_rain_losses).mean()\n val_mean_temperature_loss = np.array(val_t2m_losses).mean()\n print(f'Epoch {epoch} val loss: {val_mean_loss}. Temperature: {val_mean_temperature_loss}. Rain: {val_mean_rain_loss}.')\n print()", "_____no_output_____" ], [ "sns.lineplot(data=train_losses)", "_____no_output_____" ], [ "val_rain_losses", "_____no_output_____" ], [ "for k in model.models:\n print(k, model.models[k].tp_mu_model.weights[2][~t2m_nan_mask[0]].max())", "_____no_output_____" ], [ "dicts = []\n\nmasks = {\n 't2m': t2m_nan_mask[0],\n 'tp': tp_nan_mask[0],\n}\n\nfor monthday in monthdays[:4]:\n m = model.models[monthday]\n \n for parameter in ['mu', 'sigma']:\n for variable in ['t2m', 'tp']:\n attr_name = f'{variable}_{parameter}_model'\n linear_model = getattr(m, attr_name)\n mask = masks[variable]\n \n for component in ['weights', 'intercept']:\n model_component = getattr(linear_model, component)\n \n for lead_time in [0, 1, 2]:\n at_lead_time = model_component[lead_time]\n masked = at_lead_time[~mask].detach().numpy()\n \n datapoints = [{\n 'monthday': monthday, \n 'parameter': parameter,\n 'variable': variable,\n 'component': component,\n 'lead_time': lead_time,\n 'value': x\n } for x in masked]\n dicts.extend(datapoints)", "_____no_output_____" ], [ "monthdays[:4]", "_____no_output_____" ], [ "sns.histplot(data=masked)", "_____no_output_____" ], [ "df = pd.DataFrame(dicts)", "_____no_output_____" ], [ "df", "_____no_output_____" ], [ "sns.set_theme()", "_____no_output_____" ], [ "for variable in ['t2m', 'tp']:\n for parameter in ['mu', 'sigma']:\n for component in ['weights', 'intercept']:\n title = f'Distribution of {variable} {parameter} {component}'\n \n fig = plt.gcf()\n fig.set_size_inches(6,4)\n ax = sns.kdeplot(data=df[(df['component'] == component) & (df['variable'] == variable) & (df['parameter'] == parameter)], x='value', hue='lead_time')\n ax.set_title(title)\n plt.savefig(title.replace(' ', '_') + '.png')\n plt.show()\n ", "_____no_output_____" ], [ "sns.displot(data=df[(df['component'] == 'weights') & (df['variable'] == 't2m') & (df['parameter'] == 'mu')], x='value', row='lead_time').set_title('')", "_____no_output_____" ], [ "sns.histplot(data=df[(df['parameter'] == 'mu') & (df['variable'] == 't2m') & (df['lead_time'] == 0) & (df['component'] == 'intercept')], x='value',hue='monthday')", "_____no_output_____" ], [ "sns.histplot(data=model.models['0123'].tp_mu_model.intercept[2][~tp_nan_mask[0]].detach().numpy())", "_____no_output_____" ], [ "sns.histplot(data=model.models['0123'].tp_mu_model.weights[1][~tp_nan_mask[0]].detach().numpy())", "_____no_output_____" ], [ "sns.histplot(data=model.models['0123'].tp_sigma_model.intercept[1][~tp_nan_mask[0]].detach().numpy())", "_____no_output_____" ], [ "sns.histplot(data=model.models['0123'].tp_sigma_model.weights[1][~tp_nan_mask[0]].detach().numpy())", "_____no_output_____" ], [ "sns.histplot(data=model.models['0123'].t2m_sigma_model.weights[1][~t2m_nan_mask[0]].detach().numpy())", "_____no_output_____" ], [ "sns.histplot(data=model.models['0123'].t2m_sigma_model.intercept[1][~t2m_nan_mask[0]].detach().numpy())", "_____no_output_____" ], [ "sns.histplot(data=model.models['0123'].t2m_mu_model.intercept[1][~t2m_nan_mask[0]].detach().numpy())", "_____no_output_____" ], [ "sns.histplot(data=model.models['0123'].t2m_mu_model.weights[1][~t2m_nan_mask[0]].detach().numpy())", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code" ]

[ [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7de2df579726f5e2a8ebcbdfeb54aa2b1e1744

3,073

ipynb

Jupyter Notebook

02-Algorithms/04-Memoization.ipynb

acanozturk/algds

2a37e82e30eee5e7074519f665c20a9969e1f325

[ "Apache-2.0" ]

null

02-Algorithms/04-Memoization.ipynb

acanozturk/algds

2a37e82e30eee5e7074519f665c20a9969e1f325

[ "Apache-2.0" ]

null

02-Algorithms/04-Memoization.ipynb

acanozturk/algds

2a37e82e30eee5e7074519f665c20a9969e1f325

[ "Apache-2.0" ]

null

17.763006

44

0.40384

[ [ [ "def fact_memo(fact):\n \n memory = {}\n \n def check_memory(n):\n \n if n not in memory:\n \n memory[n] = fact(n)\n \n return memory[n]\n \n return check_memory\n", "_____no_output_____" ], [ "@fact_memo\ndef fact(n):\n \n if n == 1:\n \n return 1\n \n else:\n \n return n * fact(n - 1)", "_____no_output_____" ], [ "fact(20)", "1\n1\n1\n1\n" ], [ "def fibo_memo(fibo):\n \n memory = {}\n \n def check_memory(n):\n \n if n not in memory:\n \n memory[n] = fibo(n)\n \n return memory[n]\n \n return check_memory\n", "_____no_output_____" ], [ "@fibo_memo\ndef fibo(n):\n \n if n < 2:\n \n return n\n \n return fibo(n - 1) + fibo(n - 2)\n", "_____no_output_____" ], [ "fibo(65)", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code", "code", "code", "code", "code" ] ]

ec7e0de1f5aa238787d97b11c7c8b5654e4bec26

56,310

ipynb

Jupyter Notebook

Yelp-politness-ChiSquareTest.ipynb

gammingfreak/Dissertation

794b218036c9884e41faaac271ec5bc73c3725a3

[ "MIT" ]

null

Yelp-politness-ChiSquareTest.ipynb

gammingfreak/Dissertation

794b218036c9884e41faaac271ec5bc73c3725a3

[ "MIT" ]

null

Yelp-politness-ChiSquareTest.ipynb

gammingfreak/Dissertation

794b218036c9884e41faaac271ec5bc73c3725a3

[ "MIT" ]

null

29.872679

145

0.446759

[ [ [ "from textblob import TextBlob, Word, Blobber\nfrom textblob.classifiers import NaiveBayesClassifier\nfrom textblob.taggers import NLTKTagger\nfrom nltk.stem.porter import PorterStemmer", "_____no_output_____" ], [ "import pandas as pd\nimport textstat\nimport numpy as np\nfrom sklearn.metrics import precision_recall_fscore_support\nfrom sklearn.metrics import classification_report\nfrom sklearn import tree\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import accuracy_score\nfrom sklearn.ensemble import RandomForestClassifier\nfrom sklearn import svm\nfrom sklearn.metrics import roc_auc_score\nfrom sklearn.svm import SVC\nfrom sklearn.linear_model import LinearRegression\nfrom sklearn.metrics import mean_squared_error, r2_score\nfrom sklearn.ensemble import RandomForestRegressor\nfrom sklearn import preprocessing\nfrom sklearn.feature_extraction.text import CountVectorizer\nfrom sklearn.feature_extraction.text import TfidfVectorizer\nfrom nltk.stem.porter import PorterStemmer\nfrom sklearn.linear_model import Ridge\nimport datetime as dt\nfrom sklearn.metrics import r2_score\nimport seaborn as sns\nfrom scipy import stats\nfrom spellchecker import SpellChecker", "_____no_output_____" ], [ "from collections import OrderedDict\nimport matplotlib.pyplot as plt\n%matplotlib inline", "_____no_output_____" ], [ "path_review=\"D:/Trinity_DS/Dissertations/201907/politness/politness_predicted/yelp_polite.csv\"", "_____no_output_____" ], [ "df_review = pd.read_csv(path_review,low_memory=False,encoding='iso-8859-1')", "_____no_output_____" ], [ "df_review=df_review.drop(df_review.columns[0], axis=1)\ndf_review.head()", "_____no_output_____" ], [ "analysis_df = pd.DataFrame()\ndf_review['date'] = pd.to_datetime(df_review['date'])", "_____no_output_____" ], [ "analysis_df['day_of_week'] = df_review['date'].dt.weekday_name\nanalysis_df['stars'] = df_review['stars']", "_____no_output_____" ], [ "analysis_df.stars.unique()", "_____no_output_____" ], [ "np_matix= np.array(analysis_df)\nnp_matix\ndays_rating=pd.crosstab(np_matix[:,0],np_matix[:,1],\n rownames=['Days'], colnames=['Ratings'],)\nchi2 , p ,dof ,expected = stats.chi2_contingency(days_rating)\ndays_rating", "_____no_output_____" ], [ "days_rating=pd.crosstab(np_matix[:,0],np_matix[:,1],\n rownames=['Days'], colnames=['Ratings'],)\nchi2 , p ,dof ,expected = stats.chi2_contingency(days_rating)\nprint(\"Reviews and DaysOfWeek chi_2 value---------------\",chi2)\nprint(\"Reviews and DaysOfWeek p value-------------------\",p)\nprint(\"Reviews and DaysOfWeek degreeoffreedom value-----\",dof)", "Reviews and DaysOfWeek chi_2 value--------------- 161.90234724872974\nReviews and DaysOfWeek p value------------------- 1.9730334914915967e-22\nReviews and DaysOfWeek degreeoffreedom value----- 24\n" ], [ "prob=0.95 \ncritical_day = stats.chi2.ppf(prob, dof)\nif abs(chi2) >= critical_day:\n print('Dependent (reject H0)')\n print('Reviews Dependent on Days of the Week')\nelse:\n print('Independent (fail to reject H0)')\nalpha = 1.0 - prob\nprint('significance=%.3f, p=%.3f' % (alpha, p))\nif p <= alpha:\n print('Dependent (reject H0)')\nelse:\n print('Independent (fail to reject H0)')", "Dependent (reject H0)\nReviews Dependent on Days of the Week\nsignificance=0.050, p=0.000\nDependent (reject H0)\n" ], [ "transformed_df=analysis_df\ntransformed_df['Weekend_WeekDays'] = np.where((transformed_df['day_of_week']=='Sunday') | (transformed_df['day_of_week']=='Saturday'),\n 'Weekend','Weekday')\ntransformed_df['high_low_Rating'] = np.where(transformed_df['stars']<=3,\n 'low','high')\ntransformed_df", "_____no_output_____" ], [ "transformedMatrix= np.array(transformed_df)\ncrWeekendWeekdays=pd.crosstab(transformedMatrix[:,3],transformedMatrix[:,2],\n rownames=['high_low_Rating'], colnames=['WeekendWeekdays'])\ncrWeekendWeekdays", "_____no_output_____" ], [ "\nchi2_v4 , p_v4 ,dof_v4 ,expected_v4 = stats.chi2_contingency(crWeekendWeekdays)\nprint(\"HighLowRating and WeekendWeekdays chi_2 value---------------\",chi2_v4)\nprint(\"HighLowRating and WeekendWeekdays p value-------------------\",p_v4)\nprint(\"HighLowRating and WeekendWeekdays degreeoffreedom value-----\",dof_v4)\nprob=0.95\ncritical_hour = stats.chi2.ppf(prob, dof_v4)\nif abs(chi2_v4) >= critical_hour:\n print('Dependent (reject H0) Hypothesis tested with probability of 95% and alpha of 0.5')\n print('High Low Ratings Dependent on Six hours bands')\nelse:\n print('Independent (fail to reject H0) Hypothesis tested with probability of 95% and alpha of 0.5')\nif p_v4 <= alpha:\n print('Dependent (reject H0) Hypothesis tested with probability of 95% and alpha of 0.5')\nelse:\n print('Independent (fail to reject H0) Hypothesis tested with probability of 95% and alpha of 0.5')\n", "HighLowRating and WeekendWeekdays chi_2 value--------------- 9.144751181089555\nHighLowRating and WeekendWeekdays p value------------------- 0.002494325563510335\nHighLowRating and WeekendWeekdays degreeoffreedom value----- 1\nDependent (reject H0) Hypothesis tested with probability of 95% and alpha of 0.5\nHigh Low Ratings Dependent on Six hours bands\nDependent (reject H0) Hypothesis tested with probability of 95% and alpha of 0.5\n" ], [ "list(analysis_df)", "_____no_output_____" ], [ "#textstat.flesch_reading_ease(.id)\nreadablity = []\nfor text in df_review['text']:\n readablity.append(textstat.flesch_reading_ease((text)))\nanalysis_df['flesch_reading_ease']=readablity", "_____no_output_____" ], [ "temp_df= df_review[['stars','useful']]\ndf_review.describe()", "_____no_output_____" ], [ "smog = []\nfor text in df_review['text']:\n smog.append(textstat.smog_index(text))", "_____no_output_____" ], [ "coleman_liau=[]\nfor text in df_review['text']:\n coleman_liau.append(textstat.coleman_liau_index(text))\n", "_____no_output_____" ], [ "sentence_count=[]\nfor text in df_review['text']:\n sentence_count.append(textstat.sentence_count(text))", "_____no_output_____" ], [ "gunning_fog=[]\nfor text in df_review['text']:\n gunning_fog.append(textstat.gunning_fog(text))", "_____no_output_____" ], [ "flesch_kincaid_grade=[]\nfor text in df_review['text']:\n flesch_kincaid_grade.append(textstat.flesch_kincaid_grade(text))", "_____no_output_____" ], [ "subjectivity_list=[]\npolarity_list=[]\nfor text in df_review['text']:\n subjectivity_list.append(TextBlob(text).sentiment.subjectivity)\n polarity_list.append(TextBlob(text).sentiment.subjectivity)", "_____no_output_____" ], [ "spell = SpellChecker()\nspelling_errors=[]\n\nfor text in df_review['text']:\n spelling_errors.append(len(spell.unknown(str(text).split(' '))))", "_____no_output_____" ], [ "analysis_df['smog']=smog\nanalysis_df['coleman_liau']=coleman_liau\nanalysis_df['sentence_count']=sentence_count\nanalysis_df['gunning_fog']=gunning_fog\nanalysis_df['flesch_kincaid_grade']=flesch_kincaid_grade\nanalysis_df['subjectivity']=subjectivity_list\nanalysis_df['polarity']=polarity_list\nanalysis_df['spelling_errors']=spelling_errors", "_____no_output_____" ], [ "analysis_df=analysis_df.drop(analysis_df.columns[8], axis=1)", "_____no_output_____" ], [ "analysis_df['stars']=df_review['stars']\nanalysis_df['useful']=df_review['useful']\nanalysis_df['agg']=df_review['useful']+df_review['funny']+df_review['cool']\nanalysis_df['politness']=df_review['politness']\nanalysis_df['usefull_bin'] = np.where(analysis_df['useful']==0, '0', '1')", "_____no_output_____" ], [ "analysis_df['stars']=df_review['stars']", "_____no_output_____" ], [ "analysis_df.dtypes\nanalysis_df.groupby('usefull_bin').count()\n#analysis_df.to_csv(\"D:/Trinity_DS/Dissertations/201907/yelp/analysis.csv\",index=False)\n#analysis_df.groupby('usefull_bin').mean()\n#analysis_df.groupby('usefull_bin').var()", "_____no_output_____" ], [ "X= np.array(analysis_df.drop(['usefull_bin','agg','useful'], axis=1))\n#X= np.array(analysis_df['smog'])\nY= np.array(analysis_df['usefull_bin'])\nX_train, X_test, y_train, y_test = train_test_split(\n X, Y, test_size=0.33, random_state=303)\nX_train_scaled = preprocessing.scale(X_train)\nX_test_scaled = preprocessing.scale(X_test)", "_____no_output_____" ] ], [ [ "# Decision Tree", "_____no_output_____" ] ], [ [ "clf = tree.DecisionTreeClassifier()\nclf = clf.fit(X_train_scaled, y_train)\nY_train_Pred=clf.predict(X_train_scaled)\naccuracy_score(y_train, Y_train_Pred)", "_____no_output_____" ], [ "Y_test_Pred=clf.predict(X_test_scaled)\ntarget_names = ['0', '1']\nprint(classification_report(y_test, Y_test_Pred, target_names=target_names))", "_____no_output_____" ] ], [ [ "# Support Vector Machines", "_____no_output_____" ] ], [ [ "list(analysis_df)\nlist(analysis_df)", "_____no_output_____" ], [ "list(analysis_df)\nX= np.array(analysis_df.drop(['usefull_bin','agg','useful','date','usefull_diff','flesch_kincaid_grade','subjectivity'], axis=1))\n#X= np.array(analysis_df['smog'])\nY= np.array(analysis_df['usefull_bin'])\nX_train, X_test, y_train, y_test = train_test_split(\n X, Y, test_size=0.33, random_state=303)\nX_train_scaled = preprocessing.scale(X_train)\nX_test_scaled = preprocessing.scale(X_test)", "_____no_output_____" ], [ "clf_SVM = SVC(gamma='auto',kernel='rbf',C=10)\nclf_SVM.fit(X_train_scaled, y_train) ", "_____no_output_____" ], [ "Y_train_Pred=clf_SVM.predict(X_train_scaled)\naccuracy_score(y_train, Y_train_Pred)", "_____no_output_____" ], [ "Y_test_SVM_Pred=clf_SVM.predict(X_test_scaled)\nY_test_SVM_Pred\ntarget_names=['0','1']\nprint(classification_report(y_test, Y_test_SVM_Pred, target_names=target_names))", "_____no_output_____" ] ], [ [ "# Random Forest", "_____no_output_____" ] ], [ [ "clf_RF = RandomForestClassifier(n_estimators=100,random_state=0)\nclf_RF.fit(X_train_scaled, y_train)\nY_train_Pred=clf_RF.predict(X_train_scaled)\nY_test_RF_Pred=clf_RF.predict(X_test_scaled)\nprint(\"Training Accuracy\",accuracy_score(y_train, Y_train_Pred))\ntarget_names=['0','1']\nprint(classification_report(y_test, Y_test_RF_Pred, target_names=target_names))", "_____no_output_____" ], [ "clf_RF = RandomForestClassifier(n_estimators=1000,random_state=0,max_depth=4)\nclf_RF.fit(X_train_scaled, y_train)\nY_train_Pred=clf_RF.predict(X_train_scaled)\nY_test_RF_Pred=clf_RF.predict(X_test_scaled)\nprint(\"Training Accuracy\",accuracy_score(y_train, Y_train_Pred))\ntarget_names=['0','1']\nprint(classification_report(y_test, Y_test_RF_Pred, target_names=target_names))", "_____no_output_____" ], [ "clf_RF = RandomForestClassifier(n_estimators=2000,random_state=0)\nclf_RF.fit(X_train_scaled, y_train)", "_____no_output_____" ], [ "Y_train_Pred=clf_RF.predict(X_train_scaled)\naccuracy_score(y_train, Y_train_Pred)", "_____no_output_____" ], [ "Y_test_RF_Pred=clf_RF.predict(X_test)\ntarget_names=['0','1']\nprint(classification_report(y_test, Y_test_RF_Pred, target_names=target_names))", "_____no_output_____" ] ], [ [ "# Regression Analysis", "_____no_output_____" ] ], [ [ "sns.set(style=\"whitegrid\")\nax = sns.boxplot(x=analysis_df[\"useful\"])\n", "_____no_output_____" ], [ "print(list(analysis_df))\nanalysis_df['log_usefull']=np.log(analysis_df['useful']+1)\nprint(analysis_df.dtypes)\nsns.distplot(analysis_df[\"log_usefull\"],bins=int(180/5), hist=True, kde=True, color = 'darkblue', \n )", "_____no_output_____" ], [ "analysis_df['date'] = pd.to_datetime(df_review['date']).dt.date\n#filtered_df = data_df[data_df['reviews_dateAdded_Date_time'].notnull()]\n#filtered_df = data_df[data_df['reviews_date_Date_time'].notnull()]\n#analysis_df['date']-dt.datetime.now().date()-", "_____no_output_____" ], [ "analysis_df['diff_days'] = dt.datetime.now().date() - analysis_df['date']\nanalysis_df['diff_days']=(analysis_df['diff_days']/np.timedelta64(1,'M'))\n\n\nanalysis_df['usefull_diff'] = (analysis_df['useful']/analysis_df['diff_days'])\nanalysis_df['usefull_diff'].describe()\n#usefull_diff", "_____no_output_____" ], [ "analysis_df['diff_days'] = dt.datetime.now().date() - analysis_df['date']\nanalysis_df['diff_days']=(analysis_df['diff_days']/np.timedelta64(1,'M'))\n#(analysis_df['diff_days']/np.timedelta64(1,'M')).describe()\ncorr_df = analysis_df.drop(['agg','log_usefull','usefull_diff','flesch_kincaid_grade','subjectivity'], axis=1)\ncorr = abs(corr_df.corr())\n\nplt.figure(figsize= (10, 10))\nsns.heatmap(corr_df.corr())\nfig, ax = plt.subplots(figsize=(10, 10)) \nmask = np.zeros_like(corr_df.corr())\n\nmask[np.triu_indices_from(mask)] = 1\nsns.heatmap(corr_df.corr(), mask= mask, ax= ax, annot= True,annot_kws={\"size\": 11},fmt='.2f')\n\n\ncorr = np.round(abs(corr_df.corr()),2)\ncorr.style.background_gradient(cmap='coolwarm') ", "_____no_output_____" ], [ "analysis_df[['subjectivity','polarity']]", "_____no_output_____" ], [ "X= np.array(analysis_df.drop(['usefull_bin','agg','useful','usefull_diff','date','subjectivity'], axis=1))\n#X= np.array(analysis_df['smog'])\nY= np.array(analysis_df['usefull_diff'])\nX_train, X_test, y_train, y_test = train_test_split(\n X, Y, test_size=0.33, random_state=303)\nX_train_scaled = preprocessing.scale(X_train)\nX_test_scaled = preprocessing.scale(X_test)", "_____no_output_____" ], [ "reg = LinearRegression().fit(X_train_scaled, y_train)\ny_pred=reg.predict(X_test_scaled)\n#scaler.transform(X_test)\n(mean_squared_error(y_test, y_pred)**0.5)*100\n", "_____no_output_____" ], [ "\nX= np.array(analysis_df.drop(['usefull_bin','agg','useful','usefull_diff','date','log_usefull','subjectivity'], axis=1))\n#X= np.array(analysis_df['smog'])\nY= np.array(analysis_df['log_usefull'])\nX_train, X_test, y_train, y_test = train_test_split(\n X, Y, test_size=0.33, random_state=303)\nX_train_scaled = preprocessing.scale(X_train)\nX_test_scaled = preprocessing.scale(X_test)\nreg = LinearRegression().fit(X_train_scaled, y_train)\ny_pred=reg.predict(X_test_scaled)\n#scaler.transform(X_test)\nprint(r2_score(y_test, (np.exp(y_pred))))\n(mean_squared_error(y_test, (np.exp(y_pred)-1))**0.5)*100\n", "_____no_output_____" ], [ "#analysis_df.to_csv(\"D:/Trinity_DS/Dissertations/201907/dataset_v2/analysis_df.csv\", index = None, header=True)\nclf_ridge = Ridge(alpha=100000)\nclf_ridge.fit(X_train_scaled, y_train) \ny_pred_0=clf_ridge.predict(X_train_scaled)\n#scaler.transform(X_test)\nmean_squared_error(y_train,np.exp(y_pred_0)-1)*100", "_____no_output_____" ], [ "y_pred=clf_ridge.predict(X_test_scaled)\n#scaler.transform(X_test)\nmean_squared_error(np.exp(y_pred)-1,y_test)*100\n", "_____no_output_____" ], [ "r2_score(y_test,(np.exp(y_pred)-1))", "_____no_output_____" ], [ "print(((np.exp(y_pred))-1).mean())\nprint(y_test.mean())", "_____no_output_____" ], [ "y_pred.std()", "_____no_output_____" ], [ "((sum((y_pred-y_test)**2))/len(y_pred))**0.5*100", "_____no_output_____" ] ], [ [ "# Politness Vs Helpfull and Polite vs Stars", "_____no_output_____" ] ], [ [ "X= np.array(analysis_df['politness'])\n#X= np.array(analysis_df['smog'])\nY= np.array(analysis_df['log_usefull'])\nX_train, X_test, y_train, y_test = train_test_split(\n X, Y, test_size=0.33, random_state=303)\nX_train_scaled = preprocessing.scale(X_train)\nX_test_scaled = preprocessing.scale(X_test)\nreg = LinearRegression().fit(X_train_scaled.reshape(-1, 1), y_train)\ny_pred=reg.predict(X_test_scaled.reshape(-1, 1))\n#scaler.transform(X_test)\nprint(r2_score(y_test, (np.exp(y_pred))-1))\n(mean_squared_error(y_test, (np.exp(y_pred)-1))**0.5)*100", "_____no_output_____" ] ], [ [ "# Merged Analysis", "_____no_output_____" ] ], [ [ "list(analysis_df)\nreview_roll_up_df=analysis_df.drop(['usefull_bin','date','diff_days'], axis=1)\nreview_roll_up_df['user_id'] = df_review['user_id']\n#business_id\nreview_roll_up_df['business_id'] = df_review['business_id']\ng1 = review_roll_up_df.groupby(['user_id','business_id']).sum()", "_____no_output_____" ], [ "review_roll_up_df=g1.reset_index()\njoin_review_user=pd.merge(df_user, review_roll_up_df,on='user_id')", "_____no_output_____" ], [ "list(join_review_user)\njoin_review_user=join_review_user.drop([\n 'user_id',\n 'name',\n 'yelping_since',\n 'funny',\n 'cool',\n 'fans',\n 'compliment_hot',\n 'compliment_more',\n 'compliment_profile',\n 'compliment_cute',\n 'compliment_list',\n 'compliment_note',\n 'compliment_plain',\n 'compliment_cool',\n 'compliment_funny',\n 'compliment_writer',\n 'compliment_photos'\n], axis=1)\ng2 = join_review_user.groupby(['business_id']).sum()\nbusiness_lvl_review_user=g2.reset_index()", "_____no_output_____" ], [ "join_review_user=pd.merge(df_business.dropna(subset='business_id'),business_lvl_review_user.dropna(subset='business_id') ,on='business_id')", "_____no_output_____" ], [ "list(analysis_df)", "_____no_output_____" ], [ "join_review_user=pd.merge(df_user, review_roll_up_df1,on='user_id')\nlist(join_review_user)\njoin_review_user['business_id']", "_____no_output_____" ], [ "merged_final_df=pd.merge(df_business, join_review_user,on='business_id',how='right')", "_____no_output_____" ], [ "from spellchecker import SpellChecker\n\nspell = SpellChecker() # loads default word frequency list\n\n\n# if I just want to make sure some words are not flagged as misspelled\nspell.word_frequency.load_words(['microsoft', 'apple', 'google'])\nspell.known(['microsoft', 'google','asd']) # will return both now!", "_____no_output_____" ], [ "misspelled = spell.unknown(['something', 'is', 'hapenning', 'here'])\nlen(misspelled)", "_____no_output_____" ] ] ]

[ "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7e1017e65e9280d6382e312e25f6bba07d039f

258,617

ipynb

Jupyter Notebook

Final_Project/Annotations.ipynb

alirezash97/Machine-Learning-Course

a1f07c180992c2a605f355c432c8f59e69f225a5

[ "MIT" ]

1

2020-11-30T14:53:39.000Z

Final_Project/Annotations.ipynb

alirezash97/Machine-Learning-Course

a1f07c180992c2a605f355c432c8f59e69f225a5

[ "MIT" ]

null

Final_Project/Annotations.ipynb

alirezash97/Machine-Learning-Course

a1f07c180992c2a605f355c432c8f59e69f225a5

[ "MIT" ]

null

140.323928

83,614

0.800442

[ [ [ "<a href=\"https://colab.research.google.com/github/alirezash97/Machine-Learning-Course/blob/main/Final_Project/Annotations.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>", "_____no_output_____" ] ], [ [ "# from google.colab import drive\n# drive.mount('/content/drive')", "Mounted at /content/drive\n" ], [ "# !wget 'https://storage.googleapis.com/kaggle-competitions-data/kaggle-v2/23870/1781260/compressed/train.zip?GoogleAccessId=web-data@kaggle-161607.iam.gserviceaccount.com&Expires=1615290685&Signature=Io6pu1QbSFcOxSiCzQY%2BFCMzOtM44LAo2whB9pg3HCxEZb%2B7trwjhfDcSUQSVVIW7myMOe9X6ZAc4XEw8Zk6AWC8vtFXo5V%2FaYDTGuY%2Fa2pbz8iR0uEoFZLR5kvMSh669u1SAFbGGN8lxkqwGy8OLCXgBhS9k1pEh3yWF57JlvAgSS0MQ5mWOeYXgMB95QZqrBis4o5ans7JGXq9rjyYmhEzgecfQ8oH%2F%2Bjn8aOFA8LIIqCtU2bUp95NReFnh1L%2F4i6sxemRFdFv4buBx1TwNI6vB1FMFfstwF7ooB5Ft5V1JMtPskRP%2BkKDP71EBWQQEl%2FpUZnUxYxJd0eo8jb76w%3D%3D&response-content-disposition=attachment%3B+filename%3Dtrain.zip'\r\n\r\n\r\n\r\n# !wget 'https://storage.googleapis.com/kaggle-competitions-data/kaggle-v2/23870/1781260/compressed/train_annotations.csv.zip?GoogleAccessId=web-data@kaggle-161607.iam.gserviceaccount.com&Expires=1615290713&Signature=mva9XeQQT0So9jwmVVGijIGUn9roXIXjYoC2ZC8yKtROD3C%2FJfdrn5aHKzAfTULE0sCVuPdrTg8Vd%2BePI3h2Y0u8ZNwuyjC0sRhR%2FJh0RgqGrkxS0wWVthcg9wXf4utQYiOPvxQSM%2FRszJS6z7kiwAs9XcDeaEvpx1UBN2lOEcGLHJS6ZhwD7S7jnvneTFBF1rylK5DX%2Bn1TheuacV1fFT%2BOZeZxJvR7i5Oxf%2BpuJW0glnwucHHN%2Fw6nSpPWaYPzFbbGJ%2BpNbw%2B5NR%2FJpJI4MYX9npoiEJF2J4mxSXOqrNvUCc%2BVWsq1kGQa5se337uMuFY4%2FV7a%2B5jNeOqTImiXIw%3D%3D&response-content-disposition=attachment%3B+filename%3Dtrain_annotations.csv.zip'", "_____no_output_____" ], [ "# !mkdir /content/trainset\r\n# !mkdir /content/trainset/data/\r\n# !mkdir /content/trainset/data/1/\r\n# !unzip '/content/train.zip' -d /content/trainset/data/1/\r\n# !unzip '/content/train_annotations.csv.zip' -d /content/trainset/annotations/", "_____no_output_____" ], [ "import pandas as pd \r\nannotations = pd.read_csv('/content/trainset/annotations/train_annotations.csv')\r\nannotations.head()", "_____no_output_____" ], [ "len(annotations)", "_____no_output_____" ], [ "import re\r\nimport ast\r\nimport numpy as np\r\ndef str2array(s):\r\n # Remove space after [\r\n s=re.sub('\\[ +', '[', s.strip())\r\n # Replace commas and spaces\r\n s=re.sub('[,\\s]+', ', ', s)\r\n return np.array(ast.literal_eval(s))", "_____no_output_____" ], [ "import numpy as np \r\n\r\n##############\r\nmsk_for_dataset_subset = np.random.rand(len(annotations)) < 1\r\ndataset_subset = annotations[msk_for_dataset_subset]\r\n##############\r\n\r\nmsk = np.random.rand(len(dataset_subset)) > 0.25\r\ntrain_samples = dataset_subset[msk]\r\nvalidation_samples = dataset_subset[~msk]\r\n\r\n##############\r\n# train_samples = train_samples[:5000]\r\n# validation_samples = validation_samples[:1500]\r\n##############\r\nprint('number of train samples: ', len(train_samples))\r\nprint('number of validation samples: ', len(validation_samples))\r\n\r\n\r\n\r\nsample = dataset_subset.iloc[1, :]\r\nlandmarks = sample['data']\r\n# print(landmarks)\r\nlandmarks = np.array(str2array(landmarks))\r\n# print(type(landmarks))\r\n# landmarks = np.array(list(landmarks))\r\nprint(\"sample landmark shape: \", landmarks.shape)\r\n# print(landmarks)\r\n\r\n# print('Image name: {}'.format(img_name))\r\n# print('Landmarks shape: {}'.format(landmarks.shape))\r\n# print('First 4 Landmarks: {}'.format(landmarks[:4]))", "number of train samples: 13449\nnumber of validation samples: 4550\nsample landmark shape: (13, 2)\n" ], [ "def distance(x1, x2, y1, y2):\r\n return np.sqrt( (x1-x2)**2 + (y1-y2)**2 )", "_____no_output_____" ], [ "from PIL import Image\r\nimport random\r\nimport torch.nn.functional as F\r\nfrom math import cos, sin, radians\r\nimport imutils\r\nimport cv2\r\nfrom matplotlib import cm\r\nimport scipy.misc\r\nfrom skimage.draw import line\r\n\r\nclass RANZCRDataset():\r\n\r\n\r\n def __init__(self, csv_file='/content/trainset/train.csv', root_dir='/content/trainset/data/1', transform=None,\r\n transform_label=None, images_name=None):\r\n \r\n \r\n \"\"\"\r\n Args:\r\n csv_file (string): Path to the csv file with annotations.\r\n root_dir (string): Directory with all the images.\r\n transform (callable, optional): Optional transform to be applied\r\n on a sample.\r\n \"\"\"\r\n self.Images_name = images_name\r\n self.root_dir = root_dir\r\n self.transform = transform\r\n self.transform_label = transform_label\r\n \r\n\r\n def __len__(self):\r\n return len(self.Images_name)\r\n\r\n #############\r\n\r\n\r\n # def get_rot_mat(self, theta):\r\n\r\n # theta = torch.tensor(theta)\r\n # return torch.tensor([[torch.cos(theta), -torch.sin(theta), 0],\r\n # [torch.sin(theta), torch.cos(theta), 0]])\r\n\r\n\r\n\r\n # def rot_img_landmark(self, x, landmarks, theta, dtype):\r\n # rot_mat = self.get_rot_mat(theta)[None, ...].type(dtype).repeat(x.shape[0],1,1)\r\n # grid = F.affine_grid(rot_mat, x.size()).type(dtype)\r\n # image = F.grid_sample(x, grid)\r\n # landmarks = landmarks - 0.5\r\n # new_landmarks = np.matmul(landmarks, transformation_matrix)\r\n # new_landmarks = new_landmarks + 0.5\r\n # return image, new_landmarks\r\n\r\n\r\n ##############\r\n\r\n def __getitem__(self, idx):\r\n centerCrop_value = 256\r\n if torch.is_tensor(idx):\r\n idx = idx.tolist()\r\n\r\n img_name = os.path.join(self.root_dir,\r\n self.Images_name.iloc[idx, 0])\r\n image = Image.open(img_name + '.jpg').convert('RGB')\r\n labels = self.Images_name.iloc[idx, -1]\r\n labels = torch.from_numpy(str2array(labels))\r\n sample = {'image': image, 'label': labels }\r\n\r\n\r\n if self.transform:\r\n\r\n \r\n tmp = np.zeros((100, 2))\r\n for i in range(0, (sample['label'].shape[0]) ):\r\n \r\n tmp[i, 0] = ( (centerCrop_value / np.array(image).shape[1]) * np.array(sample['label'])[i, 0] ) \r\n tmp[i, 1] = ( (centerCrop_value / np.array(image).shape[0]) * np.array(sample['label'])[i, 1] ) \r\n\r\n ############ just for show\r\n temp = np.zeros((100, 2))\r\n for i in range(0, (sample['label'].shape[0]) ):\r\n \r\n temp[i, 0] = ( (1032 / np.array(image).shape[1]) * np.array(sample['label'])[i, 0] ) \r\n temp[i, 1] = ( (1032 / np.array(image).shape[0]) * np.array(sample['label'])[i, 1] ) \r\n sample['landmark'] = torch.from_numpy(temp).type(torch.float16)\r\n ##########################\r\n \r\n \r\n\r\n ############################## landmark to segment\r\n label_img = self.transform_label(sample['image']).numpy()\r\n label_img = np.zeros((label_img.shape[1], label_img.shape[2]))\r\n \r\n for index_point1, landmark_point1 in enumerate(tmp):\r\n \r\n if (landmark_point1 != 0).any() :\r\n distance_dict = {}\r\n for index_point2, landmark_point2 in enumerate(tmp[index_point1+1:]):\r\n if (landmark_point2 != 0).any() :\r\n distance_dict[index_point2] = distance(landmark_point1[0], landmark_point2[0],\r\n landmark_point1[1], landmark_point2[1])\r\n else:\r\n pass\r\n if distance_dict:\r\n my_point = min(distance_dict, key=distance_dict.get)\r\n rr, cc = line(int(landmark_point1[0]), int(landmark_point1[1]), \r\n int(tmp[(index_point1+my_point+1), 0]), int(tmp[(index_point1+my_point+1), 1]))\r\n\r\n for index, value in enumerate(rr):\r\n\r\n if value == 256:\r\n rr[index]-=3\r\n elif value == 255:\r\n rr[index]-=2\r\n elif value == 254:\r\n rr[index]-=1\r\n elif value == 0:\r\n rr[index]+=2\r\n elif value == 1:\r\n rr[index]+=1\r\n else: \r\n pass\r\n for index, value in enumerate(cc):\r\n\r\n if value == 256:\r\n cc[index]-=3\r\n elif value == 255:\r\n cc[index]-=2\r\n elif value == 254:\r\n cc[index]-=1\r\n elif value == 0:\r\n cc[index]+=2\r\n elif value == 1:\r\n cc[index]+=1\r\n else: \r\n pass\r\n # place ones lines\r\n label_img[cc-2, rr-2] = 1\r\n label_img[cc-1, rr-1] = 1\r\n label_img[cc, rr] = 1\r\n label_img[cc+1, rr+1] = 1\r\n label_img[cc+2, rr+2] = 1\r\n\r\n else:\r\n pass\r\n ####################################################\r\n\r\n\r\n # sample['label'] = torch.from_numpy(tmp).type(torch.float16)\r\n\r\n sample['label'] = torch.from_numpy(label_img).type(torch.long)\r\n\r\n sample['image'] = self.transform(sample['image'])\r\n\r\n \r\n\r\n\r\n # # random rotation\r\n # image, landmark = self.rot_img_landmark(sample['image'], sample['label'], np.pi/2, dtype= torch.FloatTensor)\r\n # print(type(image), image.shape)\r\n # print(type(landmark), landmark.shape)\r\n \r\n\r\n return sample\r\n\r\n# my_dataset = RANZCRDataset\r\n# my_dataset.__getitem__(self, 4)", "_____no_output_____" ], [ "\r\nfrom torch.utils.data import Dataset, DataLoader\r\nfrom torchvision import transforms, utils\r\nimport torch\r\nimport torchvision\r\nimport os\r\n\r\n# batch_size = 8\r\nmean = np.array([0.4823, 0.4823, 0.4823])\r\nstd = np.array([0.191473164, 0.191473164, 0.191473164])\r\n\r\n\r\ndef load_data(csv_file='/content/trainset/annotations/train_annotations.csv', root_dir='/content/trainset/data/1'):\r\n\r\n centerCrop_value = 1032\r\n transform = transforms.Compose([transforms.ToTensor(),\r\n transforms.Resize((1056, 1056)),\r\n transforms.CenterCrop(centerCrop_value),\r\n transforms.Normalize(mean, std)])\r\n \r\n transform_label = transforms.Compose([transforms.ToTensor(),\r\n transforms.Resize((270, 270)),\r\n transforms.CenterCrop(256),\r\n transforms.Normalize(mean, std)])\r\n\r\n\r\n\r\n trainset = RANZCRDataset(csv_file='/content/trainset/annotations/train_annotations.csv',\r\n root_dir='/content/trainset/data/1', transform=transform, \r\n transform_label = transform_label, images_name=train_samples)\r\n\r\n\r\n\r\n\r\n validation_set = RANZCRDataset(csv_file='/content/trainset/annotations/train_annotations.csv',\r\n root_dir='/content/trainset/data/1', transform=transform,\r\n transform_label = transform_label, images_name=validation_samples)\r\n \r\n\r\n return trainset, validation_set\r\n\r\n\r\n\r\n\r\n", "_____no_output_____" ], [ "# trainset, testset = load_data()\r\n\r\n\r\n# train_loader = torch.utils.data.DataLoader(trainset,\r\n# batch_size=2,\r\n# num_workers=0,\r\n# shuffle=True)\r\n\r\n\r\n\r\n# validation_loader = torch.utils.data.DataLoader(testset,\r\n# batch_size=2,\r\n# num_workers=0,\r\n# shuffle=True)\r\n", "_____no_output_____" ], [ "import matplotlib.pyplot as plt\r\nimport copy\r\n\r\ncenterCrop_value = 904\r\ndef imshow_landmark(img, landmarks):\r\n npimg = img.numpy()\r\n npimg = ((npimg * std[0]) + mean[0]) # unnormalize\r\n plt.imshow((np.transpose(npimg, (1, 2, 0)) * 255).astype(np.uint8))\r\n show_landmark = copy.deepcopy(landmarks)\r\n for i in range(show_landmark.shape[0]):\r\n show_landmark[i, :, 0] = show_landmark[i, :, 0] + (centerCrop_value*i)\r\n plt.scatter(show_landmark[:, :, 0], show_landmark[:, :, 1], s=10, marker='.', c='r')\r\n plt.pause(0.001) # pause a bit so that plots are updated\r\n plt.show()\r\n\r\n", "_____no_output_____" ], [ "import matplotlib.pyplot as plt\r\n\r\n\r\ndef imshow_segmented(segment_label):\r\n \r\n # npimg = img.numpy()\r\n # npimg = ((npimg * std[0]) + mean[0]) # unnormalize\r\n # plt.imshow((np.transpose(npimg, (1, 2, 0)) * 255).astype(np.uint8))\r\n \r\n\r\n npimg = (segment_label.detach().cpu().numpy()*255)\r\n dummy_channel = np.zeros((1, npimg.shape[1], npimg.shape[2]))\r\n npimg = np.concatenate([npimg, dummy_channel], axis=0)\r\n npimg = ((npimg * std[0]) + mean[0]) # unnormalize\r\n plt.imshow((np.transpose(npimg, (1, 2, 0)) * 255).astype(np.uint8))\r\n \r\n plt.show()\r\n\r\ndef imshow(img):\r\n \r\n npimg = img.numpy()\r\n npimg = ((npimg * std[0]) + mean[0]) # unnormalize\r\n plt.imshow((np.transpose(npimg, (1, 2, 0)) * 255).astype(np.uint8))\r\n\r\n plt.show()\r\n\r\n\r\n# get some random training images\r\n# dataiter = iter(train_loader)\r\n# sample = dataiter.next()\r\n\r\n\r\n# print(sample['image'].shape)\r\n# imshow(torchvision.utils.make_grid(sample['image']))\r\n\r\n# print(sample['label'].shape)\r\n# imshow_segmented(torchvision.utils.make_grid(sample['label']))\r\n\r\n# print(sample['landmark'].shape)\r\n# imshow_landmark(torchvision.utils.make_grid(sample['image']), sample['landmark'])\r\n", "_____no_output_____" ], [ "# class Network(nn.Module):\r\n \r\n# def __init__(self):\r\n# super(Network, self).__init__()\r\n# self.model = model\r\n# self.conv1 = nn.Conv2d(3, 3, 5)\r\n# self.conv2 = nn.Conv2d(3, 3, 1)\r\n# self.pool2 = nn.MaxPool2d(2, 2)\r\n# self.sigmoid = nn.Sigmoid()\r\n# self.fc_final = nn.Linear(1000, 11)\r\n\r\n# def forward(self, x):\r\n\r\n# x = self.pool2(F.relu(self.conv1(x)))\r\n# x = self.pool2(F.relu(self.conv2(x)))\r\n# x = self.model(x)\r\n# x = self.sigmoid(self.fc_final(x))\r\n# return x\r\n\r\n# Network = Network()", "_____no_output_____" ], [ "import torch.nn as nn\r\nfrom torchvision import models\r\nfrom torch import optim\r\nimport time\r\n\r\nmodel = torchvision.models.segmentation.deeplabv3_resnet50(pretrained=False, progress=True, num_classes=2, aux_loss=None)\r\n", "Downloading: \"https://download.pytorch.org/models/resnet50-19c8e357.pth\" to /root/.cache/torch/hub/checkpoints/resnet50-19c8e357.pth\n" ], [ "model", "_____no_output_____" ], [ "import torch.nn as nn\r\nimport torch.nn.functional as F \r\n\r\nclass Net(nn.Module):\r\n \r\n def __init__(self, c1=3):\r\n super(Net, self).__init__()\r\n self.model = model\r\n self.conv1 = nn.Conv2d(3, c1, 5)\r\n self.conv2 = nn.Conv2d(c1, 3, 3)\r\n self.pool2 = nn.MaxPool2d(2, 2)\r\n\r\n def forward(self, x):\r\n\r\n x = self.pool2(F.relu(self.conv1(x)))\r\n x = self.pool2(F.relu(self.conv2(x)))\r\n x = self.model(x)\r\n\r\n return x\r\n", "_____no_output_____" ], [ "\r\n# ###################################\r\n# !pip install ray\r\n# !pip install tensorboardX", "_____no_output_____" ], [ "from functools import partial\r\nimport numpy as np\r\nimport os\r\nimport torch\r\nimport torch.nn as nn\r\nimport torch.nn.functional as F\r\nimport torch.optim as optim\r\nfrom torch.utils.data import random_split\r\nimport torchvision\r\nimport torchvision.transforms as transforms\r\nfrom ray import tune\r\nfrom ray.tune import CLIReporter\r\nfrom ray.tune.schedulers import ASHAScheduler\r\nimport tensorboardX", "_____no_output_____" ], [ "from sklearn.metrics import accuracy_score\r\n\r\ndef train_ranzcr_landmark(config, checkpoint_dir=None, data_dir=None):\r\n net = Net()\r\n checkpoint_dir = '/content/drive/MyDrive/RANZCR/'\r\n\r\n device = \"cpu\"\r\n if torch.cuda.is_available():\r\n device = \"cuda:0\"\r\n if torch.cuda.device_count() > 1:\r\n net = nn.DataParallel(net)\r\n net.to(device)\r\n\r\n criterion = nn.CrossEntropyLoss()\r\n optimizer = optim.SGD(net.parameters(), lr=config[\"lr\"], weight_decay=config[\"wd\"])\r\n\r\n if checkpoint_dir:\r\n model_state, optimizer_state = torch.load(\r\n os.path.join(checkpoint_dir, 'model_landmark_11_0.04266818.pth'))\r\n net.load_state_dict(model_state)\r\n optimizer.load_state_dict(optimizer_state)\r\n\r\n trainset, testset = load_data(data_dir)\r\n\r\n test_abs = int(len(trainset) * 0.8)\r\n train_subset, val_subset = random_split(\r\n trainset, [test_abs, len(trainset) - test_abs])\r\n \r\n trainloader = torch.utils.data.DataLoader(\r\n train_subset,\r\n batch_size=int(config[\"batch_size\"]),\r\n shuffle=True,\r\n num_workers=8)\r\n valloader = torch.utils.data.DataLoader(\r\n val_subset,\r\n batch_size=int(config[\"batch_size\"]),\r\n shuffle=True,\r\n num_workers=8)\r\n\r\n for epoch in range(15): # loop over the dataset multiple times\r\n running_loss = 0.0\r\n epoch_steps = 0\r\n for i, data in enumerate(trainloader, 0):\r\n # get the inputs; data is a list of [inputs, labels]\r\n inputs, labels = data['image'].float(), data['label']\r\n \r\n # batch_size = int(inputs.shape[0])\r\n # labels = np.zeros((batch_size, 200))\r\n # for i in range(batch_size):\r\n # labels[i, :100] = labels_temp[i, :, 0]\r\n # labels[i, 100:] = labels_temp[i, :, 1]\r\n\r\n # labels = (torch.from_numpy(labels)).type(torch.float16)\r\n inputs, labels = inputs.to(device), labels.to(device)\r\n # zero the parameter gradients\r\n optimizer.zero_grad()\r\n\r\n # forward + backward + optimize\r\n \r\n net.eval()\r\n outputs = net(inputs)['out']\r\n loss = criterion(outputs, labels)\r\n loss.backward()\r\n optimizer.step()\r\n\r\n # print statistics\r\n running_loss += loss.item()\r\n epoch_steps += 1\r\n if i % 2000 == 1999: # print every 2000 mini-batches\r\n print(\"[%d, %5d] loss: %.3f\" % (epoch + 1, i + 1,\r\n running_loss / epoch_steps))\r\n running_loss = 0.0\r\n\r\n # Validation loss\r\n val_loss = 0.0\r\n val_steps = 0\r\n total_train = 0\r\n correct_train = 0\r\n accuracy = 0\r\n batch_counter = 0\r\n for i, data in enumerate(valloader, 0):\r\n with torch.no_grad():\r\n inputs, labels = data['image'].float(), data['label']\r\n # batch_size = int(inputs.shape[0])\r\n # labels = np.zeros((batch_size, 200))\r\n # for i in range(batch_size):\r\n # labels[i, :100] = labels_temp[i, :, 0]\r\n # labels[i, 100:] = labels_temp[i, :, 1]\r\n\r\n # labels = (torch.from_numpy(labels)).type(torch.float16)\r\n inputs, labels = inputs.to(device), labels.to(device)\r\n\r\n net.eval()\r\n outputs = net(inputs)['out']\r\n ############\r\n _, predicted = torch.max(outputs.data, 1)\r\n total_train += labels.nelement()\r\n correct_train += predicted.eq(labels.data).sum().item()\r\n train_accuracy = 100 * correct_train / total_train\r\n #############\r\n\r\n batch_counter += 1\r\n \r\n loss = criterion(outputs, labels)\r\n val_loss += loss.cpu().numpy()\r\n val_steps += 1\r\n\r\n with tune.checkpoint_dir(epoch) as checkpoint_dir:\r\n path = os.path.join(checkpoint_dir, \"checkpoint\")\r\n torch.save((net.state_dict(), optimizer.state_dict()), path)\r\n\r\n tune.report(loss=(val_loss / val_steps), accuracy=(train_accuracy))\r\n torch.save((net.state_dict(), optimizer.state_dict()), '/content/drive/MyDrive/RANZCR/model_landmark_%.8f.pth'%((val_loss / val_steps)))\r\n print(\"Finished Training\")\r\n\r\n\r\n ", "_____no_output_____" ], [ "\r\ndef test_accuracy(net, device=\"cpu\"):\r\n trainset, testset = load_data()\r\n\r\n criterion = nn.CrossEntropyLoss()\r\n testloader = torch.utils.data.DataLoader(\r\n testset, batch_size=4, shuffle=False, num_workers=2)\r\n\r\n correct_test = 0\r\n total_test = 0\r\n batch_counter_test = 0\r\n accuracy_test = 0\r\n \r\n for data in testloader:\r\n with torch.no_grad():\r\n\r\n\r\n inputs, labels = data['image'].float(), data['label']\r\n batch_size = int(inputs.shape[0])\r\n batch_counter_test += 1 \r\n # labels = np.zeros((batch_size, 200))\r\n # for i in range(batch_size):\r\n # labels[i, :100] = labels_temp[i, :, 0]\r\n # labels[i, 100:] = labels_temp[i, :, 1]\r\n\r\n # labels = (torch.from_numpy(labels)).type(torch.float16)\r\n inputs, labels = inputs.to(device), labels.to(device)\r\n\r\n net.eval()\r\n outputs = outputs = net(inputs)['out']\r\n # _, predicted = torch.max(outputs.data, 1)\r\n # total += labels.size(0)\r\n # correct += (predicted == labels).sum().item()\r\n # my_validation_outputs = (outputs > 0.5)\r\n # print('------------------------------------------------------------------------------')\r\n # print(labels)\r\n # print(my_validation_outputs)\r\n # print('------------------------------------------------------------------------------')\r\n # accuracy += auc_s(labels, my_validation_outputs)\r\n # batch_counter += 1\r\n ############\r\n _, predicted = torch.max(outputs.data, 1)\r\n total_test += labels.nelement()\r\n correct_test += predicted.eq(labels.data).sum().item()\r\n test_accuracy = 100 * correct_test / total_test\r\n #############\r\n\r\n\r\n return test_accuracy", "_____no_output_____" ], [ "\r\ndef main(num_samples=10, max_num_epochs=10, gpus_per_trial=2):\r\n data_dir = os.path.abspath(\"/content/trainset/data\")\r\n load_data(data_dir)\r\n config = {\r\n \"lr\": tune.choice([0.0422581]),\r\n \"wd\": tune.choice([4.12557e-06]),\r\n \"c1\": tune.choice([3]),\r\n \"batch_size\": tune.choice([8])\r\n }\r\n scheduler = ASHAScheduler(\r\n metric=\"loss\",\r\n mode=\"min\",\r\n max_t=max_num_epochs,\r\n grace_period=1,\r\n reduction_factor=2)\r\n reporter = CLIReporter(\r\n # parameter_columns=[\"l1\", \"l2\", \"lr\", \"batch_size\"],\r\n metric_columns=[\"loss\", \"accuracy\", \"training_iteration\"])\r\n\r\n result = tune.run(\r\n partial(train_ranzcr_landmark, data_dir=data_dir),\r\n resources_per_trial={\"cpu\": 2, \"gpu\": gpus_per_trial},\r\n config=config,\r\n num_samples=num_samples,\r\n scheduler=scheduler,\r\n progress_reporter=reporter)\r\n \r\n best_trial = result.get_best_trial(\"loss\", \"min\", \"last\")\r\n print(\"Best trial config: {}\".format(best_trial.config))\r\n print(\"Best trial final validation loss: {}\".format(\r\n best_trial.last_result[\"loss\"]))\r\n print(\"Best trial final validation accuracy: {}\".format(\r\n best_trial.last_result[\"accuracy\"]))\r\n\r\n best_trained_model = Net()\r\n device = \"cpu\"\r\n if torch.cuda.is_available():\r\n device = \"cuda:0\"\r\n if gpus_per_trial > 1:\r\n best_trained_model = nn.DataParallel(best_trained_model)\r\n best_trained_model.to(device)\r\n\r\n best_checkpoint_dir = best_trial.checkpoint.value\r\n model_state, optimizer_state = torch.load(os.path.join(best_checkpoint_dir, \"checkpoint\"))\r\n best_trained_model.load_state_dict(model_state)\r\n test_acc = test_accuracy(best_trained_model, device)\r\n torch.save(best_trained_model.state_dict(), '/content/drive/MyDrive/RANZCR/model_landmark_%.3f.pth'%(test_acc) )\r\n print(\"Best trial test set accuracy: {}\".format(test_acc))\r\n\r\n\r\nif __name__ == \"__main__\":\r\n # You can change the number of GPUs per trial here:\r\n main(num_samples=1, max_num_epochs=20, gpus_per_trial=1)", "2021-03-06 12:23:31,300\tINFO services.py:1174 -- View the Ray dashboard at \u001b[1m\u001b[32mhttp://127.0.0.1:8265\u001b[39m\u001b[22m\n2021-03-06 12:23:34,125\tWARNING experiment.py:292 -- No name detected on trainable. Using DEFAULT.\n2021-03-06 12:23:34,126\tINFO registry.py:65 -- Detected unknown callable for trainable. Converting to class.\n2021-03-06 12:23:40,164\tWARNING worker.py:1107 -- Warning: The actor ImplicitFunc has size 163342186 when pickled. It will be stored in Redis, which could cause memory issues. This may mean that its definition uses a large array or other object.\n2021-03-06 12:23:40,356\tWARNING util.py:152 -- The `start_trial` operation took 2.644 s, which may be a performance bottleneck.\n" ], [ "# def count_parameters(model):\r\n# return sum(p.numel() for p in model.parameters() if p.requires_grad)\r\n# print(count_parameters(model))\r\n# print(count_parameters(Network))", "_____no_output_____" ], [ "# trained_model = torch.load()\r\nmy_net = Net()\r\nnet_std, net_optdict = torch.load('/content/drive/MyDrive/RANZCR/model_landmark_11_0.04266818.pth')\r\nmy_net.load_state_dict(net_std)", "_____no_output_____" ], [ "\r\nfrom torch.autograd import Variable\r\n\r\n_, my_testset = load_data()\r\n\r\nmy_testloader = torch.utils.data.DataLoader(\r\n my_testset, batch_size=4, shuffle=False, num_workers=2)\r\n\r\ndataiter = iter(my_testloader)\r\nsample = dataiter.next()\r\n\r\n\r\n\r\nprint('input : ', sample['image'].shape)\r\nimshow(torchvision.utils.make_grid(sample['image']))\r\n\r\n\r\nprint('landmarks : ',sample['landmark'].shape)\r\nimshow_landmark(torchvision.utils.make_grid(sample['image']), sample['landmark'])\r\n\r\n\r\nmy_labels = torch.unsqueeze(sample['label'], 1)\r\nprint('segmentation label : ',my_labels.shape)\r\nimshow_segmented(torchvision.utils.make_grid(my_labels))\r\n\r\n\r\n####################################\r\ndevice = \"cpu\"\r\nif torch.cuda.is_available():\r\n device = \"cuda:0\"\r\n if torch.cuda.device_count() > 1:\r\n my_net = nn.DataParallel(my_net)\r\nmy_net.to(device)\r\n\r\n\r\ninputs, labels = sample['image'].float(), sample['label']\r\nbatch_size = int(inputs.shape[0])\r\ninputs, labels = inputs.to(device), labels.to(device)\r\nmy_outputs = my_net(inputs)['out']\r\noutputs = torch.squeeze(my_outputs, 0)\r\nprint('predicted outputs : ',outputs.shape)\r\nimshow_segmented(torchvision.utils.make_grid(outputs))", "input : torch.Size([4, 3, 1032, 1032])\n" ], [ "", "_____no_output_____" ] ] ]

[ "markdown", "code" ]

[ [ "markdown" ], [ "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code", "code" ] ]

ec7e173e38474993bb2e091c8059c7bcf3c5f461

12,148

ipynb

Jupyter Notebook

colabs/monthly_budget_mover.ipynb

danieldjewell/starthinker

3327d5874f01d7563603b8a82c1ecd6615b9768d

[ "Apache-2.0" ]

1

2020-12-04T17:13:35.000Z

colabs/monthly_budget_mover.ipynb

danieldjewell/starthinker

3327d5874f01d7563603b8a82c1ecd6615b9768d

[ "Apache-2.0" ]

null

colabs/monthly_budget_mover.ipynb

danieldjewell/starthinker

3327d5874f01d7563603b8a82c1ecd6615b9768d

[ "Apache-2.0" ]

null

46.015152

264

0.505186

[ [ [ "#1. Install Dependencies\nFirst install the libraries needed to execute recipes, this only needs to be done once, then click play.\n", "_____no_output_____" ] ], [ [ "!pip install git+https://github.com/google/starthinker\n", "_____no_output_____" ] ], [ [ "#2. Get Cloud Project ID\nTo run this recipe [requires a Google Cloud Project](https://github.com/google/starthinker/blob/master/tutorials/cloud_project.md), this only needs to be done once, then click play.\n", "_____no_output_____" ] ], [ [ "CLOUD_PROJECT = 'PASTE PROJECT ID HERE'\n\nprint(\"Cloud Project Set To: %s\" % CLOUD_PROJECT)\n", "_____no_output_____" ] ], [ [ "#3. Get Client Credentials\nTo read and write to various endpoints requires [downloading client credentials](https://github.com/google/starthinker/blob/master/tutorials/cloud_client_installed.md), this only needs to be done once, then click play.\n", "_____no_output_____" ] ], [ [ "CLIENT_CREDENTIALS = 'PASTE CREDENTIALS HERE'\n\nprint(\"Client Credentials Set To: %s\" % CLIENT_CREDENTIALS)\n", "_____no_output_____" ] ], [ [ "#4. Enter Monthly Budget Mover Parameters\nApply the previous month's budget/spend delta to the current month. Aggregate up the budget and spend from the previous month of each category declared then apply the delta of the spend and budget equally to each Line Item under that Category.\n 1. No changes made can be made in DV360 from the start to the end of this process\n 1. Make sure there is budget information for the current and previous month's IOs in DV360\n 1. Make sure the provided spend report has spend data for every IO in the previous month\n 1. Spend report must contain 'Revenue (Adv Currency)' and 'Insertion Order ID'\n 1. There are no duplicate IO Ids in the categories outlined below\n 1. This process must be ran during the month of the budget it is updating\n 1. If you receive a 502 error then you must separate your jobs into two, because there is too much information being pulled in the sdf\n 1. Manually run this job\n 1. Once the job has completed go to the table for the new sdf and export to a csv\n 1. Take the new sdf and upload it into DV360\nModify the values below for your use case, can be done multiple times, then click play.\n", "_____no_output_____" ] ], [ [ "FIELDS = {\n 'recipe_timezone': 'America/Los_Angeles', # Timezone for report dates.\n 'recipe_name': '', # \n 'auth_write': 'service', # Credentials used for writing data.\n 'auth_read': 'user', # Credentials used for reading data.\n 'partner_id': '', # The sdf file types.\n 'budget_categories': '{}', # A dictionary to show which IO Ids go under which Category. {\"CATEGORY1\":[12345,12345,12345], \"CATEGORY2\":[12345,12345]}\n 'filter_ids': [], # Comma separated list of filter ids for the request.\n 'excluded_ios': '', # A comma separated list of Inserion Order Ids that should be exluded from the budget calculations\n 'version': '5', # The sdf version to be returned.\n 'is_colab': True, # Are you running this in Colab? (This will store the files in Colab instead of Bigquery)\n 'dataset': '', # Dataset that you would like your output tables to be produced in.\n}\n\nprint(\"Parameters Set To: %s\" % FIELDS)\n", "_____no_output_____" ] ], [ [ "#5. Execute Monthly Budget Mover\nThis does NOT need to be modified unles you are changing the recipe, click play.\n", "_____no_output_____" ] ], [ [ "from starthinker.util.project import project\nfrom starthinker.script.parse import json_set_fields\n\nUSER_CREDENTIALS = '/content/user.json'\n\nTASKS = [\n {\n 'dataset': {\n 'description': 'Create a dataset where data will be combined and transfored for upload.',\n 'auth': 'user',\n 'dataset': {'field': {'name': 'dataset','kind': 'string','order': 1,'description': 'Place where tables will be created in BigQuery.'}}\n }\n },\n {\n 'dbm': {\n 'auth': 'user',\n 'report': {\n 'timeout': 90,\n 'filters': {\n 'FILTER_ADVERTISER': {\n 'values': {'field': {'name': 'filter_ids','kind': 'integer_list','order': 7,'default': '','description': 'The comma separated list of Advertiser Ids.'}}\n }\n },\n 'body': {\n 'timezoneCode': {'field': {'name': 'recipe_timezone','kind': 'timezone','description': 'Timezone for report dates.','default': 'America/Los_Angeles'}},\n 'metadata': {\n 'title': {'field': {'name': 'recipe_name','kind': 'string','prefix': 'Monthly_Budget_Mover_','order': 1,'description': 'Name of report in DV360, should be unique.'}},\n 'dataRange': 'PREVIOUS_MONTH',\n 'format': 'CSV'\n },\n 'params': {\n 'type': 'TYPE_GENERAL',\n 'groupBys': [\n 'FILTER_ADVERTISER_CURRENCY',\n 'FILTER_INSERTION_ORDER'\n ],\n 'metrics': [\n 'METRIC_REVENUE_ADVERTISER'\n ]\n }\n }\n },\n 'delete': False\n }\n },\n {\n 'monthly_budget_mover': {\n 'auth': 'user',\n 'is_colab': {'field': {'name': 'is_colab','kind': 'boolean','default': True,'order': 7,'description': 'Are you running this in Colab? (This will store the files in Colab instead of Bigquery)'}},\n 'report_name': {'field': {'name': 'recipe_name','kind': 'string','prefix': 'Monthly_Budget_Mover_','order': 1,'description': 'Name of report in DV360, should be unique.'}},\n 'budget_categories': {'field': {'name': 'budget_categories','kind': 'json','order': 3,'default': '{}','description': 'A dictionary to show which IO Ids go under which Category. {\"CATEGORY1\":[12345,12345,12345], \"CATEGORY2\":[12345,12345]}'}},\n 'excluded_ios': {'field': {'name': 'excluded_ios','kind': 'integer_list','order': 4,'description': 'A comma separated list of Inserion Order Ids that should be exluded from the budget calculations'}},\n 'sdf': {\n 'auth': 'user',\n 'version': {'field': {'name': 'version','kind': 'choice','order': 6,'default': '5','description': 'The sdf version to be returned.','choices': ['SDF_VERSION_5','SDF_VERSION_5_1']}},\n 'partner_id': {'field': {'name': 'partner_id','kind': 'integer','order': 1,'description': 'The sdf file types.'}},\n 'file_types': 'INSERTION_ORDER',\n 'filter_type': 'FILTER_TYPE_ADVERTISER_ID',\n 'read': {\n 'filter_ids': {\n 'single_cell': True,\n 'values': {'field': {'name': 'filter_ids','kind': 'integer_list','order': 4,'default': [],'description': 'Comma separated list of filter ids for the request.'}}\n }\n },\n 'time_partitioned_table': False,\n 'create_single_day_table': False,\n 'dataset': {'field': {'name': 'dataset','kind': 'string','order': 6,'default': '','description': 'Dataset to be written to in BigQuery.'}},\n 'table_suffix': ''\n },\n 'out_old_sdf': {\n 'bigquery': {\n 'dataset': {'field': {'name': 'dataset','kind': 'string','order': 8,'default': '','description': 'Dataset that you would like your output tables to be produced in.'}},\n 'table': {'field': {'name': 'recipe_name','kind': 'string','prefix': 'SDF_OLD_','description': ''}},\n 'schema': [\n ],\n 'skip_rows': 0,\n 'disposition': 'WRITE_TRUNCATE'\n },\n 'file': '/content/old_sdf.csv'\n },\n 'out_new_sdf': {\n 'bigquery': {\n 'dataset': {'field': {'name': 'dataset','kind': 'string','order': 8,'default': '','description': 'Dataset that you would like your output tables to be produced in.'}},\n 'table': {'field': {'name': 'recipe_name','kind': 'string','prefix': 'SDF_NEW_','description': ''}},\n 'schema': [\n ],\n 'skip_rows': 0,\n 'disposition': 'WRITE_TRUNCATE'\n },\n 'file': '/content/new_sdf.csv'\n },\n 'out_changes': {\n 'bigquery': {\n 'dataset': {'field': {'name': 'dataset','kind': 'string','order': 8,'default': '','description': 'Dataset that you would like your output tables to be produced in.'}},\n 'table': {'field': {'name': 'recipe_name','kind': 'string','prefix': 'SDF_BUDGET_MOVER_LOG_','description': ''}},\n 'schema': [\n ],\n 'skip_rows': 0,\n 'disposition': 'WRITE_TRUNCATE'\n },\n 'file': '/content/log.csv'\n }\n }\n }\n]\n\njson_set_fields(TASKS, FIELDS)\n\nproject.initialize(_recipe={ 'tasks':TASKS }, _project=CLOUD_PROJECT, _user=USER_CREDENTIALS, _client=CLIENT_CREDENTIALS, _verbose=True, _force=True)\nproject.execute(_force=True)\n", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ] ]

ec7e20f901d1c12d57dbd95a15dd9e84c335f03a

917,916

ipynb

Jupyter Notebook

notebooks/benchmarks/multiple_example/002_p-gnn2D_BCE_loss_variable_embed_batch_norm_sum_combo_benchmark.ipynb

FordyceLab/tessellate

a3f0c38b4392027a7503828f48d65ec02eb24698

[ "MIT" ]

null

notebooks/benchmarks/multiple_example/002_p-gnn2D_BCE_loss_variable_embed_batch_norm_sum_combo_benchmark.ipynb

FordyceLab/tessellate

a3f0c38b4392027a7503828f48d65ec02eb24698

[ "MIT" ]

7

2021-03-31T19:41:46.000Z

2022-01-13T02:39:45.000Z

notebooks/benchmarks/multiple_example/002_p-gnn2D_BCE_loss_variable_embed_batch_norm_sum_combo_benchmark.ipynb

FordyceLab/tessellate

a3f0c38b4392027a7503828f48d65ec02eb24698

[ "MIT" ]

null

105.253526

1,785

0.552753

[ [ [ "# Benchmark position-aware graph neural network/2D CNN architecture\n\nThis notebook contains all of the code to overfit a P-GNN/2D CNN to four contact channels of a single structure (6E6O).\n\n## Setup", "_____no_output_____" ], [ "### Dataloader code ", "_____no_output_____" ] ], [ [ "import torch\nfrom torch.utils.data import Dataset\nimport numpy as np\nimport periodictable as pt\nimport pandas as pd\nimport os\nimport matplotlib.pyplot as plt\nimport networkx as nx\nimport itertools as it\n\n\ndef read_files(acc, model, graph_dir, contacts_dir):\n \"\"\"\n Read graph and contacts files.\n\n Args:\n - acc (str) - String of the PDB ID (lowercese).\n - model (int) - Model number of the desired bioassembly.\n - graph_dir (str) - Directory containing the nodes, edges,\n and mask files.\n - contacts_dir (str) - Directory containing the .contacts\n files from get_contacts.py.\n\n Returns:\n - Dictionary of DataFrames and lists corresponding to\n graph nodes, edges, mask, and contacts.\n \"\"\"\n\n # Get the file names for the graph files\n node_file = os.path.join(graph_dir, '{}-{}_nodes.csv'.format(acc, model))\n edge_file = os.path.join(graph_dir, '{}-{}_edges.csv'.format(acc, model))\n mask_file = os.path.join(graph_dir, '{}-{}_mask.csv'.format(acc, model))\n\n # Get the contacts file\n contacts_file = os.path.join(contacts_dir, '{}-{}.contacts'.format(acc, model))\n\n # Read the nodes and edges\n nodes = pd.read_csv(node_file)\n edges = pd.read_csv(edge_file)\n\n # Check if the mask is empty\n if os.path.getsize(mask_file) > 0:\n with open(mask_file) as f:\n mask = f.read().split('\\n')\n else:\n mask = []\n\n # Read the contacts\n contacts = pd.read_table(contacts_file, sep='\\t',\n header=None, names=['type', 'start', 'end'])\n\n # Return the data\n data = {\n 'nodes': nodes,\n 'edges': edges,\n 'mask': mask,\n 'contacts': contacts\n }\n\n return data\n\n\ndef process_res_data(data):\n \"\"\"\n Process residue-level data from atom-level data.\n\n Args:\n - data (dict) - Dictionary of graph data output from `read_files`.\n\n Returns:\n - Dictionary of atom and residue graph and contact data.\n \"\"\"\n\n # Extract data form dict\n nodes = data['nodes']\n edges = data['edges']\n mask = data['mask']\n contacts = data['contacts']\n\n # Get residue nodes\n res_nodes = pd.DataFrame()\n res_nodes['res'] = [':'.join(atom.split(':')[:3]) for atom in nodes['atom']]\n res_nodes = res_nodes.drop_duplicates().reset_index(drop=True)\n\n # Get residue edges\n res_edges = edges.copy()\n res_edges['start'] = [':'.join(atom.split(':')[:3]) for atom in res_edges['start']]\n res_edges['end'] = [':'.join(atom.split(':')[:3]) for atom in res_edges['end']]\n res_edges = res_edges[res_edges['start'] != res_edges['end']].drop_duplicates().reset_index(drop=True)\n\n # Get residue contacts\n res_contacts = contacts.copy()\n res_contacts['start'] = [':'.join(atom.split(':')[:3]) for atom in res_contacts['start']]\n res_contacts['end'] = [':'.join(atom.split(':')[:3]) for atom in res_contacts['end']]\n res_contacts = res_contacts[res_contacts['start'] != res_contacts['end']].drop_duplicates().reset_index(drop=True)\n\n # Get residue mask\n res_mask = list(set([':'.join(atom.split(':')[:3]) for atom in mask]))\n\n # Return data dict\n data = {\n 'atom_nodes': nodes,\n 'atom_edges': edges,\n 'atom_contact': contacts,\n 'atom_mask': mask,\n 'res_nodes': res_nodes,\n 'res_edges': res_edges,\n 'res_contact': res_contacts,\n 'res_mask': res_mask\n }\n\n return data\n\n\ndef get_map_dicts(entity_list):\n \"\"\"\n Map identifiers to indices and vice versa.\n\n Args:\n - entity_list (list) - List of entities (atoms, residues, etc.)\n to index.\n\n Returns:\n - Tuple of the entity to index and index to entity dicts, respectively.\n \"\"\"\n\n # Create the entity:index dictionary\n ent2idx_dict = {entity: idx for idx, entity in enumerate(entity_list)}\n\n # Create the index:entity dictionary\n idx2ent_dict = {idx: entity for entity, idx in ent2idx_dict.items()}\n\n return (ent2idx_dict, idx2ent_dict)\n\n\ndef create_adj_mat(data, dict_map, mat_type):\n \"\"\"\n Creates an adjacency matrix.\n\n Args:\n - data (DataFrame) - Dataframe with 'start' and 'end' column\n for each interaction. For atom-level adjacency, 'order'\n column is also required. For atom or residue conatcts,\n 'type' column is also required.\n\n Returns:\n - Coordinate format matrix (numpy). For atom adjacency, third column\n corresponds to bond order. For contacts, third column\n corresponds to channel.\n\n Channel mappings (shorthand from get_contacts.py source):\n\n 0:\n hp hydrophobic interactions\n 1:\n hb hydrogen bonds\n lhb ligand hydrogen bonds\n hbbb backbone-backbone hydrogen bonds\n hbsb backbone-sidechain hydrogen bonds\n hbss sidechain-sidechain hydrogen bonds\n hbls ligand-sidechain residue hydrogen bonds\n hblb ligand-backbone residue hydrogen bonds\n 2:\n vdw van der Waals\n 3:\n wb water bridges\n wb2 extended water bridges\n lwb ligand water bridges\n lwb2 extended ligand water bridges\n 4:\n sb salt bridges\n 5:\n ps pi-stacking\n 6:\n pc pi-cation\n 7:\n ts t-stacking\n \"\"\"\n\n # Initialize the coordinate list\n coord_mat = []\n\n # Map channel names to numeric channels\n channel = {\n # Hydrophobic interactions in first channel\n 'hp': 0,\n 'hplp': 0,\n 'hpll': 0,\n\n # Hydrogen bonds in second channel\n 'hb': 1,\n 'lhb': 1,\n 'hbbb': 1,\n 'hbsb': 1,\n 'hbss': 1,\n 'hbls': 1,\n 'hblb': 1,\n\n # VdW in third channel\n 'vdw': 2,\n\n # Water bridges\n 'wb': 3,\n 'wb2': 3,\n 'lwb': 3,\n 'lwb2': 3,\n\n # Salt bridges\n 'sb': 4,\n 'sbpl': 4,\n 'sbll': 4,\n\n # Other interactions\n 'ps': 5,\n 'pc': 6,\n 'ts': 7,\n }\n\n # Assemble the contacts\n for idx, row in data.iterrows():\n\n if row['start'] in dict_map and row['end'] in dict_map:\n\n entry = [dict_map[row['start']], dict_map[row['end']]]\n\n # Add order or type if necessary\n if mat_type == 'atom_graph':\n entry.append(row['order'])\n elif mat_type == 'atom_contact':\n entry.append(channel[row['type']])\n elif mat_type == 'res_contact':\n entry.append(channel[row['type']])\n\n coord_mat.append(entry)\n\n return np.array(coord_mat)\n\n\ndef create_conn_adj_mat(adj):\n \"\"\"\n Create connection adjacency matrix\n \"\"\"\n\n conn_map = {(a, b): idx for idx, (a, b) in enumerate(zip(*np.triu_indices_from(adj)))}\n\n one_hop_neighbors = {idx: np.argwhere(row > 0).squeeze().tolist() for idx, row in enumerate(adj)}\n\n conns = []\n\n for i, j in it.combinations_with_replacement(one_hop_neighbors, 2):\n row = conn_map[(i, j)]\n adj_conn_coords = set([(m, n) if m <= n else (n, m)\n for m, n in it.product(one_hop_neighbors[i],\n one_hop_neighbors[j])])\n adj_conns = [conn_map[x] for x in adj_conn_coords]\n\n conns.append(np.array(list(it.product([row], adj_conns))))\n\n conn_adj = np.concatenate(conns, axis=0).T\n\n return conn_adj\n\n\ndef create_mem_mat(atom_dict, res_dict):\n \"\"\"\n Create a membership matrix mapping atoms to residues.\n\n Args:\n - atom_dict (dict) - Dictionary mapping atoms to indices.\n - res_dict (dict) - Dictionary mapping residues to indices.\n\n Returns:\n - Coordinate format membership matrix (numpy) with first\n row being residue number and the second column being\n atom number.\n \"\"\"\n\n # Initialize the coordinate list\n mem_coord = []\n\n # Map atoms to residues\n for atom, atom_idx in atom_dict.items():\n res_idx = res_dict[':'.join(atom.split(':')[:3])]\n\n mem_coord.append([res_idx, atom_idx])\n\n mem_coord = np.array(mem_coord)\n\n return mem_coord\n\n\ndef create_idx_list(id_list, dict_map):\n \"\"\"\n Create list of indices.\n\n Args:\n - id_list (list) - List of masked atom or residue identifiers.\n - dict_map (dict) - Dictionary mapping entities to indices.\n\n Returns:\n - A numpy array of the masked indices.\n \"\"\"\n\n # Generate the numpy index array\n idx_array = np.array([dict_map[iden] for iden in id_list])\n\n return idx_array\n\n\nclass TesselateDataset(Dataset):\n \"\"\"\n Dataset class for structural data.\n\n Args:\n - accession_list (str) - File path from which to read PDB IDs for dataset.\n - graph_dir (str) - Directory containing the nodes, edges, and mask files.\n - contacts_dir (str) - Directory containing the .contacts files from\n get_contacts.py.\n - return_data (list) - List of datasets to return. Value must be 'all' or\n a subset of the following list:\n - pdb_id\n - model\n - atom_nodes\n - atom_adj\n - atom_contact\n - atom_mask\n - res_adj\n - res_dist\n - res_contact\n - res_mask\n - mem_mat\n - idx2atom_dict\n - idx2res_dict\n \"\"\"\n\n def __init__(self, accession_list, graph_dir, contacts_dir, add_covalent=False, return_data='all', in_memory=False):\n\n if return_data == 'all':\n self.return_data = [\n 'pdb_id',\n 'model',\n 'atom_nodes',\n 'atom_adj',\n 'atom_contact',\n 'atom_mask',\n 'res_adj',\n 'res_dist',\n 'conn_adj',\n 'res_contact',\n 'res_mask',\n 'mem_mat',\n 'idx2atom_dict',\n 'idx2res_dict'\n ]\n\n else:\n self.return_data = return_data\n\n # Store reference to accession list file\n self.accession_list = accession_list\n\n # Store references to the necessary directories\n self.graph_dir = graph_dir\n self.contacts_dir = contacts_dir\n\n # Whether to add covalent bonds to prediction task and\n # remove sequence non-deterministic covalent bonds from the adjacency matrix\n self.add_covalent=add_covalent\n\n # Read in and store a list of accession IDs\n with open(accession_list, 'r') as handle:\n self.accessions = np.array([acc.strip().lower().split() for acc in handle.readlines()])\n\n self.data = {}\n\n\n def __len__(self):\n \"\"\"\n Return the length of the dataset.\n\n Returns:\n - Integer count of number of examples.\n \"\"\"\n return len(self.accessions)\n\n\n def __getitem__(self, idx):\n \"\"\"\n Get an item with a particular index value.\n\n Args:\n - idx (int) - Index of desired sample.\n\n Returns:\n - Dictionary of dataset example. All tensors are sparse when possible.\n \"\"\"\n\n try:\n if idx in self.data:\n return self.data[idx]\n\n # initialize the return dictionary\n return_dict = {}\n\n acc_entry = self.accessions[idx]\n\n # Get the PDB ID\n acc = acc_entry[0]\n\n # Get the model number if one exists\n if len(acc_entry) == 1:\n model = 1\n else:\n model = acc_entry[1]\n\n # Read and process the files\n data = read_files(acc, model, self.graph_dir, self.contacts_dir)\n data = process_res_data(data)\n\n # Generate the mapping dictionaries\n atom2idx_dict, idx2atom_dict = get_map_dicts(data['atom_nodes']['atom'].unique())\n res2idx_dict, idx2res_dict = get_map_dicts(data['res_nodes']['res'].unique())\n\n # Get numbers of atoms and residues per sample\n n_atoms = len(atom2idx_dict)\n n_res = len(res2idx_dict)\n\n # Handle all of the possible returned datasets\n if 'pdb_id' in self.return_data:\n return_dict['pdb_id'] = acc\n\n if 'model' in self.return_data:\n return_dict['model'] = model\n\n if 'atom_nodes' in self.return_data:\n ele_nums = [pt.elements.symbol(element).number for element in data['atom_nodes']['element']]\n return_dict['atom_nodes'] = torch.LongTensor(ele_nums)\n assert not torch.isnan(return_dict['atom_nodes']).any()\n\n if 'atom_adj' in self.return_data:\n\n adj = create_adj_mat(data['atom_edges'], atom2idx_dict, mat_type='atom_graph').T\n\n x = torch.LongTensor(adj[0, :]).squeeze()\n y = torch.LongTensor(adj[1, :]).squeeze()\n val = torch.FloatTensor(adj[2, :]).squeeze()\n\n atom_adj = torch.zeros([n_atoms, n_atoms]).index_put_((x, y), val, accumulate=False)\n\n atom_adj = atom_adj.index_put_((y, x), val, accumulate=False)\n\n atom_adj[range(n_atoms), range(n_atoms)] = 1\n\n atom_adj = (atom_adj > 0).float()\n\n return_dict['atom_adj'] = atom_adj\n\n assert not torch.isnan(return_dict['atom_adj']).any()\n\n if 'atom_contact' in self.return_data:\n atom_contact = create_adj_mat(data['atom_contact'], atom2idx_dict, mat_type='atom_contact').T\n\n x = torch.LongTensor(atom_contact[0, :]).squeeze()\n y = torch.LongTensor(atom_contact[1, :]).squeeze()\n z = torch.LongTensor(atom_contact[2, :]).squeeze()\n\n atom_contact = torch.zeros([n_atoms, n_atoms, 8]).index_put_((x, y, z),\n torch.ones(len(x)))\n atom_contact = atom_contact.index_put_((y, x, z),\n torch.ones(len(x)))\n\n return_dict['atom_contact'] = atom_contact\n\n assert not torch.isnan(return_dict['atom_contact']).any()\n\n if 'atom_mask' in self.return_data:\n atom_mask = create_idx_list(data['atom_mask'], atom2idx_dict)\n\n masked_pos = torch.from_numpy(atom_mask)\n\n if self.add_covalent:\n channels = 9\n else:\n channels = 8\n\n mask = torch.ones([n_atoms, n_atoms, channels])\n\n if len(masked_pos) > 0:\n mask[masked_pos, :, :] = 0\n mask[:, masked_pos, :] = 0\n\n return_dict['atom_mask'] = mask\n\n assert not torch.isnan(return_dict['atom_mask']).any()\n\n if 'res_adj' in self.return_data:\n adj = create_adj_mat(data['res_edges'], res2idx_dict, mat_type='res_graph').T\n\n x = torch.LongTensor(adj[0, :]).squeeze()\n y = torch.LongTensor(adj[1, :]).squeeze()\n\n res_adj = torch.zeros([n_res, n_res]).index_put_((x, y), torch.ones(len(x)))\n\n res_adj = res_adj.index_put_((y, x), torch.ones(len(x)))\n\n res_adj[range(n_res), range(n_res)] = 1\n\n return_dict['res_adj'] = res_adj\n\n assert not torch.isnan(return_dict['res_adj']).any()\n\n if 'res_dist' in self.return_data:\n G = nx.from_numpy_matrix(return_dict['res_adj'].numpy())\n res_dist = torch.from_numpy(nx.floyd_warshall_numpy(G)).float()\n\n res_dist[torch.isinf(res_dist)] = -1\n\n return_dict['res_dist'] = res_dist\n\n chain_mem = torch.zeros(res_dist.shape)\n chain_mem[~torch.isinf(return_dict['res_dist'])] = 1\n\n return_dict['chain_mem'] = chain_mem\n\n assert not torch.isnan(return_dict['res_dist']).any()\n assert not torch.isinf(return_dict['res_dist']).any()\n assert not torch.isnan(return_dict['chain_mem']).any()\n assert not torch.isinf(return_dict['chain_mem']).any()\n\n if 'conn_adj' in self.return_data:\n\n conn_adj = create_conn_adj_mat(return_dict['res_adj'].numpy())\n\n return_dict['conn_adj'] = torch.from_numpy(conn_adj)\n\n if 'res_contact' in self.return_data:\n res_contact = create_adj_mat(data['res_contact'], res2idx_dict, mat_type='res_contact').T\n\n x = torch.LongTensor(res_contact[0, :]).squeeze()\n y = torch.LongTensor(res_contact[1, :]).squeeze()\n z = torch.LongTensor(res_contact[2, :]).squeeze()\n\n res_contact = torch.zeros([n_res, n_res, 8]).index_put_((x, y, z),\n torch.ones(len(x)))\n\n res_contact = res_contact.index_put_((y, x, z),\n torch.ones(len(x)))\n\n return_dict['res_contact'] = res_contact\n\n assert not torch.isnan(return_dict['res_contact']).any()\n\n if 'res_mask' in self.return_data:\n res_mask = create_idx_list(data['res_mask'], res2idx_dict)\n\n masked_pos = torch.from_numpy(res_mask)\n\n if self.add_covalent:\n channels = 9\n else:\n channels = 8\n\n mask = torch.ones([n_res, n_res, channels])\n\n if len(masked_pos) > 0:\n mask[masked_pos, :, :] = 0\n mask[:, masked_pos, :] = 0\n\n return_dict['res_mask'] = mask\n\n assert not torch.isnan(return_dict['res_mask']).any()\n\n if 'mem_mat' in self.return_data:\n mem_mat = create_mem_mat(atom2idx_dict, res2idx_dict).T\n\n x = torch.LongTensor(mem_mat[0, :]).squeeze()\n y = torch.LongTensor(mem_mat[1, :]).squeeze()\n\n mem_mat = torch.zeros([n_res, n_atoms]).index_put_((x, y),\n torch.ones(len(x)))\n\n return_dict['mem_mat'] = mem_mat\n\n assert not torch.isnan(return_dict['mem_mat']).any()\n\n if 'idx2atom_dict' in self.return_data:\n return_dict['idx2atom_dict'] = idx2atom_dict\n\n if 'idx2res_dict' in self.return_data:\n return_dict['idx2res_dict'] = idx2res_dict\n\n self.data[idx] = return_dict\n\n # Return the processed data\n return return_dict\n\n except Exception as e:\n print(\"Error:\", acc, str(e))\n return np.array([0])", "_____no_output_____" ] ], [ [ "### Modules", "_____no_output_____" ] ], [ [ "import torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nimport torch.optim as optim\nimport numpy as np\n\n\n####################\n# Embedding layers #\n####################\n\nclass AtomEmbed(nn.Module):\n \"\"\"\n Embed the atoms to fixed-length input vectors.\n\n Args:\n - num_features (int) - Size of the returned embedding vectors.\n - scale_grad_by_freq (bool) - Scale gradients by the inverse of\n frequency (default=True).\n \"\"\"\n\n def __init__(self, n_features, scale_grad_by_freq=True):\n super(AtomEmbed, self).__init__()\n self.embedding = nn.Embedding(118,\n n_features,\n scale_grad_by_freq=scale_grad_by_freq)\n\n def forward(self, atomic_numbers):\n \"\"\"\n Return the embeddings for each atom in the graph.\n\n Args:\n - atoms (torch.LongTensor) - Tensor (n_atoms) containing atomic numbers.\n\n Returns:\n - torch.FloatTensor of dimension (n_atoms, n_features) containing\n the embedding vectors.\n \"\"\"\n\n # Get and return the embeddings for each atom\n embedded_atoms = self.embedding(atomic_numbers)\n return embedded_atoms\n\n\n#########################\n# Position-aware layers #\n#########################\n\n# # PGNN layer, only pick closest node for message passing\nclass PGNN_layer(nn.Module):\n def __init__(self, input_dim, output_dim,dist_trainable=False):\n super(PGNN_layer, self).__init__()\n self.input_dim = input_dim\n self.dist_trainable = dist_trainable\n\n if self.dist_trainable:\n self.dist_compute = Nonlinear(1, output_dim, 1)\n\n self.linear_hidden = nn.Linear(input_dim*2, output_dim)\n self.linear_out_position = nn.Linear(output_dim,1)\n self.act = nn.ReLU()\n\n for m in self.modules():\n if isinstance(m, nn.Linear):\n m.weight.data = nn.init.xavier_uniform_(m.weight.data, gain=nn.init.calculate_gain('relu'))\n if m.bias is not None:\n m.bias.data = nn.init.constant_(m.bias.data, 0.0)\n\n def forward(self, feature, dists_max, dists_argmax):\n if self.dist_trainable:\n dists_max = self.dist_compute(dists_max.unsqueeze(-1)).squeeze()\n\n subset_features = feature[dists_argmax.flatten(), :]\n subset_features = subset_features.reshape((dists_argmax.shape[0], dists_argmax.shape[1],\n feature.shape[1]))\n\n messages = (subset_features * dists_max.unsqueeze(-1))\n\n self_feature = feature.unsqueeze(1).repeat(1, dists_max.shape[1], 1)\n messages = torch.cat((messages, self_feature), dim=-1)\n\n messages = self.linear_hidden(messages).squeeze()\n messages = self.act(messages) # n*m*d\n\n out_position = self.linear_out_position(messages).squeeze(-1) # n*m_out\n out_structure = torch.mean(messages, dim=1) # n*d\n\n return out_position, out_structure\n\n\n### Non linearity\nclass Nonlinear(nn.Module):\n def __init__(self, input_dim, hidden_dim, output_dim):\n super(Nonlinear, self).__init__()\n\n self.linear1 = nn.Linear(input_dim, hidden_dim)\n self.linear2 = nn.Linear(hidden_dim, output_dim)\n\n self.act = nn.ReLU()\n\n for m in self.modules():\n if isinstance(m, nn.Linear):\n m.weight.data = nn.init.xavier_uniform_(m.weight.data, gain=nn.init.calculate_gain('relu'))\n if m.bias is not None:\n m.bias.data = nn.init.constant_(m.bias.data, 0.0)\n\n def forward(self, x):\n x = self.linear1(x)\n x = self.act(x)\n x = self.linear2(x)\n return x", "_____no_output_____" ] ], [ [ "# Functions", "_____no_output_____" ] ], [ [ "import numpy as np\nimport torch\n\n########################################\n# Pairwise matrix generation functions #\n########################################\n\ndef pairwise_mat(nodes, method='mean'):\n \"\"\"\n Generate matrix for pairwise determination of interactions.\n\n Args:\n - nodes (torch.FloatTensor) - Tensor of node (n_nodes, n_features) features.\n - method (str) - One of 'sum' or 'mean' for combination startegy for\n pairwise combination matrix (default = 'mean').\n\n Returns:\n - torch.FloatTensor of shape (n_pairwise, n_nodes) than can be used used to\n combine feature vectors. Values are 1 if method == \"sum\" and 0.5 if\n method == \"mean\".\n \"\"\"\n\n # Get the upper triangle indices\n triu = np.vstack(np.triu_indices(nodes.shape[0]))\n\n # Loop through all indices and add to list with\n idxs = torch.from_numpy(triu).T\n\n # Convert to tensor\n combos = torch.zeros([idxs.shape[0], nodes.shape[0]]).scatter(1, idxs, 1)\n\n # Set to 0.5 if method is 'mean'\n if method == 'mean':\n combos *= 0.5\n\n return combos.to(nodes.device)\n\n\ndef pairwise_3d(nodes):\n # Get the upper triangle indices\n repeated_nodes = nodes.unsqueeze(0).expand(nodes.shape[0], -1, -1)\n repeated_nodes2 = repeated_nodes.permute(1, 0, 2)\n\n return torch.cat((repeated_nodes, repeated_nodes2), dim=-1)\n\n\n############################\n# Upper triangle functions #\n############################\n\ndef triu_condense(input_tensor):\n \"\"\"\n Condense the upper triangle of a tensor into a 2d dense representation.\n\n Args:\n - input_tensor (torch.Tensor) - Tensor of shape (n, n, m).\n\n Returns:\n - Tensor of shape (n(n+1)/2, m) where elements along the third dimension in\n the original tensor are packed row-wise according to the upper\n triangular indices.\n \"\"\"\n\n # Get upper triangle index info\n row_idx, col_idx = np.triu_indices(input_tensor.shape[0])\n row_idx = torch.LongTensor(row_idx)\n col_idx = torch.LongTensor(col_idx)\n\n # Return the packed matrix\n output = input_tensor[row_idx, col_idx, :]\n\n return output\n\n\ndef triu_expand(input_matrix):\n \"\"\"\n Expand a dense representation of the upper triangle of a tensor into\n a 3D squareform representation.\n\n Args:\n - input_matrix (torch.Tensor) - Tensor of shape (n(n+1)/2, m).\n\n Returns:\n - Tensor of shape (n, n, m) where elements along the third dimension in the\n original tensor are packed row-wise according to the upper triangular\n indices.\n \"\"\"\n # Get the edge size n of the tensor\n n_elements = input_matrix.shape[0]\n n_chan = input_matrix.shape[1]\n n_res = int((-1 + np.sqrt(1 + 4 * 2 * (n_elements))) / 2)\n\n # Get upper triangle index info\n row_idx, col_idx = np.triu_indices(n_res)\n row_idx = torch.LongTensor(row_idx)\n col_idx = torch.LongTensor(col_idx)\n\n # Generate the output tensor\n output = torch.zeros((n_res, n_res, n_chan), device=input_matrix.device)\n\n # Input the triu values\n for i in range(n_chan):\n i_tens = torch.full((len(row_idx),), i, dtype=torch.long)\n output.index_put_((row_idx, col_idx, i_tens), input_matrix[:, i])\n\n # Input the tril values\n for i in range(n_chan):\n i_tens = torch.full((len(row_idx),), i, dtype=torch.long)\n output.index_put_((col_idx, row_idx, i_tens), input_matrix[:, i])\n\n return output\n\n\n###################\n# P-GNN functions #\n###################\n\ndef generate_dists(adj_mat):\n adj_mask = adj_mat == 0\n\n dist = adj_mat - torch.eye(adj_mat.shape[0], device=adj_mat.device)\n dist = 1 / (dist + 1)\n dist[adj_mask] = 0\n\n return dist.squeeze()\n\n\ndef get_dist_max(anchorset_id, dist):\n dist_max = torch.zeros((dist.shape[0],len(anchorset_id)),\n device=dist.device)\n dist_argmax = torch.zeros((dist.shape[0],len(anchorset_id)),\n device=dist.device).long()\n for i in range(len(anchorset_id)):\n temp_id = anchorset_id[i]\n dist_temp = dist[:, temp_id]\n dist_max_temp, dist_argmax_temp = torch.max(dist_temp, dim=-1)\n dist_max[:,i] = dist_max_temp\n dist_argmax[:,i] = dist_argmax_temp\n return dist_max, dist_argmax\n\n\ndef get_random_anchorset(n,c=0.5):\n m = int(np.log2(n))\n copy = int(c*m)\n anchorset_id = []\n for i in range(m):\n anchor_size = int(n/np.exp2(i + 1))\n for j in range(copy):\n anchorset_id.append(np.random.choice(n,size=anchor_size,replace=False))\n return anchorset_id\n\n\ndef preselect_anchor(n_nodes, dists):\n anchorset_id = get_random_anchorset(n_nodes, c=1)\n return get_dist_max(anchorset_id, dists)", "_____no_output_____" ] ], [ [ "### Model", "_____no_output_____" ] ], [ [ "import torch\nimport torch.nn as nn\nfrom torch.nn import functional as F\nimport pytorch_lightning as pl\nfrom torch.utils.data import DataLoader\nimport wandb\n\n\nclass PGNN2D(pl.LightningModule):\n def __init__(self, input_dim, hidden_dim, output_dim,\n n_contact_channels,\n layer_num=2, train_data=None,\n val_data=None, test_data=None):\n super(PGNN2D, self).__init__()\n\n # Parameters\n self.n_contact_channels = n_contact_channels\n self.layer_num = layer_num\n\n # Datasets\n self.train_data = train_data\n self.val_data = val_data\n self.test_data = test_data\n\n # Embedding\n self.embed = AtomEmbed(input_dim)\n\n # First P-GNN layer\n self.conv_atom_first = PGNN_layer(input_dim, hidden_dim)\n\n self.batch_norm_first = nn.BatchNorm1d(hidden_dim)\n\n # All other P-GNN layers\n if layer_num>1:\n self.conv_atom_hidden = nn.ModuleList([PGNN_layer(hidden_dim,\n hidden_dim)\n for i in range(layer_num - 1)])\n\n self.batch_norm_hidden = nn.ModuleList([nn.BatchNorm1d(hidden_dim)\n for i in range(layer_num - 1)])\n\n # RNN to condense position-aware embeddings\n self.embed_condense = torch.nn.GRU(input_size=1, hidden_size=25, batch_first=True)\n self.batch_norm_condense = nn.BatchNorm1d(25)\n\n self.batch_norm_pairwise = nn.BatchNorm1d(25)\n\n # 2D convolutional layers\n self.conv1 = nn.Conv2d(25,\n 25, 3, stride=1, padding=1)\n self.batch_norm_conv1 = nn.BatchNorm2d(25)\n self.conv2 = nn.Conv2d(25,\n 25, 3, stride=1, padding=1)\n self.batch_norm_conv2 = nn.BatchNorm2d(25)\n\n self.conv3 = nn.Conv2d(25,\n 25, 3, stride=1, padding=1)\n self.batch_norm_conv3 = nn.BatchNorm2d(25)\n\n # Final 2D CNN\n self.conv_final = nn.Conv2d(25, n_contact_channels, 3, stride=1, padding=1)\n\n # Focal loss\n self.loss = nn.BCEWithLogitsLoss()\n\n # Validation information\n self.reset_epoch_metrics()\n\n \n def forward(self, data):\n\n data['atom_dist'] = generate_dists(data['atom_adj'])\n\n data['atom_dist_max'], data['atom_dist_argmax'] = preselect_anchor(data['atom_adj'].squeeze().shape[0], data['atom_dist'])\n\n x = self.embed(data['atom_nodes'].squeeze())\n\n atom_embed = x.detach().cpu().numpy()\n\n x_position, x = self.conv_atom_first(x, data['atom_dist_max'], data['atom_dist_argmax'])\n x = F.relu(x)\n x = self.batch_norm_first(x)\n\n for i in range(self.layer_num-1):\n\n data['atom_dist'] = generate_dists(data['atom_adj'])\n\n data['atom_dist_max'], data['atom_dist_argmax'] = preselect_anchor(data['atom_adj'].squeeze().shape[0], data['atom_dist'])\n\n x_position, x = self.conv_atom_hidden[i](x, data['atom_dist_max'], data['atom_dist_argmax'])\n x = self.batch_norm_hidden[i](x)\n x = F.relu(x)\n\n x_position = data['mem_mat'].squeeze().matmul(x_position)\n\n _, x_position = self.embed_condense(x_position.unsqueeze(-1))\n\n x_position = x_position.squeeze()\n x_position = self.batch_norm_condense(x_position)\n\n pairwise = pairwise_mat(x_position, method='sum').matmul(x_position)\n pairwise = triu_expand(self.batch_norm_pairwise(pairwise)).permute(2, 0, 1).unsqueeze(0)\n\n conv1_out = F.relu(self.conv1(pairwise))\n conv1_out = self.batch_norm_conv1(conv1_out)\n conv2_out = F.relu(self.conv2(conv1_out))\n conv2_out = self.batch_norm_conv2(conv2_out)\n conv3_out = F.relu(self.conv3(conv2_out))\n conv3_out = self.batch_norm_conv3(conv3_out)\n square_preds = self.conv_final(conv3_out)\n\n square_preds = square_preds + square_preds.permute(0, 1, 3, 2)\n\n preds = triu_condense(square_preds.squeeze().permute(1, 2, 0))\n\n return preds\n\n def configure_optimizers(self):\n parameters = filter(lambda p: p.requires_grad, self.parameters())\n optimizer = torch.optim.SGD(parameters, lr=.1, momentum=0.9)\n scheduler = torch.optim.lr_scheduler.ReduceLROnPlateau(optimizer)\n\n return [optimizer], [scheduler]\n \n def training_step(self, batch, batch_nb):\n # REQUIRED\n y_hat = self.forward(batch).squeeze()\n y = triu_condense(batch['res_contact'].squeeze())\n weights = triu_condense(batch['res_mask'].squeeze())\n\n loss = F.binary_cross_entropy_with_logits(y_hat, y, weight=weights)\n\n self.train_losses.append(loss.item())\n\n return {'loss': loss}\n\n def validation_step(self, batch, batch_nb):\n # OPTIONAL\n y_hat = self.forward(batch).squeeze()\n y = triu_condense(batch['res_contact'].squeeze())\n weights = triu_condense(batch['res_mask'].squeeze())\n\n loss = F.binary_cross_entropy_with_logits(y_hat, y, weight=weights)\n\n self.val_losses.append(loss.item())\n\n _, _, auroc = calc_metric_curve(y_hat.sigmoid().cpu(), y.cpu(), 'ROC', squareform=False)\n _, _, auprc = calc_metric_curve(y_hat.sigmoid().cpu(), y.cpu(), 'PRC', squareform=False)\n\n self.auroc['Hydrophobic'].append(auroc[0])\n self.auroc['Hydrogen bond'].append(auroc[1])\n self.auroc['Van der Waals'].append(auroc[2])\n self.auroc['Water bridges'].append(auroc[3])\n self.auroc['Salt bridges'].append(auroc[4])\n self.auroc['Pi-stacking'].append(auroc[5])\n self.auroc['Pi-cation'].append(auroc[6])\n self.auroc['T-stacking'].append(auroc[7])\n\n self.auprc['Hydrophobic'].append(auprc[0])\n self.auprc['Hydrogen bond'].append(auprc[1])\n self.auprc['Van der Waals'].append(auprc[2])\n self.auprc['Water bridges'].append(auprc[3])\n self.auprc['Salt bridges'].append(auprc[4])\n self.auprc['Pi-stacking'].append(auprc[5])\n self.auprc['Pi-cation'].append(auprc[6])\n self.auprc['T-stacking'].append(auprc[7])\n \n self.pred_example.append(wandb.Image(plot_channels(triu_expand(y_hat.detach().cpu()))))\n \n def validation_end(self, outputs):\n avg_loss = torch.tensor(self.val_losses).mean().item()\n return {'val_loss': avg_loss}\n\n def on_post_performance_check(self):\n\n epoch_metrics = {}\n all_metrics = {}\n\n for key in self.auroc:\n values = np.asarray(self.auroc[key])\n all_metrics[f'AUROC - {key}'] = values\n\n values = values[~np.isnan(values)]\n epoch_metrics[f'AUROC - {key}'] = wandb.Histogram(values)\n\n\n for key in self.auprc:\n values = np.asarray(self.auprc[key])\n all_metrics[f'AUPRC - {key}'] = values\n\n values = values[~np.isnan(values)]\n epoch_metrics[f'AUPRC - {key}'] = wandb.Histogram(values)\n\n epoch_metrics['Train Losses'] = wandb.Histogram(self.train_losses)\n\n epoch_metrics['Val Losses'] = wandb.Histogram(self.val_losses)\n all_metrics['Val Losses'] = np.asarray(self.val_losses)\n\n all_metrics = pd.DataFrame(all_metrics)\n cor_mat = all_metrics.corr()\n\n corr_plot = metric_corr_plot(cor_mat)\n epoch_metrics['Metric Corr Plot'] = wandb.Image(corr_plot)\n\n epoch_metrics['train_loss'] = np.asarray(self.train_losses).mean()\n epoch_metrics['val_loss'] = np.asarray(self.val_losses).mean()\n\n epoch_metrics['Example Prediction'] = self.pred_example\n\n self.logger.experiment.log(epoch_metrics)\n\n self.reset_epoch_metrics()\n\n def reset_epoch_metrics(self):\n\n # Validation information\n self.auroc = {\n 'Hydrophobic': [],\n 'Hydrogen bond': [],\n 'Van der Waals': [],\n 'Water bridges': [],\n 'Salt bridges': [],\n 'Pi-stacking': [],\n 'Pi-cation': [],\n 'T-stacking': []\n }\n\n self.auprc = {\n 'Hydrophobic': [],\n 'Hydrogen bond': [],\n 'Van der Waals': [],\n 'Water bridges': [],\n 'Salt bridges': [],\n 'Pi-stacking': [],\n 'Pi-cation': [],\n 'T-stacking': []\n }\n\n self.val_losses = []\n self.train_losses = []\n self.pred_example = []\n\n\n# def validation_end(self, outputs):\n# # OPTIONAL\n# # avg_loss = torch.stack([x['val_loss'] for x in outputs]).mean()\n# # return {'avg_val_loss': avg_loss}\n# pass\n\n @pl.data_loader\n def train_dataloader(self):\n # REQUIRED\n return DataLoader(self.train_data, shuffle=True, num_workers=30, pin_memory=True)\n\n @pl.data_loader\n def val_dataloader(self):\n # OPTIONAL\n return DataLoader(self.val_data, shuffle=False, num_workers=20, pin_memory=True)", "_____no_output_____" ] ], [ [ "### Analysis functions", "_____no_output_____" ] ], [ [ "import seaborn as sns\n\ndef plot_channels(values):\n fig, ax = plt.subplots(nrows=2, ncols=4, figsize=(20, 10))\n\n ax = ax.flatten()\n\n channel_names = [\n 'Hydrophobic',\n 'Hydrogen bond',\n 'Van der Waals',\n 'Water bridges',\n 'Salt bridges',\n 'Pi-stacking',\n 'Pi-cation',\n 'T-stacking'\n ]\n\n for channel in range(values.shape[-1]):\n ax[channel].imshow(values[:, :, channel].squeeze(), vmin=0, vmax=1)\n ax[channel].set(title=channel_names[channel], xlabel='Residue #', ylabel='Residue #')\n\n plt.close()\n return fig\n\n\n######################\n# ROC and PRC curves #\n######################\nfrom sklearn.metrics import precision_recall_curve, roc_curve, auc\n\ndef calc_metric_curve(preds, target, curve_type, squareform=False):\n \"\"\"\n Calculate ROC or PRC curves and area for the predicted contact channels.\n\n Args:\n - preds (np.ndarray) - Numpy array of model predictions either of form\n (n_res, n_res, n_chan) or (n_res * [n_res - 1] / 2, n_chan).\n - target (np.ndarray) - Numpy array of target values either of form\n (n_res, n_res, n_chan) or (n_res * [n_res - 1] / 2, n_chan),\n must match form of preds.\n - curve_type (str) - One of 'ROC' or 'PRC' to denote type of curve.\n - squareform (bool) - True if tensors are of shape (n_res, n_res, n_chan),\n False if they are of shape (n_res * [n_res - 1] / 2, n_chan)\n (default = True).\n\n Returns:\n - Tuple of x, y, and AUC values to be used for plotting the curves\n using plot_curve metric.\n \"\"\"\n\n # Get correct curve function\n if curve_type.upper() == 'ROC':\n curve_func = roc_curve\n elif curve_type.upper() == 'PRC':\n curve_func = precision_recall_curve\n\n # Generate dicts to hold outputs from curve generation functions\n x = dict()\n y = dict()\n auc_ = dict()\n\n # Handle case of squareform matrix (only get non-redundant triu indices)\n if squareform:\n indices = np.triu_indices(target.shape[0])\n\n # For each channel\n for i in range(target.shape[-1]):\n\n # Handle case of squareform\n if squareform:\n var1, var2, _ = curve_func(target[:, :, i][indices],\n preds[:, :, i][indices])\n\n # Handle case of pairwise\n else:\n var1, var2, _ = curve_func(target[:, i], preds[:, i])\n\n # Assign outputs to correct dict for plotting\n if curve_type.upper() == 'ROC':\n x[i] = var1\n y[i] = var2\n elif curve_type.upper() == 'PRC':\n x[i] = var2\n y[i] = var1\n\n # Calc AUC\n auc_[i] = auc(x[i], y[i])\n\n return (x, y, auc_)\n\n\ndef plot_curve_metric(x, y, auc, curve_type, title=None, labels=None):\n \"\"\"\n Plot ROC or PRC curves per output channel.\n\n Args:\n - x (dict) - Dict of numpy arrays for values to plot on x axis.\n - y (dict) - Dict of numpy arrays for values to plot on x axis.\n - auc (dict) - Dict of numpy arrays for areas under each curve.\n - curve_type (str) - One of 'ROC' or 'PRC' to denote type of curve.\n - title\n - labels\n\n Returns:\n - pyplot object of curves.\n \"\"\"\n\n # Generate figure\n plt.figure()\n\n # Linetype spec\n lw = 2\n curve_type = curve_type.upper()\n\n # Get the number of channels being plotted\n n_chan = len(x)\n\n # Make labels numeric if not provided\n if labels is None:\n labels = list(range(n_chan))\n\n # Check to make sure the labels are the right length\n if len(labels) != n_chan:\n raise ValueError('Number of labels ({}) does not match number of prediction channels ({}).'.format(len(labels), n_chan))\n\n # Get a lit of colors for all the channels\n color_list = plt.cm.Set1(np.linspace(0, 1, n_chan))\n\n # Plot each line\n for i, color in enumerate(color_list):\n plt.plot(x[i], y[i], color=color,\n lw=lw, label='{} (area = {:0.2f})'.format(labels[i], auc[i]))\n\n # Add labels and diagonal line for ROC\n if curve_type == 'ROC':\n xlab = 'FPR'\n ylab = 'TPR'\n plt.plot([0, 1], [0, 1], color='navy', lw=lw, linestyle='--')\n plt.legend(loc=\"lower right\")\n\n # Add labels for PRC\n elif curve_type == 'PRC':\n xlab = 'Recall'\n ylab = 'Precision'\n plt.legend(loc=\"lower left\")\n\n # Extend limits, add labels and title\n plt.xlim([-0.05, 1.05])\n plt.ylim([-0.05, 1.05])\n plt.xlabel(xlab)\n plt.ylabel(ylab)\n\n if title is not None:\n plt.title('{} for {}'.format(curve_type, title))\n else:\n plt.title('{}'.format(curve_type))\n\n plt.show()\n\ndef plot_curve(preds, target, curve_type, title=None, labels=None,\n squareform=False):\n \"\"\"\n Wrapper to directly plot curves from model output and target.\n\n Args:\n - preds (np array-like) - Array or tensor of predicted values output by\n model.\n - target (np array-like) - Array or tensor of target values.\n - curve_type (str) - One of 'ROC' or 'PRC'.\n - title (str) - Title of plot (default = None).\n - labels (list) - List of labels for each channel on the plot\n (default = None).\n - squareform (bool) - Whether the predictions and targets are in square form\n (default = False).\n \"\"\"\n x, y, auc_ = calc_metric_curve(preds, target, curve_type, squareform)\n return plot_curve_metric(x, y, auc_, curve_type, title, labels)\n\n\n################################\n# Correlations between metrics #\n################################\n\ndef metric_corr_plot(corr):\n mask = np.triu(np.ones_like(corr, dtype=np.bool))\n\n fig, ax = plt.subplots(figsize=(11, 9))\n cmap = sns.diverging_palette(220, 10, as_cmap=True)\n\n sns.heatmap(corr, mask=mask, cmap=cmap, vmax=1.0, vmin = -1.0, center=0,\n square=True, linewidths=.5, cbar_kws={\"shrink\": .5}, ax=ax)\n\n plt.tight_layout()\n plt.close()\n return fig", "_____no_output_____" ] ], [ [ "## Training\n\n### Instantiate dataloader and model", "_____no_output_____" ] ], [ [ "from pytorch_lightning.callbacks import ModelCheckpoint\n\ndata_base = '/home/tshimko/tesselate/'\n\ntorch.manual_seed(10)\n\n\ndata_load = [\n 'pdb_id',\n# 'model',\n 'atom_nodes',\n 'atom_adj',\n# 'atom_contact',\n# 'atom_mask',\n 'res_adj',\n 'res_dist',\n 'chain_mem',\n 'res_contact',\n 'conn_adj',\n 'res_mask',\n 'mem_mat',\n# 'idx2atom_dict',\n# 'idx2res_dict'\n ]\n\n# train_data = TesselateDataset(data_base + 'id_lists/ligand_free_monomers/small_train.txt',\n# graph_dir=data_base + 'data/graphs',\n# contacts_dir=data_base + 'data/contacts',\n# return_data='all', in_memory=False)\n\ntrain_data = TesselateDataset(data_base + 'test4.txt',\n graph_dir=data_base + 'data/graphs',\n contacts_dir=data_base + 'data/contacts',\n return_data=data_load, in_memory=False)\n\nval_data = TesselateDataset(data_base + 'id_lists/ligand_free_monomers/small_val.txt',\n graph_dir=data_base + 'data/graphs',\n contacts_dir=data_base + 'data/contacts',\n return_data='all', in_memory=False)\n\n\ncheckpoint_callback = ModelCheckpoint(\n filepath=os.getcwd() + '/checkpoints',\n save_top_k=0,\n verbose=False,\n monitor='val_loss',\n mode='min',\n prefix=''\n)\n\nmodel = PGNN2D(input_dim=10, hidden_dim=10, output_dim=10,\n n_contact_channels=8,\n layer_num=4,\n train_data=train_data,\n val_data=val_data, test_data=None)", "_____no_output_____" ], [ "%%capture --no-stdout\nimport logging\nlogging.getLogger().setLevel(logging.CRITICAL)\n\nfrom pytorch_lightning import Trainer\nfrom pytorch_lightning.logging import WandbLogger\n\nlogger = WandbLogger(project='tesselate')\n\ntrainer = Trainer(max_nb_epochs=10000,\n accumulate_grad_batches=8,\n checkpoint_callback=checkpoint_callback,\n early_stop_callback=False,\n gpus=1,\n# distributed_backend='dp',\n logger=logger,\n fast_dev_run=False)\n# amp_level='O2', use_amp=True) # Needed for 16-bit training\ntrainer.fit(model)", "Epoch 1: 51%|█████▏ | 55/107 [00:32<00:25, 2.07batch/s, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 1: 52%|█████▏ | 56/107 [00:36<01:17, 1.53s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 53%|█████▎ | 57/107 [00:41<02:08, 2.57s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 54%|█████▍ | 58/107 [00:42<01:48, 2.21s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 55%|█████▌ | 59/107 [00:47<02:12, 2.76s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 56%|█████▌ | 60/107 [00:48<01:49, 2.34s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 57%|█████▋ | 61/107 [00:49<01:36, 2.10s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 58%|█████▊ | 62/107 [00:53<01:48, 2.41s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 59%|█████▉ | 63/107 [00:54<01:37, 2.21s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 60%|█████▉ | 64/107 [00:57<01:37, 2.28s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 61%|██████ | 65/107 [00:59<01:37, 2.32s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 62%|██████▏ | 66/107 [01:02<01:45, 2.56s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 63%|██████▎ | 67/107 [01:05<01:41, 2.54s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 64%|██████▎ | 68/107 [01:06<01:24, 2.16s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 64%|██████▍ | 69/107 [01:07<01:10, 1.86s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 65%|██████▌ | 70/107 [01:08<01:01, 1.65s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 66%|██████▋ | 71/107 [01:10<01:04, 1.78s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 67%|██████▋ | 72/107 [01:12<00:57, 1.63s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 68%|██████▊ | 73/107 [01:12<00:45, 1.35s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 69%|██████▉ | 74/107 [01:14<00:43, 1.31s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 70%|███████ | 75/107 [01:14<00:37, 1.16s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 71%|███████ | 76/107 [01:15<00:33, 1.08s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 72%|███████▏ | 77/107 [01:16<00:31, 1.06s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 73%|███████▎ | 78/107 [01:18<00:37, 1.28s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 74%|███████▍ | 79/107 [01:19<00:31, 1.14s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 75%|███████▍ | 80/107 [01:21<00:33, 1.26s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 76%|███████▌ | 81/107 [01:21<00:27, 1.06s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 77%|███████▋ | 82/107 [01:22<00:24, 1.01batch/s, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 78%|███████▊ | 83/107 [01:23<00:24, 1.02s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 79%|███████▊ | 84/107 [01:24<00:21, 1.06batch/s, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 79%|███████▉ | 85/107 [01:25<00:19, 1.10batch/s, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 80%|████████ | 86/107 [01:26<00:19, 1.08batch/s, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 81%|████████▏ | 87/107 [01:26<00:17, 1.11batch/s, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 82%|████████▏ | 88/107 [01:27<00:16, 1.16batch/s, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 83%|████████▎ | 89/107 [01:28<00:16, 1.10batch/s, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 84%|████████▍ | 90/107 [01:29<00:17, 1.02s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 85%|████████▌ | 91/107 [01:30<00:15, 1.05batch/s, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 86%|████████▌ | 92/107 [01:33<00:20, 1.38s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 87%|████████▋ | 93/107 [01:34<00:17, 1.22s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 88%|████████▊ | 94/107 [01:35<00:16, 1.30s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 89%|████████▉ | 95/107 [01:40<00:27, 2.27s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 90%|████████▉ | 96/107 [01:41<00:21, 1.92s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 91%|█████████ | 97/107 [01:43<00:20, 2.01s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 92%|█████████▏| 98/107 [01:44<00:15, 1.67s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 93%|█████████▎| 99/107 [01:45<00:11, 1.45s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 93%|█████████▎| 100/107 [01:45<00:08, 1.24s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 94%|█████████▍| 101/107 [01:46<00:06, 1.10s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 95%|█████████▌| 102/107 [01:48<00:05, 1.17s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 96%|█████████▋| 103/107 [01:48<00:04, 1.06s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 97%|█████████▋| 104/107 [01:49<00:02, 1.01batch/s, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 98%|█████████▊| 105/107 [01:50<00:01, 1.03batch/s, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 99%|█████████▉| 106/107 [01:52<00:01, 1.22s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 1: 100%|██████████| 107/107 [01:57<00:00, 1.07s/batch, batch_idx=54, gpu=0, loss=0.503, v_num=zf852wxn]\nEpoch 2: 51%|█████▏ | 55/107 [00:32<00:29, 1.76batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 2: 52%|█████▏ | 56/107 [00:36<01:23, 1.64s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 53%|█████▎ | 57/107 [00:40<01:50, 2.21s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 54%|█████▍ | 58/107 [00:41<01:34, 1.93s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 55%|█████▌ | 59/107 [00:45<02:11, 2.74s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 56%|█████▌ | 60/107 [00:47<01:46, 2.27s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 57%|█████▋ | 61/107 [00:48<01:30, 1.96s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 58%|█████▊ | 62/107 [00:52<02:02, 2.72s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 59%|█████▉ | 63/107 [00:53<01:39, 2.25s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 60%|█████▉ | 64/107 [00:55<01:31, 2.12s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 61%|██████ | 65/107 [00:57<01:19, 1.90s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 62%|██████▏ | 66/107 [00:58<01:12, 1.78s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 63%|██████▎ | 67/107 [00:59<01:02, 1.55s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 64%|██████▎ | 68/107 [01:00<00:52, 1.35s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 64%|██████▍ | 69/107 [01:03<01:07, 1.77s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 65%|██████▌ | 70/107 [01:05<01:07, 1.82s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 66%|██████▋ | 71/107 [01:06<01:03, 1.77s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 67%|██████▋ | 72/107 [01:08<00:56, 1.60s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 68%|██████▊ | 73/107 [01:08<00:45, 1.33s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 69%|██████▉ | 74/107 [01:09<00:42, 1.28s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 70%|███████ | 75/107 [01:10<00:36, 1.13s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 71%|███████ | 76/107 [01:12<00:41, 1.33s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 72%|███████▏ | 77/107 [01:13<00:36, 1.23s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 73%|███████▎ | 78/107 [01:15<00:40, 1.39s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 74%|███████▍ | 79/107 [01:16<00:33, 1.21s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 75%|███████▍ | 80/107 [01:17<00:31, 1.16s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 76%|███████▌ | 81/107 [01:17<00:25, 1.01batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 77%|███████▋ | 82/107 [01:18<00:23, 1.08batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 78%|███████▊ | 83/107 [01:19<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 79%|███████▊ | 84/107 [01:20<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 79%|███████▉ | 85/107 [01:21<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 80%|████████ | 86/107 [01:21<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 81%|████████▏ | 87/107 [01:22<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 82%|████████▏ | 88/107 [01:23<00:15, 1.21batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 83%|████████▎ | 89/107 [01:24<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 84%|████████▍ | 90/107 [01:25<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 85%|████████▌ | 91/107 [01:26<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 86%|████████▌ | 92/107 [01:28<00:17, 1.16s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 87%|████████▋ | 93/107 [01:28<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 88%|████████▊ | 94/107 [01:30<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 89%|████████▉ | 95/107 [01:34<00:25, 2.10s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 90%|████████▉ | 96/107 [01:35<00:19, 1.79s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 91%|█████████ | 97/107 [01:37<00:19, 1.90s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 92%|█████████▏| 98/107 [01:38<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 93%|█████████▎| 99/107 [01:40<00:14, 1.77s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 93%|█████████▎| 100/107 [01:41<00:10, 1.46s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 94%|█████████▍| 101/107 [01:42<00:07, 1.23s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 95%|█████████▌| 102/107 [01:43<00:06, 1.24s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 96%|█████████▋| 103/107 [01:44<00:04, 1.09s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 97%|█████████▋| 104/107 [01:45<00:02, 1.01batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 98%|█████████▊| 105/107 [01:46<00:01, 1.04batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 99%|█████████▉| 106/107 [01:47<00:01, 1.04s/batch, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 2: 100%|██████████| 107/107 [01:51<00:00, 1.05batch/s, batch_idx=54, gpu=0, loss=0.395, v_num=zf852wxn]\nEpoch 3: 51%|█████▏ | 55/107 [00:32<00:29, 1.75batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 3: 52%|█████▏ | 56/107 [00:37<01:35, 1.87s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 53%|█████▎ | 57/107 [00:40<01:51, 2.23s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 54%|█████▍ | 58/107 [00:42<01:37, 1.99s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 55%|█████▌ | 59/107 [00:45<01:59, 2.49s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 56%|█████▌ | 60/107 [00:47<01:44, 2.22s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 57%|█████▋ | 61/107 [00:48<01:27, 1.91s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 58%|█████▊ | 62/107 [00:50<01:26, 1.93s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 59%|█████▉ | 63/107 [00:51<01:14, 1.70s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 60%|█████▉ | 64/107 [00:55<01:33, 2.18s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 61%|██████ | 65/107 [00:56<01:16, 1.83s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 62%|██████▏ | 66/107 [00:57<01:06, 1.63s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 63%|██████▎ | 67/107 [00:58<00:59, 1.49s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 64%|██████▎ | 68/107 [00:59<00:50, 1.31s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 64%|██████▍ | 69/107 [01:00<00:49, 1.29s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 65%|██████▌ | 70/107 [01:01<00:45, 1.22s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 66%|██████▋ | 71/107 [01:03<00:54, 1.51s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 67%|██████▋ | 72/107 [01:05<00:49, 1.42s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 68%|██████▊ | 73/107 [01:05<00:40, 1.18s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 69%|██████▉ | 74/107 [01:06<00:38, 1.18s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 70%|███████ | 75/107 [01:07<00:33, 1.05s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 71%|███████ | 76/107 [01:08<00:30, 1.03batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 72%|███████▏ | 77/107 [01:09<00:28, 1.04batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 73%|███████▎ | 78/107 [01:11<00:34, 1.19s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 74%|███████▍ | 79/107 [01:11<00:29, 1.07s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 75%|███████▍ | 80/107 [01:12<00:28, 1.06s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 76%|███████▌ | 81/107 [01:13<00:24, 1.08batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 77%|███████▋ | 82/107 [01:14<00:21, 1.14batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 78%|███████▊ | 83/107 [01:15<00:22, 1.08batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 79%|███████▊ | 84/107 [01:16<00:19, 1.16batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 79%|███████▉ | 85/107 [01:16<00:18, 1.21batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 80%|████████ | 86/107 [01:17<00:17, 1.19batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 81%|████████▏ | 87/107 [01:18<00:16, 1.22batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 82%|████████▏ | 88/107 [01:19<00:15, 1.27batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 83%|████████▎ | 89/107 [01:20<00:14, 1.21batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 84%|████████▍ | 90/107 [01:21<00:16, 1.06batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 85%|████████▌ | 91/107 [01:21<00:14, 1.13batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 86%|████████▌ | 92/107 [01:23<00:17, 1.16s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 87%|████████▋ | 93/107 [01:24<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 88%|████████▊ | 94/107 [01:26<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 89%|████████▉ | 95/107 [01:31<00:30, 2.58s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 90%|████████▉ | 96/107 [01:32<00:23, 2.13s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 91%|█████████ | 97/107 [01:35<00:21, 2.15s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 92%|█████████▏| 98/107 [01:36<00:15, 1.77s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 93%|█████████▎| 99/107 [01:36<00:12, 1.50s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 93%|█████████▎| 100/107 [01:37<00:08, 1.27s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 94%|█████████▍| 101/107 [01:38<00:06, 1.11s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 95%|█████████▌| 102/107 [01:39<00:05, 1.16s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 96%|█████████▋| 103/107 [01:40<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 97%|█████████▋| 104/107 [01:41<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 98%|█████████▊| 105/107 [01:41<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 99%|█████████▉| 106/107 [01:43<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 3: 100%|██████████| 107/107 [01:46<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.289, v_num=zf852wxn]\nEpoch 4: 51%|█████▏ | 55/107 [00:33<00:37, 1.38batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 4: 52%|█████▏ | 56/107 [00:38<01:39, 1.96s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 53%|█████▎ | 57/107 [00:42<01:58, 2.38s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 54%|█████▍ | 58/107 [00:43<01:43, 2.11s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 55%|█████▌ | 59/107 [00:45<01:37, 2.03s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 56%|█████▌ | 60/107 [00:46<01:29, 1.90s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 57%|█████▋ | 61/107 [00:48<01:16, 1.67s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 58%|█████▊ | 62/107 [00:50<01:24, 1.88s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 59%|█████▉ | 63/107 [00:51<01:12, 1.65s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 60%|█████▉ | 64/107 [00:53<01:10, 1.64s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 61%|██████ | 65/107 [00:54<00:58, 1.40s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 62%|██████▏ | 66/107 [00:55<00:57, 1.39s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 63%|██████▎ | 67/107 [00:56<00:55, 1.39s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 64%|██████▎ | 68/107 [00:57<00:47, 1.23s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 64%|██████▍ | 69/107 [00:58<00:47, 1.24s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 65%|██████▌ | 70/107 [01:02<01:06, 1.81s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 66%|██████▋ | 71/107 [01:03<01:02, 1.75s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 67%|██████▋ | 72/107 [01:05<01:01, 1.76s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 68%|██████▊ | 73/107 [01:06<00:48, 1.42s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 69%|██████▉ | 74/107 [01:07<00:44, 1.34s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 70%|███████ | 75/107 [01:08<00:37, 1.16s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 71%|███████ | 76/107 [01:08<00:32, 1.05s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 72%|███████▏ | 77/107 [01:09<00:30, 1.02s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 73%|███████▎ | 78/107 [01:11<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 74%|███████▍ | 79/107 [01:12<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 75%|███████▍ | 80/107 [01:13<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 76%|███████▌ | 81/107 [01:13<00:24, 1.08batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 77%|███████▋ | 82/107 [01:14<00:21, 1.14batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 78%|███████▊ | 83/107 [01:15<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 79%|███████▊ | 84/107 [01:16<00:19, 1.16batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 79%|███████▉ | 85/107 [01:17<00:18, 1.21batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 80%|████████ | 86/107 [01:17<00:17, 1.20batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 81%|████████▏ | 87/107 [01:18<00:16, 1.23batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 82%|████████▏ | 88/107 [01:19<00:14, 1.28batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 83%|████████▎ | 89/107 [01:20<00:14, 1.22batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 84%|████████▍ | 90/107 [01:21<00:15, 1.06batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 85%|████████▌ | 91/107 [01:22<00:14, 1.13batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 86%|████████▌ | 92/107 [01:24<00:17, 1.15s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 87%|████████▋ | 93/107 [01:24<00:14, 1.03s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 88%|████████▊ | 94/107 [01:26<00:14, 1.15s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 89%|████████▉ | 95/107 [01:30<00:24, 2.08s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 90%|████████▉ | 96/107 [01:31<00:19, 1.77s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 91%|█████████ | 97/107 [01:33<00:18, 1.88s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 92%|█████████▏| 98/107 [01:34<00:14, 1.57s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 93%|█████████▎| 99/107 [01:35<00:10, 1.36s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 93%|█████████▎| 100/107 [01:36<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 94%|█████████▍| 101/107 [01:36<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 95%|█████████▌| 102/107 [01:38<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 96%|█████████▋| 103/107 [01:38<00:03, 1.00batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 97%|█████████▋| 104/107 [01:39<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 98%|█████████▊| 105/107 [01:40<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 99%|█████████▉| 106/107 [01:41<00:00, 1.04batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 4: 100%|██████████| 107/107 [01:45<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.227, v_num=zf852wxn]\nEpoch 5: 51%|█████▏ | 55/107 [00:35<00:28, 1.85batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 5: 52%|█████▏ | 56/107 [00:40<01:43, 2.02s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 53%|█████▎ | 57/107 [00:43<01:51, 2.24s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 54%|█████▍ | 58/107 [00:44<01:34, 1.92s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 55%|█████▌ | 59/107 [00:47<01:54, 2.39s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 56%|█████▌ | 60/107 [00:52<02:27, 3.14s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 57%|█████▋ | 61/107 [00:53<01:54, 2.48s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 58%|█████▊ | 62/107 [00:57<02:04, 2.76s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 59%|█████▉ | 63/107 [00:58<01:41, 2.30s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 60%|█████▉ | 64/107 [00:59<01:27, 2.03s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 61%|██████ | 65/107 [01:01<01:15, 1.80s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 62%|██████▏ | 66/107 [01:02<01:12, 1.77s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 63%|██████▎ | 67/107 [01:03<01:02, 1.55s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 64%|██████▎ | 68/107 [01:04<00:53, 1.37s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 64%|██████▍ | 69/107 [01:06<00:57, 1.52s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 65%|██████▌ | 70/107 [01:07<00:51, 1.39s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 66%|██████▋ | 71/107 [01:09<00:52, 1.45s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 67%|██████▋ | 72/107 [01:10<00:48, 1.38s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 68%|██████▊ | 73/107 [01:11<00:39, 1.16s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 69%|██████▉ | 74/107 [01:12<00:38, 1.17s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 70%|███████ | 75/107 [01:13<00:33, 1.06s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 71%|███████ | 76/107 [01:14<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 72%|███████▏ | 77/107 [01:15<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 73%|███████▎ | 78/107 [01:16<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 74%|███████▍ | 79/107 [01:17<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 75%|███████▍ | 80/107 [01:18<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 76%|███████▌ | 81/107 [01:19<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 77%|███████▋ | 82/107 [01:20<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 78%|███████▊ | 83/107 [01:21<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 79%|███████▊ | 84/107 [01:21<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 79%|███████▉ | 85/107 [01:22<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 80%|████████ | 86/107 [01:23<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 81%|████████▏ | 87/107 [01:24<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 82%|████████▏ | 88/107 [01:25<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 83%|████████▎ | 89/107 [01:26<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 84%|████████▍ | 90/107 [01:27<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 85%|████████▌ | 91/107 [01:28<00:14, 1.08batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 86%|████████▌ | 92/107 [01:29<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 87%|████████▋ | 93/107 [01:30<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 88%|████████▊ | 94/107 [01:32<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 89%|████████▉ | 95/107 [01:36<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 90%|████████▉ | 96/107 [01:37<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 91%|█████████ | 97/107 [01:39<00:19, 1.91s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 92%|█████████▏| 98/107 [01:40<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 93%|█████████▎| 99/107 [01:41<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 93%|█████████▎| 100/107 [01:42<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 94%|█████████▍| 101/107 [01:42<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 95%|█████████▌| 102/107 [01:44<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 96%|█████████▋| 103/107 [01:44<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 97%|█████████▋| 104/107 [01:45<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 98%|█████████▊| 105/107 [01:46<00:01, 1.08batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 99%|█████████▉| 106/107 [01:47<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 5: 100%|██████████| 107/107 [01:51<00:00, 1.10batch/s, batch_idx=54, gpu=0, loss=0.191, v_num=zf852wxn]\nEpoch 6: 51%|█████▏ | 55/107 [00:33<00:26, 1.93batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 6: 52%|█████▏ | 56/107 [00:42<02:37, 3.08s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 53%|█████▎ | 57/107 [00:44<02:19, 2.79s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 54%|█████▍ | 58/107 [00:46<01:56, 2.39s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 55%|█████▌ | 59/107 [00:48<01:46, 2.21s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 56%|█████▌ | 60/107 [00:49<01:32, 1.96s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 57%|█████▋ | 61/107 [00:50<01:15, 1.64s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 58%|█████▊ | 62/107 [00:54<01:41, 2.25s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 59%|█████▉ | 63/107 [00:55<01:26, 1.97s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 60%|█████▉ | 64/107 [00:56<01:17, 1.80s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 61%|██████ | 65/107 [00:57<01:06, 1.57s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 62%|██████▏ | 66/107 [00:59<01:05, 1.61s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 63%|██████▎ | 67/107 [01:00<00:58, 1.47s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 64%|██████▎ | 68/107 [01:01<00:49, 1.28s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 64%|██████▍ | 69/107 [01:02<00:47, 1.26s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 65%|██████▌ | 70/107 [01:03<00:43, 1.18s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 66%|██████▋ | 71/107 [01:05<00:47, 1.31s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 67%|██████▋ | 72/107 [01:06<00:45, 1.29s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 68%|██████▊ | 73/107 [01:07<00:37, 1.11s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 69%|██████▉ | 74/107 [01:08<00:37, 1.13s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 70%|███████ | 75/107 [01:09<00:32, 1.02s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 71%|███████ | 76/107 [01:10<00:30, 1.02batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 72%|███████▏ | 77/107 [01:11<00:35, 1.18s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 73%|███████▎ | 78/107 [01:13<00:38, 1.34s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 74%|███████▍ | 79/107 [01:14<00:32, 1.18s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 75%|███████▍ | 80/107 [01:15<00:30, 1.14s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 76%|███████▌ | 81/107 [01:15<00:25, 1.03batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 77%|███████▋ | 82/107 [01:16<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 78%|███████▊ | 83/107 [01:17<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 79%|███████▊ | 84/107 [01:18<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 79%|███████▉ | 85/107 [01:19<00:18, 1.19batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 80%|████████ | 86/107 [01:20<00:17, 1.18batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 81%|████████▏ | 87/107 [01:20<00:16, 1.21batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 82%|████████▏ | 88/107 [01:21<00:15, 1.26batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 83%|████████▎ | 89/107 [01:22<00:14, 1.21batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 84%|████████▍ | 90/107 [01:23<00:15, 1.07batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 85%|████████▌ | 91/107 [01:24<00:14, 1.13batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 86%|████████▌ | 92/107 [01:26<00:17, 1.16s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 87%|████████▋ | 93/107 [01:26<00:14, 1.04s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 88%|████████▊ | 94/107 [01:28<00:14, 1.15s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 89%|████████▉ | 95/107 [01:32<00:25, 2.09s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 90%|████████▉ | 96/107 [01:33<00:19, 1.77s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 91%|█████████ | 97/107 [01:35<00:18, 1.88s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 92%|█████████▏| 98/107 [01:36<00:14, 1.56s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 93%|█████████▎| 99/107 [01:37<00:10, 1.35s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 93%|█████████▎| 100/107 [01:38<00:08, 1.16s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 94%|█████████▍| 101/107 [01:38<00:06, 1.02s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 95%|█████████▌| 102/107 [01:40<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 96%|█████████▋| 103/107 [01:40<00:03, 1.01batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 97%|█████████▋| 104/107 [01:41<00:02, 1.10batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 98%|█████████▊| 105/107 [01:42<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 99%|█████████▉| 106/107 [01:43<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 6: 100%|██████████| 107/107 [01:47<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.166, v_num=zf852wxn]\nEpoch 7: 51%|█████▏ | 55/107 [00:34<00:32, 1.58batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 7: 52%|█████▏ | 56/107 [00:39<01:44, 2.05s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 53%|█████▎ | 57/107 [00:48<03:29, 4.19s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 54%|█████▍ | 58/107 [00:50<02:43, 3.33s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 55%|█████▌ | 59/107 [00:51<02:19, 2.91s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 56%|█████▌ | 60/107 [00:53<01:57, 2.51s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 57%|█████▋ | 61/107 [00:54<01:36, 2.09s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 58%|█████▊ | 62/107 [00:56<01:32, 2.05s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 59%|█████▉ | 63/107 [00:57<01:17, 1.75s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 60%|█████▉ | 64/107 [00:59<01:15, 1.75s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 61%|██████ | 65/107 [01:00<01:03, 1.52s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 62%|██████▏ | 66/107 [01:01<00:59, 1.45s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 63%|██████▎ | 67/107 [01:02<00:53, 1.34s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 64%|██████▎ | 68/107 [01:03<00:47, 1.21s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 64%|██████▍ | 69/107 [01:05<00:53, 1.40s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 65%|██████▌ | 70/107 [01:06<00:47, 1.28s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 66%|██████▋ | 71/107 [01:08<00:49, 1.37s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 67%|██████▋ | 72/107 [01:09<00:45, 1.31s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 68%|██████▊ | 73/107 [01:09<00:37, 1.11s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 69%|██████▉ | 74/107 [01:11<00:37, 1.12s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 70%|███████ | 75/107 [01:11<00:32, 1.02s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 71%|███████ | 76/107 [01:12<00:29, 1.06batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 72%|███████▏ | 77/107 [01:13<00:28, 1.06batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 73%|███████▎ | 78/107 [01:15<00:34, 1.17s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 74%|███████▍ | 79/107 [01:16<00:29, 1.05s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 75%|███████▍ | 80/107 [01:17<00:28, 1.06s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 76%|███████▌ | 81/107 [01:17<00:23, 1.09batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 77%|███████▋ | 82/107 [01:18<00:21, 1.15batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 78%|███████▊ | 83/107 [01:19<00:21, 1.09batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 79%|███████▊ | 84/107 [01:20<00:19, 1.17batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 79%|███████▉ | 85/107 [01:20<00:18, 1.21batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 80%|████████ | 86/107 [01:21<00:17, 1.20batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 81%|████████▏ | 87/107 [01:22<00:16, 1.23batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 82%|████████▏ | 88/107 [01:23<00:14, 1.27batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 83%|████████▎ | 89/107 [01:24<00:14, 1.20batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 84%|████████▍ | 90/107 [01:25<00:16, 1.05batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 85%|████████▌ | 91/107 [01:26<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 86%|████████▌ | 92/107 [01:28<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 87%|████████▋ | 93/107 [01:28<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 88%|████████▊ | 94/107 [01:30<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 89%|████████▉ | 95/107 [01:34<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 90%|████████▉ | 96/107 [01:35<00:19, 1.79s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 91%|█████████ | 97/107 [01:37<00:18, 1.90s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 92%|█████████▏| 98/107 [01:38<00:14, 1.56s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 93%|█████████▎| 99/107 [01:39<00:10, 1.35s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 93%|█████████▎| 100/107 [01:40<00:08, 1.16s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 94%|█████████▍| 101/107 [01:40<00:06, 1.03s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 95%|█████████▌| 102/107 [01:42<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 96%|█████████▋| 103/107 [01:42<00:03, 1.01batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 97%|█████████▋| 104/107 [01:43<00:02, 1.10batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 98%|█████████▊| 105/107 [01:44<00:01, 1.13batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 99%|█████████▉| 106/107 [01:45<00:00, 1.04batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 7: 100%|██████████| 107/107 [01:49<00:00, 1.15batch/s, batch_idx=54, gpu=0, loss=0.148, v_num=zf852wxn]\nEpoch 8: 51%|█████▏ | 55/107 [00:34<00:28, 1.82batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 8: 52%|█████▏ | 56/107 [00:40<01:48, 2.12s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 53%|█████▎ | 57/107 [00:45<02:20, 2.80s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 54%|█████▍ | 58/107 [00:46<01:58, 2.43s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 55%|█████▌ | 59/107 [00:49<01:56, 2.42s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 56%|█████▌ | 60/107 [00:50<01:38, 2.09s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 57%|█████▋ | 61/107 [00:51<01:19, 1.72s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 58%|█████▊ | 62/107 [00:56<02:00, 2.67s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 59%|█████▉ | 63/107 [00:57<01:36, 2.19s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 60%|█████▉ | 64/107 [00:58<01:23, 1.93s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 61%|██████ | 65/107 [01:09<03:10, 4.55s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 62%|██████▏ | 66/107 [01:10<02:32, 3.73s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 63%|██████▎ | 67/107 [01:12<02:01, 3.03s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 64%|██████▎ | 68/107 [01:14<01:42, 2.63s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 64%|██████▍ | 69/107 [01:15<01:22, 2.18s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 65%|██████▌ | 70/107 [01:16<01:08, 1.84s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 66%|██████▋ | 71/107 [01:17<01:04, 1.80s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 67%|██████▋ | 72/107 [01:19<00:56, 1.62s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 68%|██████▊ | 73/107 [01:19<00:45, 1.33s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 69%|██████▉ | 74/107 [01:20<00:42, 1.29s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 70%|███████ | 75/107 [01:21<00:36, 1.13s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 71%|███████ | 76/107 [01:22<00:31, 1.03s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 72%|███████▏ | 77/107 [01:23<00:30, 1.02s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 73%|███████▎ | 78/107 [01:25<00:36, 1.25s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 74%|███████▍ | 79/107 [01:26<00:31, 1.11s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 75%|███████▍ | 80/107 [01:27<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 76%|███████▌ | 81/107 [01:27<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 77%|███████▋ | 82/107 [01:28<00:22, 1.12batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 78%|███████▊ | 83/107 [01:29<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 79%|███████▊ | 84/107 [01:30<00:20, 1.15batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 79%|███████▉ | 85/107 [01:31<00:18, 1.19batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 80%|████████ | 86/107 [01:31<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 81%|████████▏ | 87/107 [01:32<00:16, 1.20batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 82%|████████▏ | 88/107 [01:33<00:15, 1.25batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 83%|████████▎ | 89/107 [01:34<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 84%|████████▍ | 90/107 [01:35<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 85%|████████▌ | 91/107 [01:36<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 86%|████████▌ | 92/107 [01:38<00:18, 1.20s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 87%|████████▋ | 93/107 [01:39<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 88%|████████▊ | 94/107 [01:40<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 89%|████████▉ | 95/107 [01:45<00:27, 2.25s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 90%|████████▉ | 96/107 [01:46<00:20, 1.89s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 91%|█████████ | 97/107 [01:48<00:20, 2.01s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 92%|█████████▏| 98/107 [01:49<00:14, 1.65s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 93%|█████████▎| 99/107 [01:50<00:11, 1.42s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 93%|█████████▎| 100/107 [01:51<00:08, 1.22s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 94%|█████████▍| 101/107 [01:51<00:06, 1.06s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 95%|█████████▌| 102/107 [01:53<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 96%|█████████▋| 103/107 [01:53<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 97%|█████████▋| 104/107 [01:54<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 98%|█████████▊| 105/107 [01:55<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 99%|█████████▉| 106/107 [01:56<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 8: 100%|██████████| 107/107 [02:00<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.134, v_num=zf852wxn]\nEpoch 9: 51%|█████▏ | 55/107 [00:36<00:23, 2.22batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 9: 52%|█████▏ | 56/107 [00:43<02:01, 2.39s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 53%|█████▎ | 57/107 [00:46<02:07, 2.56s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 54%|█████▍ | 58/107 [00:48<01:48, 2.22s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 55%|█████▌ | 59/107 [00:51<02:10, 2.72s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 56%|█████▌ | 60/107 [00:53<01:48, 2.32s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 57%|█████▋ | 61/107 [00:54<01:28, 1.92s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 58%|█████▊ | 62/107 [00:58<01:59, 2.66s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 59%|█████▉ | 63/107 [00:59<01:37, 2.22s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 60%|█████▉ | 64/107 [01:01<01:24, 1.97s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 61%|██████ | 65/107 [01:02<01:11, 1.70s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 62%|██████▏ | 66/107 [01:04<01:10, 1.73s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 63%|██████▎ | 67/107 [01:05<01:02, 1.55s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 64%|██████▎ | 68/107 [01:06<00:53, 1.38s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 64%|██████▍ | 69/107 [01:07<00:51, 1.36s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 65%|██████▌ | 70/107 [01:08<00:46, 1.26s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 66%|██████▋ | 71/107 [01:10<00:55, 1.55s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 67%|██████▋ | 72/107 [01:12<00:50, 1.45s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 68%|██████▊ | 73/107 [01:12<00:41, 1.21s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 69%|██████▉ | 74/107 [01:13<00:39, 1.20s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 70%|███████ | 75/107 [01:14<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 71%|███████ | 76/107 [01:15<00:30, 1.00batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 72%|███████▏ | 77/107 [01:16<00:29, 1.00batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 73%|███████▎ | 78/107 [01:18<00:35, 1.24s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 74%|███████▍ | 79/107 [01:18<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 75%|███████▍ | 80/107 [01:20<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 76%|███████▌ | 81/107 [01:20<00:24, 1.07batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 77%|███████▋ | 82/107 [01:21<00:22, 1.13batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 78%|███████▊ | 83/107 [01:22<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 79%|███████▊ | 84/107 [01:23<00:20, 1.15batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 79%|███████▉ | 85/107 [01:23<00:18, 1.20batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 80%|████████ | 86/107 [01:24<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 81%|████████▏ | 87/107 [01:25<00:16, 1.21batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 82%|████████▏ | 88/107 [01:26<00:15, 1.26batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 83%|████████▎ | 89/107 [01:27<00:14, 1.20batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 84%|████████▍ | 90/107 [01:28<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 85%|████████▌ | 91/107 [01:29<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 86%|████████▌ | 92/107 [01:31<00:18, 1.20s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 87%|████████▋ | 93/107 [01:37<00:39, 2.85s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 88%|████████▊ | 94/107 [01:39<00:31, 2.44s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 89%|████████▉ | 95/107 [01:43<00:37, 3.10s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 90%|████████▉ | 96/107 [01:45<00:27, 2.49s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 91%|█████████ | 97/107 [01:47<00:24, 2.42s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 92%|█████████▏| 98/107 [01:48<00:17, 1.94s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 93%|█████████▎| 99/107 [01:48<00:12, 1.62s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 93%|█████████▎| 100/107 [01:49<00:09, 1.35s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 94%|█████████▍| 101/107 [01:50<00:06, 1.16s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 95%|█████████▌| 102/107 [01:51<00:06, 1.21s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 96%|█████████▋| 103/107 [01:52<00:04, 1.07s/batch, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 97%|█████████▋| 104/107 [01:53<00:02, 1.03batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 98%|█████████▊| 105/107 [01:54<00:01, 1.07batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 99%|█████████▉| 106/107 [01:55<00:00, 1.00batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 9: 100%|██████████| 107/107 [01:59<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.123, v_num=zf852wxn]\nEpoch 10: 51%|█████▏ | 55/107 [00:36<00:19, 2.70batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 10: 52%|█████▏ | 56/107 [00:43<01:59, 2.35s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 53%|█████▎ | 57/107 [00:46<02:11, 2.63s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 54%|█████▍ | 58/107 [00:48<01:48, 2.22s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 55%|█████▌ | 59/107 [00:51<02:06, 2.64s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 56%|█████▌ | 60/107 [00:53<01:47, 2.29s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 57%|█████▋ | 61/107 [00:53<01:25, 1.86s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 58%|█████▊ | 62/107 [00:58<01:56, 2.58s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 59%|█████▉ | 63/107 [00:59<01:35, 2.18s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 60%|█████▉ | 64/107 [01:00<01:22, 1.91s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 61%|██████ | 65/107 [01:01<01:10, 1.68s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 62%|██████▏ | 66/107 [01:03<01:10, 1.72s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 63%|██████▎ | 67/107 [01:05<01:04, 1.62s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 64%|██████▎ | 68/107 [01:06<00:54, 1.40s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 64%|██████▍ | 69/107 [01:07<00:52, 1.38s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 65%|██████▌ | 70/107 [01:08<00:47, 1.29s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 66%|██████▋ | 71/107 [01:10<00:49, 1.38s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 67%|██████▋ | 72/107 [01:11<00:53, 1.52s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 68%|██████▊ | 73/107 [01:12<00:42, 1.26s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 69%|██████▉ | 74/107 [01:13<00:40, 1.24s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 70%|███████ | 75/107 [01:14<00:34, 1.09s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 71%|███████ | 76/107 [01:15<00:31, 1.00s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 72%|███████▏ | 77/107 [01:16<00:29, 1.00batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 73%|███████▎ | 78/107 [01:18<00:35, 1.24s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 74%|███████▍ | 79/107 [01:18<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 75%|███████▍ | 80/107 [01:19<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 76%|███████▌ | 81/107 [01:20<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 77%|███████▋ | 82/107 [01:21<00:22, 1.12batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 78%|███████▊ | 83/107 [01:22<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 79%|███████▊ | 84/107 [01:22<00:19, 1.15batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 79%|███████▉ | 85/107 [01:23<00:18, 1.20batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 80%|████████ | 86/107 [01:24<00:17, 1.18batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 81%|████████▏ | 87/107 [01:25<00:16, 1.20batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 82%|████████▏ | 88/107 [01:26<00:15, 1.25batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 83%|████████▎ | 89/107 [01:27<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 84%|████████▍ | 90/107 [01:28<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 85%|████████▌ | 91/107 [01:29<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 86%|████████▌ | 92/107 [01:30<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 87%|████████▋ | 93/107 [01:31<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 88%|████████▊ | 94/107 [01:33<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 89%|████████▉ | 95/107 [01:37<00:26, 2.22s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 90%|████████▉ | 96/107 [01:38<00:20, 1.88s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 91%|█████████ | 97/107 [01:41<00:19, 2.00s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 92%|█████████▏| 98/107 [01:41<00:14, 1.64s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 93%|█████████▎| 99/107 [01:42<00:11, 1.41s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 93%|█████████▎| 100/107 [01:43<00:08, 1.20s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 94%|█████████▍| 101/107 [01:44<00:06, 1.06s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 95%|█████████▌| 102/107 [01:45<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 96%|█████████▋| 103/107 [01:46<00:04, 1.04s/batch, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 97%|█████████▋| 104/107 [01:47<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 98%|█████████▊| 105/107 [01:47<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 99%|█████████▉| 106/107 [01:49<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 10: 100%|██████████| 107/107 [01:52<00:00, 1.10batch/s, batch_idx=54, gpu=0, loss=0.114, v_num=zf852wxn]\nEpoch 11: 51%|█████▏ | 55/107 [00:37<00:32, 1.61batch/s, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 11: 52%|█████▏ | 56/107 [00:44<02:12, 2.61s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 53%|█████▎ | 57/107 [00:48<02:21, 2.83s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 54%|█████▍ | 58/107 [00:49<01:57, 2.40s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 55%|█████▌ | 59/107 [00:52<02:04, 2.59s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 56%|█████▌ | 60/107 [00:54<01:46, 2.27s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 57%|█████▋ | 61/107 [00:55<01:25, 1.86s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 58%|█████▊ | 62/107 [00:59<02:01, 2.69s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 59%|█████▉ | 63/107 [01:00<01:36, 2.19s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 60%|█████▉ | 64/107 [01:01<01:22, 1.93s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 61%|██████ | 65/107 [01:03<01:11, 1.71s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 62%|██████▏ | 66/107 [01:04<01:10, 1.72s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 63%|██████▎ | 67/107 [01:06<01:04, 1.61s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 64%|██████▎ | 68/107 [01:07<00:54, 1.40s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 64%|██████▍ | 69/107 [01:08<00:51, 1.35s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 65%|██████▌ | 70/107 [01:09<00:52, 1.41s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 66%|██████▋ | 71/107 [01:11<00:53, 1.48s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 67%|██████▋ | 72/107 [01:12<00:49, 1.41s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 68%|██████▊ | 73/107 [01:13<00:39, 1.18s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 69%|██████▉ | 74/107 [01:14<00:38, 1.18s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 70%|███████ | 75/107 [01:15<00:33, 1.05s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 71%|███████ | 76/107 [01:16<00:30, 1.03batch/s, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 72%|███████▏ | 77/107 [01:17<00:29, 1.03batch/s, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 73%|███████▎ | 78/107 [01:27<01:53, 3.90s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 74%|███████▍ | 79/107 [01:28<01:22, 2.95s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 75%|███████▍ | 80/107 [01:29<01:04, 2.39s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 76%|███████▌ | 81/107 [01:30<00:48, 1.85s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 77%|███████▋ | 82/107 [01:31<00:38, 1.53s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 78%|███████▊ | 83/107 [01:32<00:33, 1.38s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 79%|███████▊ | 84/107 [01:32<00:27, 1.18s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 79%|███████▉ | 85/107 [01:33<00:23, 1.05s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 80%|████████ | 86/107 [01:34<00:20, 1.00batch/s, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 81%|████████▏ | 87/107 [01:35<00:18, 1.08batch/s, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 82%|████████▏ | 88/107 [01:35<00:16, 1.15batch/s, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 83%|████████▎ | 89/107 [01:36<00:15, 1.13batch/s, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 84%|████████▍ | 90/107 [01:38<00:16, 1.01batch/s, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 85%|████████▌ | 91/107 [01:38<00:14, 1.08batch/s, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 86%|████████▌ | 92/107 [01:40<00:18, 1.22s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 87%|████████▋ | 93/107 [01:41<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 88%|████████▊ | 94/107 [01:43<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 89%|████████▉ | 95/107 [01:47<00:26, 2.23s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 90%|████████▉ | 96/107 [01:48<00:20, 1.87s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 91%|█████████ | 97/107 [01:50<00:19, 1.99s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 92%|█████████▏| 98/107 [01:51<00:14, 1.63s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 93%|█████████▎| 99/107 [01:52<00:11, 1.40s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 93%|█████████▎| 100/107 [01:53<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 94%|█████████▍| 101/107 [01:54<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 95%|█████████▌| 102/107 [01:55<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 96%|█████████▋| 103/107 [01:56<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 97%|█████████▋| 104/107 [01:56<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 98%|█████████▊| 105/107 [01:57<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 99%|█████████▉| 106/107 [01:58<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 11: 100%|██████████| 107/107 [02:02<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.106, v_num=zf852wxn]\nEpoch 12: 51%|█████▏ | 55/107 [00:38<00:25, 2.03batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 12: 52%|█████▏ | 56/107 [00:45<02:07, 2.50s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 53%|█████▎ | 57/107 [00:48<02:18, 2.77s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 54%|█████▍ | 58/107 [00:50<02:06, 2.58s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 55%|█████▌ | 59/107 [00:55<02:33, 3.21s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 56%|█████▌ | 60/107 [00:56<02:03, 2.63s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 57%|█████▋ | 61/107 [00:57<01:37, 2.12s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 58%|█████▊ | 62/107 [01:00<01:41, 2.25s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 59%|█████▉ | 63/107 [01:01<01:23, 1.89s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 60%|█████▉ | 64/107 [01:02<01:17, 1.79s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 61%|██████ | 65/107 [01:03<01:04, 1.53s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 62%|██████▏ | 66/107 [01:05<01:07, 1.65s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 63%|██████▎ | 67/107 [01:07<01:03, 1.60s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 64%|██████▎ | 68/107 [01:08<00:55, 1.42s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 64%|██████▍ | 69/107 [01:09<00:50, 1.34s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 65%|██████▌ | 70/107 [01:11<00:52, 1.42s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 66%|██████▋ | 71/107 [01:12<00:53, 1.50s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 67%|██████▋ | 72/107 [01:13<00:49, 1.41s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 68%|██████▊ | 73/107 [01:14<00:40, 1.18s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 69%|██████▉ | 74/107 [01:15<00:39, 1.19s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 70%|███████ | 75/107 [01:16<00:33, 1.06s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 71%|███████ | 76/107 [01:17<00:30, 1.02batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 72%|███████▏ | 77/107 [01:18<00:29, 1.02batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 73%|███████▎ | 78/107 [01:20<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 74%|███████▍ | 79/107 [01:20<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 75%|███████▍ | 80/107 [01:22<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 76%|███████▌ | 81/107 [01:22<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 77%|███████▋ | 82/107 [01:23<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 78%|███████▊ | 83/107 [01:24<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 79%|███████▊ | 84/107 [01:25<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 79%|███████▉ | 85/107 [01:25<00:18, 1.20batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 80%|████████ | 86/107 [01:26<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 81%|████████▏ | 87/107 [01:27<00:16, 1.20batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 82%|████████▏ | 88/107 [01:28<00:15, 1.25batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 83%|████████▎ | 89/107 [01:29<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 84%|████████▍ | 90/107 [01:30<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 85%|████████▌ | 91/107 [01:31<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 86%|████████▌ | 92/107 [01:33<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 87%|████████▋ | 93/107 [01:33<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 88%|████████▊ | 94/107 [01:35<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 89%|████████▉ | 95/107 [01:40<00:26, 2.23s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 90%|████████▉ | 96/107 [01:41<00:20, 1.87s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 91%|█████████ | 97/107 [01:43<00:19, 1.99s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 92%|█████████▏| 98/107 [01:44<00:14, 1.64s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 93%|█████████▎| 99/107 [01:45<00:11, 1.41s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 93%|█████████▎| 100/107 [01:45<00:08, 1.20s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 94%|█████████▍| 101/107 [01:46<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 95%|█████████▌| 102/107 [01:47<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 96%|█████████▋| 103/107 [01:48<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 97%|█████████▋| 104/107 [01:49<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 98%|█████████▊| 105/107 [01:50<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 99%|█████████▉| 106/107 [01:51<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 12: 100%|██████████| 107/107 [01:55<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.100, v_num=zf852wxn]\nEpoch 13: 51%|█████▏ | 55/107 [00:37<00:29, 1.76batch/s, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 13: 52%|█████▏ | 56/107 [00:45<02:18, 2.72s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 53%|█████▎ | 57/107 [00:49<02:26, 2.93s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 54%|█████▍ | 58/107 [00:50<01:57, 2.41s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 55%|█████▌ | 59/107 [00:54<02:16, 2.83s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 56%|█████▌ | 60/107 [00:55<01:53, 2.41s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 57%|█████▋ | 61/107 [00:56<01:32, 2.01s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 58%|█████▊ | 62/107 [01:00<01:55, 2.56s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 59%|█████▉ | 63/107 [01:01<01:32, 2.10s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 60%|█████▉ | 64/107 [01:02<01:22, 1.92s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 61%|██████ | 65/107 [01:04<01:11, 1.71s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 62%|██████▏ | 66/107 [01:05<01:10, 1.72s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 63%|██████▎ | 67/107 [01:07<01:03, 1.59s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 64%|██████▎ | 68/107 [01:08<00:54, 1.39s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 64%|██████▍ | 69/107 [01:09<00:51, 1.35s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 65%|██████▌ | 70/107 [01:10<00:46, 1.26s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 66%|██████▋ | 71/107 [01:12<00:55, 1.55s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 67%|██████▋ | 72/107 [01:13<00:50, 1.45s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 68%|██████▊ | 73/107 [01:14<00:41, 1.21s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 69%|██████▉ | 74/107 [01:15<00:40, 1.22s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 70%|███████ | 75/107 [01:16<00:34, 1.09s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 71%|███████ | 76/107 [01:17<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 72%|███████▏ | 77/107 [01:18<00:29, 1.00batch/s, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 73%|███████▎ | 78/107 [01:20<00:36, 1.24s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 74%|███████▍ | 79/107 [01:20<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 75%|███████▍ | 80/107 [01:22<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 76%|███████▌ | 81/107 [01:22<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 77%|███████▋ | 82/107 [01:23<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 78%|███████▊ | 83/107 [01:24<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 79%|███████▊ | 84/107 [01:33<01:20, 3.49s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 79%|███████▉ | 85/107 [01:34<00:58, 2.67s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 80%|████████ | 86/107 [01:35<00:44, 2.13s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 81%|████████▏ | 87/107 [01:36<00:34, 1.72s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 82%|████████▏ | 88/107 [01:37<00:27, 1.42s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 83%|████████▎ | 89/107 [01:37<00:22, 1.28s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 84%|████████▍ | 90/107 [01:39<00:21, 1.27s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 85%|████████▌ | 91/107 [01:40<00:17, 1.12s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 86%|████████▌ | 92/107 [01:41<00:20, 1.35s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 87%|████████▋ | 93/107 [01:42<00:16, 1.17s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 88%|████████▊ | 94/107 [01:44<00:16, 1.28s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 89%|████████▉ | 95/107 [01:48<00:27, 2.30s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 90%|████████▉ | 96/107 [01:49<00:21, 1.93s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 91%|█████████ | 97/107 [01:52<00:20, 2.05s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 92%|█████████▏| 98/107 [01:53<00:15, 1.68s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 93%|█████████▎| 99/107 [01:53<00:11, 1.43s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 93%|█████████▎| 100/107 [01:54<00:08, 1.22s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 94%|█████████▍| 101/107 [01:55<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 95%|█████████▌| 102/107 [01:56<00:05, 1.15s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 96%|█████████▋| 103/107 [01:57<00:04, 1.04s/batch, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 97%|█████████▋| 104/107 [01:58<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 98%|█████████▊| 105/107 [01:59<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 99%|█████████▉| 106/107 [02:00<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 13: 100%|██████████| 107/107 [02:04<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.095, v_num=zf852wxn]\nEpoch 14: 51%|█████▏ | 55/107 [00:38<00:30, 1.69batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 14: 52%|█████▏ | 56/107 [00:46<02:21, 2.78s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 53%|█████▎ | 57/107 [00:50<02:31, 3.03s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 54%|█████▍ | 58/107 [00:51<02:04, 2.53s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 55%|█████▌ | 59/107 [00:53<01:55, 2.40s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 56%|█████▌ | 60/107 [00:55<01:41, 2.17s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 57%|█████▋ | 61/107 [00:56<01:26, 1.88s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 58%|█████▊ | 62/107 [00:58<01:28, 1.96s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 59%|█████▉ | 63/107 [01:00<01:16, 1.73s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 60%|█████▉ | 64/107 [01:01<01:14, 1.72s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 61%|██████ | 65/107 [01:02<01:04, 1.53s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 62%|██████▏ | 66/107 [01:04<01:02, 1.53s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 63%|██████▎ | 67/107 [01:05<00:56, 1.42s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 64%|██████▎ | 68/107 [01:06<00:49, 1.27s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 64%|██████▍ | 69/107 [01:07<00:49, 1.29s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 65%|██████▌ | 70/107 [01:08<00:46, 1.25s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 66%|██████▋ | 71/107 [01:10<00:50, 1.40s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 67%|██████▋ | 72/107 [01:11<00:47, 1.35s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 68%|██████▊ | 73/107 [01:13<00:45, 1.33s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 69%|██████▉ | 74/107 [01:14<00:42, 1.29s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 70%|███████ | 75/107 [01:15<00:36, 1.13s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 71%|███████ | 76/107 [01:15<00:31, 1.03s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 72%|███████▏ | 77/107 [01:16<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 73%|███████▎ | 78/107 [01:18<00:36, 1.25s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 74%|███████▍ | 79/107 [01:19<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 75%|███████▍ | 80/107 [01:20<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 76%|███████▌ | 81/107 [01:21<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 77%|███████▋ | 82/107 [01:21<00:22, 1.12batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 78%|███████▊ | 83/107 [01:22<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 79%|███████▊ | 84/107 [01:23<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 79%|███████▉ | 85/107 [01:24<00:18, 1.19batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 80%|████████ | 86/107 [01:25<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 81%|████████▏ | 87/107 [01:26<00:16, 1.20batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 82%|████████▏ | 88/107 [01:26<00:15, 1.25batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 83%|████████▎ | 89/107 [01:27<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 84%|████████▍ | 90/107 [01:29<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 85%|████████▌ | 91/107 [01:29<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 86%|████████▌ | 92/107 [01:31<00:17, 1.20s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 87%|████████▋ | 93/107 [01:32<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 88%|████████▊ | 94/107 [01:33<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 89%|████████▉ | 95/107 [01:38<00:26, 2.20s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 90%|████████▉ | 96/107 [01:39<00:20, 1.87s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 91%|█████████ | 97/107 [01:41<00:19, 1.98s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 92%|█████████▏| 98/107 [01:42<00:14, 1.63s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 93%|█████████▎| 99/107 [01:43<00:11, 1.41s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 93%|█████████▎| 100/107 [01:44<00:08, 1.21s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 94%|█████████▍| 101/107 [01:45<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 95%|█████████▌| 102/107 [01:46<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 96%|█████████▋| 103/107 [01:47<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 97%|█████████▋| 104/107 [01:47<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 98%|█████████▊| 105/107 [01:48<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 99%|█████████▉| 106/107 [01:49<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 14: 100%|██████████| 107/107 [01:53<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.091, v_num=zf852wxn]\nEpoch 15: 51%|█████▏ | 55/107 [00:39<00:26, 1.99batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 15: 52%|█████▏ | 56/107 [00:47<02:24, 2.82s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 53%|█████▎ | 57/107 [00:50<02:25, 2.90s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 54%|█████▍ | 58/107 [00:52<01:57, 2.40s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 55%|█████▌ | 59/107 [00:55<02:04, 2.60s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 56%|█████▌ | 60/107 [00:56<01:48, 2.31s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 57%|█████▋ | 61/107 [00:57<01:29, 1.95s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 58%|█████▊ | 62/107 [00:59<01:28, 1.96s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 59%|█████▉ | 63/107 [01:00<01:14, 1.69s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 60%|█████▉ | 64/107 [01:02<01:11, 1.66s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 61%|██████ | 65/107 [01:03<01:05, 1.56s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 62%|██████▏ | 66/107 [01:05<01:02, 1.53s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 63%|██████▎ | 67/107 [01:06<00:56, 1.40s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 64%|██████▎ | 68/107 [01:07<00:48, 1.25s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 64%|██████▍ | 69/107 [01:08<00:46, 1.23s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 65%|██████▌ | 70/107 [01:09<00:44, 1.19s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 66%|██████▋ | 71/107 [01:11<00:48, 1.34s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 67%|██████▋ | 72/107 [01:12<00:46, 1.33s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 68%|██████▊ | 73/107 [01:13<00:43, 1.27s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 69%|██████▉ | 74/107 [01:14<00:41, 1.25s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 70%|███████ | 75/107 [01:15<00:35, 1.11s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 71%|███████ | 76/107 [01:16<00:31, 1.02s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 72%|███████▏ | 77/107 [01:17<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 73%|███████▎ | 78/107 [01:19<00:36, 1.27s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 74%|███████▍ | 79/107 [01:20<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 75%|███████▍ | 80/107 [01:21<00:30, 1.11s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 76%|███████▌ | 81/107 [01:21<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 77%|███████▋ | 82/107 [01:22<00:23, 1.09batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 78%|███████▊ | 83/107 [01:23<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 79%|███████▊ | 84/107 [01:24<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 79%|███████▉ | 85/107 [01:25<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 80%|████████ | 86/107 [01:26<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 81%|████████▏ | 87/107 [01:26<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 82%|████████▏ | 88/107 [01:27<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 83%|████████▎ | 89/107 [01:28<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 84%|████████▍ | 90/107 [01:29<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 85%|████████▌ | 91/107 [01:30<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 86%|████████▌ | 92/107 [01:32<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 87%|████████▋ | 93/107 [01:33<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 88%|████████▊ | 94/107 [01:34<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 89%|████████▉ | 95/107 [01:39<00:26, 2.22s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 90%|████████▉ | 96/107 [01:40<00:20, 1.87s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 91%|█████████ | 97/107 [01:42<00:19, 1.99s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 92%|█████████▏| 98/107 [01:43<00:14, 1.63s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 93%|█████████▎| 99/107 [01:44<00:11, 1.40s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 93%|█████████▎| 100/107 [01:44<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 94%|█████████▍| 101/107 [01:45<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 95%|█████████▌| 102/107 [01:47<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 96%|█████████▋| 103/107 [01:47<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 97%|█████████▋| 104/107 [01:48<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 98%|█████████▊| 105/107 [01:49<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 99%|█████████▉| 106/107 [01:50<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 15: 100%|██████████| 107/107 [01:54<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.087, v_num=zf852wxn]\nEpoch 16: 51%|█████▏ | 55/107 [00:40<00:34, 1.49batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 16: 52%|█████▏ | 56/107 [00:48<02:33, 3.02s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 53%|█████▎ | 57/107 [00:53<02:54, 3.49s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 54%|█████▍ | 58/107 [00:54<02:18, 2.83s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 55%|█████▌ | 59/107 [00:58<02:31, 3.16s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 56%|█████▌ | 60/107 [01:11<04:44, 6.05s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 57%|█████▋ | 61/107 [01:12<03:29, 4.56s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 58%|█████▊ | 62/107 [01:14<02:54, 3.87s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 59%|█████▉ | 63/107 [01:15<02:16, 3.09s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 60%|█████▉ | 64/107 [01:17<01:52, 2.62s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 61%|██████ | 65/107 [01:18<01:28, 2.11s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 62%|██████▏ | 66/107 [01:20<01:26, 2.11s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 63%|██████▎ | 67/107 [01:21<01:15, 1.89s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 64%|██████▎ | 68/107 [01:22<01:01, 1.59s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 64%|██████▍ | 69/107 [01:23<00:56, 1.47s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 65%|██████▌ | 70/107 [01:25<00:50, 1.36s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 66%|██████▋ | 71/107 [01:26<00:52, 1.45s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 67%|██████▋ | 72/107 [01:27<00:48, 1.39s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 68%|██████▊ | 73/107 [01:28<00:39, 1.16s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 69%|██████▉ | 74/107 [01:29<00:38, 1.18s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 70%|███████ | 75/107 [01:30<00:33, 1.06s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 71%|███████ | 76/107 [01:31<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 72%|███████▏ | 77/107 [01:32<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 73%|███████▎ | 78/107 [01:34<00:36, 1.26s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 74%|███████▍ | 79/107 [01:35<00:31, 1.11s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 75%|███████▍ | 80/107 [01:36<00:30, 1.11s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 76%|███████▌ | 81/107 [01:36<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 77%|███████▋ | 82/107 [01:37<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 78%|███████▊ | 83/107 [01:38<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 79%|███████▊ | 84/107 [01:39<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 79%|███████▉ | 85/107 [01:40<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 80%|████████ | 86/107 [01:41<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 81%|████████▏ | 87/107 [01:41<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 82%|████████▏ | 88/107 [01:42<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 83%|████████▎ | 89/107 [01:43<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 84%|████████▍ | 90/107 [01:44<00:16, 1.02batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 85%|████████▌ | 91/107 [01:45<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 86%|████████▌ | 92/107 [01:47<00:18, 1.22s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 87%|████████▋ | 93/107 [01:48<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 88%|████████▊ | 94/107 [01:49<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 89%|████████▉ | 95/107 [01:54<00:26, 2.25s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 90%|████████▉ | 96/107 [01:55<00:20, 1.89s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 91%|█████████ | 97/107 [01:57<00:20, 2.01s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 92%|█████████▏| 98/107 [01:58<00:14, 1.66s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 93%|█████████▎| 99/107 [01:59<00:11, 1.42s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 93%|█████████▎| 100/107 [02:00<00:08, 1.21s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 94%|█████████▍| 101/107 [02:00<00:06, 1.06s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 95%|█████████▌| 102/107 [02:02<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 96%|█████████▋| 103/107 [02:03<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 97%|█████████▋| 104/107 [02:03<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 98%|█████████▊| 105/107 [02:04<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 99%|█████████▉| 106/107 [02:05<00:00, 1.00batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 16: 100%|██████████| 107/107 [02:09<00:00, 1.10batch/s, batch_idx=54, gpu=0, loss=0.083, v_num=zf852wxn]\nEpoch 17: 51%|█████▏ | 55/107 [00:39<00:25, 2.02batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 17: 52%|█████▏ | 56/107 [00:48<02:29, 2.93s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 53%|█████▎ | 57/107 [00:51<02:38, 3.17s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 54%|█████▍ | 58/107 [00:53<02:07, 2.59s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 55%|█████▌ | 59/107 [00:55<02:03, 2.57s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 56%|█████▌ | 60/107 [00:57<01:47, 2.28s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 57%|█████▋ | 61/107 [00:58<01:29, 1.95s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 58%|█████▊ | 62/107 [01:00<01:30, 2.02s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 59%|█████▉ | 63/107 [01:02<01:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 60%|█████▉ | 64/107 [01:03<01:15, 1.75s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 61%|██████ | 65/107 [01:04<01:07, 1.61s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 62%|██████▏ | 66/107 [01:06<01:01, 1.50s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 63%|██████▎ | 67/107 [01:07<00:56, 1.40s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 64%|██████▎ | 68/107 [01:08<00:49, 1.26s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 64%|██████▍ | 69/107 [01:09<00:46, 1.23s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 65%|██████▌ | 70/107 [01:10<00:45, 1.22s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 66%|██████▋ | 71/107 [01:12<00:49, 1.38s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 67%|██████▋ | 72/107 [01:13<00:47, 1.35s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 68%|██████▊ | 73/107 [01:14<00:39, 1.15s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 69%|██████▉ | 74/107 [01:16<00:43, 1.32s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 70%|███████ | 75/107 [01:16<00:37, 1.16s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 71%|███████ | 76/107 [01:17<00:32, 1.05s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 72%|███████▏ | 77/107 [01:18<00:31, 1.04s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 73%|███████▎ | 78/107 [01:20<00:36, 1.27s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 74%|███████▍ | 79/107 [01:21<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 75%|███████▍ | 80/107 [01:22<00:29, 1.11s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 76%|███████▌ | 81/107 [01:22<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 77%|███████▋ | 82/107 [01:23<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 78%|███████▊ | 83/107 [01:24<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 79%|███████▊ | 84/107 [01:25<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 79%|███████▉ | 85/107 [01:26<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 80%|████████ | 86/107 [01:27<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 81%|████████▏ | 87/107 [01:27<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 82%|████████▏ | 88/107 [01:28<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 83%|████████▎ | 89/107 [01:29<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 84%|████████▍ | 90/107 [01:30<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 85%|████████▌ | 91/107 [01:31<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 86%|████████▌ | 92/107 [01:33<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 87%|████████▋ | 93/107 [01:34<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 88%|████████▊ | 94/107 [01:35<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 89%|████████▉ | 95/107 [01:40<00:26, 2.20s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 90%|████████▉ | 96/107 [01:41<00:20, 1.86s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 91%|█████████ | 97/107 [01:43<00:19, 1.98s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 92%|█████████▏| 98/107 [01:44<00:14, 1.63s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 93%|█████████▎| 99/107 [01:45<00:11, 1.41s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 93%|█████████▎| 100/107 [01:46<00:08, 1.20s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 94%|█████████▍| 101/107 [01:46<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 95%|█████████▌| 102/107 [01:48<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 96%|█████████▋| 103/107 [01:48<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 97%|█████████▋| 104/107 [01:49<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 98%|█████████▊| 105/107 [01:50<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 99%|█████████▉| 106/107 [01:51<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 17: 100%|██████████| 107/107 [01:55<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.071, v_num=zf852wxn]\nEpoch 18: 51%|█████▏ | 55/107 [00:40<00:40, 1.28batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 18: 52%|█████▏ | 56/107 [00:49<02:41, 3.17s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 53%|█████▎ | 57/107 [00:52<02:42, 3.25s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 54%|█████▍ | 58/107 [00:53<02:09, 2.64s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 55%|█████▌ | 59/107 [00:56<02:13, 2.77s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 56%|█████▌ | 60/107 [00:58<01:56, 2.47s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 57%|█████▋ | 61/107 [00:59<01:35, 2.09s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 58%|█████▊ | 62/107 [01:01<01:32, 2.05s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 59%|█████▉ | 63/107 [01:02<01:17, 1.76s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 60%|█████▉ | 64/107 [01:04<01:11, 1.67s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 61%|██████ | 65/107 [01:05<01:04, 1.53s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 62%|██████▏ | 66/107 [01:07<01:06, 1.62s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 63%|██████▎ | 67/107 [01:08<00:57, 1.44s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 64%|██████▎ | 68/107 [01:09<00:49, 1.27s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 64%|██████▍ | 69/107 [01:10<00:48, 1.26s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 65%|██████▌ | 70/107 [01:11<00:44, 1.20s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 66%|██████▋ | 71/107 [01:13<00:54, 1.52s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 67%|██████▋ | 72/107 [01:15<00:50, 1.43s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 68%|██████▊ | 73/107 [01:15<00:41, 1.21s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 69%|██████▉ | 74/107 [01:17<00:40, 1.21s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 70%|███████ | 75/107 [01:17<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 71%|███████ | 76/107 [01:18<00:30, 1.00batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 72%|███████▏ | 77/107 [01:19<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 73%|███████▎ | 78/107 [01:21<00:35, 1.24s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 74%|███████▍ | 79/107 [01:22<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 75%|███████▍ | 80/107 [01:23<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 76%|███████▌ | 81/107 [01:23<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 77%|███████▋ | 82/107 [01:24<00:22, 1.12batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 78%|███████▊ | 83/107 [01:25<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 79%|███████▊ | 84/107 [01:26<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 79%|███████▉ | 85/107 [01:27<00:18, 1.19batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 80%|████████ | 86/107 [01:28<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 81%|████████▏ | 87/107 [01:28<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 82%|████████▏ | 88/107 [01:29<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 83%|████████▎ | 89/107 [01:30<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 84%|████████▍ | 90/107 [01:31<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 85%|████████▌ | 91/107 [01:32<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 86%|████████▌ | 92/107 [01:34<00:17, 1.20s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 87%|████████▋ | 93/107 [01:35<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 88%|████████▊ | 94/107 [01:36<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 89%|████████▉ | 95/107 [01:41<00:26, 2.22s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 90%|████████▉ | 96/107 [01:42<00:20, 1.87s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 91%|█████████ | 97/107 [01:44<00:19, 2.00s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 92%|█████████▏| 98/107 [01:45<00:14, 1.64s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 93%|█████████▎| 99/107 [01:46<00:11, 1.41s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 93%|█████████▎| 100/107 [01:47<00:08, 1.20s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 94%|█████████▍| 101/107 [01:47<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 95%|█████████▌| 102/107 [01:49<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 96%|█████████▋| 103/107 [01:49<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 97%|█████████▋| 104/107 [01:50<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 98%|█████████▊| 105/107 [01:51<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 99%|█████████▉| 106/107 [01:52<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 18: 100%|██████████| 107/107 [01:56<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.046, v_num=zf852wxn]\nEpoch 19: 51%|█████▏ | 55/107 [00:39<00:17, 2.92batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 19: 52%|█████▏ | 56/107 [00:48<02:29, 2.93s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 53%|█████▎ | 57/107 [00:53<02:54, 3.48s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 54%|█████▍ | 58/107 [00:55<02:19, 2.84s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 55%|█████▌ | 59/107 [00:57<02:15, 2.83s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 56%|█████▌ | 60/107 [01:11<04:49, 6.15s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 57%|█████▋ | 61/107 [01:12<03:33, 4.64s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 58%|█████▊ | 62/107 [01:14<02:54, 3.88s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 59%|█████▉ | 63/107 [01:15<02:12, 3.02s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 60%|█████▉ | 64/107 [01:17<01:50, 2.57s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 61%|██████ | 65/107 [01:18<01:26, 2.06s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 62%|██████▏ | 66/107 [01:20<01:24, 2.07s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 63%|██████▎ | 67/107 [01:21<01:15, 1.88s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 64%|██████▎ | 68/107 [01:22<01:02, 1.61s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 64%|██████▍ | 69/107 [01:24<00:55, 1.47s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 65%|██████▌ | 70/107 [01:25<00:49, 1.34s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 66%|██████▋ | 71/107 [01:26<00:51, 1.44s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 67%|██████▋ | 72/107 [01:27<00:48, 1.38s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 68%|██████▊ | 73/107 [01:28<00:39, 1.16s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 69%|██████▉ | 74/107 [01:29<00:38, 1.18s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 70%|███████ | 75/107 [01:30<00:33, 1.05s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 71%|███████ | 76/107 [01:31<00:30, 1.02batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 72%|███████▏ | 77/107 [01:32<00:29, 1.02batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 73%|███████▎ | 78/107 [01:34<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 74%|███████▍ | 79/107 [01:34<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 75%|███████▍ | 80/107 [01:36<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 76%|███████▌ | 81/107 [01:36<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 77%|███████▋ | 82/107 [01:37<00:22, 1.12batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 78%|███████▊ | 83/107 [01:38<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 79%|███████▊ | 84/107 [01:39<00:20, 1.15batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 79%|███████▉ | 85/107 [01:39<00:18, 1.19batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 80%|████████ | 86/107 [01:40<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 81%|████████▏ | 87/107 [01:41<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 82%|████████▏ | 88/107 [01:42<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 83%|████████▎ | 89/107 [01:43<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 84%|████████▍ | 90/107 [01:44<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 85%|████████▌ | 91/107 [01:45<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 86%|████████▌ | 92/107 [01:47<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 87%|████████▋ | 93/107 [01:48<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 88%|████████▊ | 94/107 [01:49<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 89%|████████▉ | 95/107 [01:54<00:26, 2.25s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 90%|████████▉ | 96/107 [01:55<00:20, 1.89s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 91%|█████████ | 97/107 [01:57<00:20, 2.00s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 92%|█████████▏| 98/107 [01:58<00:14, 1.64s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 93%|█████████▎| 99/107 [01:59<00:11, 1.41s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 93%|█████████▎| 100/107 [01:59<00:08, 1.20s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 94%|█████████▍| 101/107 [02:00<00:06, 1.06s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 95%|█████████▌| 102/107 [02:01<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 96%|█████████▋| 103/107 [02:02<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 97%|█████████▋| 104/107 [02:03<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 98%|█████████▊| 105/107 [02:04<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 99%|█████████▉| 106/107 [02:05<00:00, 1.00batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 19: 100%|██████████| 107/107 [02:09<00:00, 1.10batch/s, batch_idx=54, gpu=0, loss=0.036, v_num=zf852wxn]\nEpoch 20: 51%|█████▏ | 55/107 [00:40<00:20, 2.51batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 20: 52%|█████▏ | 56/107 [00:50<02:47, 3.29s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 53%|█████▎ | 57/107 [00:54<02:55, 3.50s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 54%|█████▍ | 58/107 [00:56<02:24, 2.94s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 55%|█████▌ | 59/107 [00:58<02:05, 2.61s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 56%|█████▌ | 60/107 [01:00<01:49, 2.33s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 57%|█████▋ | 61/107 [01:01<01:31, 1.99s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 58%|█████▊ | 62/107 [01:03<01:34, 2.09s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 59%|█████▉ | 63/107 [01:04<01:20, 1.83s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 60%|█████▉ | 64/107 [01:06<01:21, 1.91s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 61%|██████ | 65/107 [01:08<01:17, 1.85s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 62%|██████▏ | 66/107 [01:10<01:16, 1.87s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 63%|██████▎ | 67/107 [01:11<01:07, 1.69s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 64%|██████▎ | 68/107 [01:12<00:57, 1.49s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 64%|██████▍ | 69/107 [01:14<00:55, 1.47s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 65%|██████▌ | 70/107 [01:15<00:50, 1.37s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 66%|██████▋ | 71/107 [01:17<00:53, 1.50s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 67%|██████▋ | 72/107 [01:19<00:56, 1.61s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 68%|██████▊ | 73/107 [01:19<00:45, 1.33s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 69%|██████▉ | 74/107 [01:20<00:43, 1.31s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 70%|███████ | 75/107 [01:21<00:36, 1.15s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 71%|███████ | 76/107 [01:22<00:32, 1.06s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 72%|███████▏ | 77/107 [01:23<00:31, 1.05s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 73%|███████▎ | 78/107 [01:25<00:37, 1.29s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 74%|███████▍ | 79/107 [01:26<00:31, 1.13s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 75%|███████▍ | 80/107 [01:27<00:30, 1.12s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 76%|███████▌ | 81/107 [01:27<00:25, 1.03batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 77%|███████▋ | 82/107 [01:28<00:23, 1.09batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 78%|███████▊ | 83/107 [01:29<00:23, 1.03batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 79%|███████▊ | 84/107 [01:30<00:20, 1.11batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 79%|███████▉ | 85/107 [01:31<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 80%|████████ | 86/107 [01:32<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 81%|████████▏ | 87/107 [01:32<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 82%|████████▏ | 88/107 [01:33<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 83%|████████▎ | 89/107 [01:34<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 84%|████████▍ | 90/107 [01:35<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 85%|████████▌ | 91/107 [01:36<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 86%|████████▌ | 92/107 [01:38<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 87%|████████▋ | 93/107 [01:39<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 88%|████████▊ | 94/107 [01:40<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 89%|████████▉ | 95/107 [01:45<00:26, 2.20s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 90%|████████▉ | 96/107 [01:46<00:20, 1.85s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 91%|█████████ | 97/107 [01:48<00:19, 1.97s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 92%|█████████▏| 98/107 [01:49<00:14, 1.62s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 93%|█████████▎| 99/107 [01:50<00:11, 1.40s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 93%|█████████▎| 100/107 [01:51<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 94%|█████████▍| 101/107 [01:51<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 95%|█████████▌| 102/107 [01:53<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 96%|█████████▋| 103/107 [01:53<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 97%|█████████▋| 104/107 [01:54<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 98%|█████████▊| 105/107 [01:55<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 99%|█████████▉| 106/107 [01:56<00:01, 1.00s/batch, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 20: 100%|██████████| 107/107 [02:00<00:00, 1.10batch/s, batch_idx=54, gpu=0, loss=0.035, v_num=zf852wxn]\nEpoch 21: 51%|█████▏ | 55/107 [00:41<00:27, 1.92batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 21: 52%|█████▏ | 56/107 [00:51<02:44, 3.23s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 53%|█████▎ | 57/107 [00:55<02:51, 3.42s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 54%|█████▍ | 58/107 [00:56<02:21, 2.88s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 55%|█████▌ | 59/107 [01:01<02:36, 3.27s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 56%|█████▌ | 60/107 [01:02<02:09, 2.76s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 57%|█████▋ | 61/107 [01:03<01:43, 2.24s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 58%|█████▊ | 62/107 [01:07<02:06, 2.81s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 59%|█████▉ | 63/107 [01:08<01:41, 2.31s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 60%|█████▉ | 64/107 [01:10<01:27, 2.03s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 61%|██████ | 65/107 [01:11<01:13, 1.74s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 62%|██████▏ | 66/107 [01:13<01:13, 1.80s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 63%|██████▎ | 67/107 [01:14<01:03, 1.58s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 64%|██████▎ | 68/107 [01:15<00:53, 1.38s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 64%|██████▍ | 69/107 [01:16<00:51, 1.36s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 65%|██████▌ | 70/107 [01:17<00:47, 1.28s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 66%|██████▋ | 71/107 [01:19<00:57, 1.60s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 67%|██████▋ | 72/107 [01:21<00:51, 1.48s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 68%|██████▊ | 73/107 [01:21<00:41, 1.23s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 69%|██████▉ | 74/107 [01:23<00:40, 1.22s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 70%|███████ | 75/107 [01:23<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 71%|███████ | 76/107 [01:24<00:30, 1.00batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 72%|███████▏ | 77/107 [01:25<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 73%|███████▎ | 78/107 [01:27<00:35, 1.24s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 74%|███████▍ | 79/107 [01:28<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 75%|███████▍ | 80/107 [01:29<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 76%|███████▌ | 81/107 [01:29<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 77%|███████▋ | 82/107 [01:30<00:22, 1.12batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 78%|███████▊ | 83/107 [01:31<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 79%|███████▊ | 84/107 [01:32<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 79%|███████▉ | 85/107 [01:33<00:18, 1.19batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 80%|████████ | 86/107 [01:34<00:17, 1.18batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 81%|████████▏ | 87/107 [01:34<00:16, 1.20batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 82%|████████▏ | 88/107 [01:35<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 83%|████████▎ | 89/107 [01:36<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 84%|████████▍ | 90/107 [01:37<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 85%|████████▌ | 91/107 [01:38<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 86%|████████▌ | 92/107 [01:40<00:18, 1.23s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 87%|████████▋ | 93/107 [01:41<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 88%|████████▊ | 94/107 [01:42<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 89%|████████▉ | 95/107 [01:47<00:27, 2.33s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 90%|████████▉ | 96/107 [01:48<00:21, 1.96s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 91%|█████████ | 97/107 [01:51<00:20, 2.04s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 92%|█████████▏| 98/107 [01:51<00:15, 1.67s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 93%|█████████▎| 99/107 [01:52<00:11, 1.46s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 93%|█████████▎| 100/107 [01:53<00:08, 1.23s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 94%|█████████▍| 101/107 [01:54<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 95%|█████████▌| 102/107 [01:55<00:05, 1.15s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 96%|█████████▋| 103/107 [01:56<00:04, 1.05s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 97%|█████████▋| 104/107 [01:57<00:02, 1.04batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 98%|█████████▊| 105/107 [01:57<00:01, 1.07batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 99%|█████████▉| 106/107 [01:59<00:01, 1.12s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 21: 100%|██████████| 107/107 [02:04<00:00, 1.19s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 51%|█████▏ | 55/107 [00:41<00:25, 2.06batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 22: 52%|█████▏ | 56/107 [00:51<02:49, 3.32s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 53%|█████▎ | 57/107 [00:54<02:47, 3.36s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 54%|█████▍ | 58/107 [00:56<02:16, 2.78s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 55%|█████▌ | 59/107 [00:58<02:12, 2.77s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 56%|█████▌ | 60/107 [01:00<01:54, 2.45s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 57%|█████▋ | 61/107 [01:01<01:35, 2.07s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 58%|█████▊ | 62/107 [01:03<01:32, 2.06s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 59%|█████▉ | 63/107 [01:04<01:17, 1.77s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 60%|█████▉ | 64/107 [01:06<01:11, 1.67s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 61%|██████ | 65/107 [01:07<01:03, 1.52s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 62%|██████▏ | 66/107 [01:09<01:06, 1.62s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 63%|██████▎ | 67/107 [01:10<00:57, 1.44s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 64%|██████▎ | 68/107 [01:11<00:50, 1.28s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 64%|██████▍ | 69/107 [01:12<00:48, 1.28s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 65%|██████▌ | 70/107 [01:13<00:45, 1.22s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 66%|██████▋ | 71/107 [01:15<00:50, 1.40s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 67%|██████▋ | 72/107 [01:16<00:47, 1.35s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 68%|██████▊ | 73/107 [01:17<00:39, 1.15s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 69%|██████▉ | 74/107 [01:19<00:44, 1.36s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 70%|███████ | 75/107 [01:20<00:38, 1.19s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 71%|███████ | 76/107 [01:20<00:33, 1.08s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 72%|███████▏ | 77/107 [01:21<00:31, 1.05s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 73%|███████▎ | 78/107 [01:23<00:37, 1.29s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 74%|███████▍ | 79/107 [01:24<00:31, 1.13s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 75%|███████▍ | 80/107 [01:25<00:30, 1.12s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 76%|███████▌ | 81/107 [01:26<00:25, 1.04batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 77%|███████▋ | 82/107 [01:27<00:23, 1.08batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 78%|███████▊ | 83/107 [01:28<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 79%|███████▊ | 84/107 [01:28<00:20, 1.11batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 79%|███████▉ | 85/107 [01:29<00:19, 1.16batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 80%|████████ | 86/107 [01:30<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 81%|████████▏ | 87/107 [01:46<01:48, 5.41s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 82%|████████▏ | 88/107 [01:47<01:16, 4.00s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 83%|████████▎ | 89/107 [01:48<00:55, 3.08s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 84%|████████▍ | 90/107 [01:49<00:43, 2.53s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 85%|████████▌ | 91/107 [01:50<00:32, 2.00s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 86%|████████▌ | 92/107 [01:52<00:29, 1.95s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 87%|████████▋ | 93/107 [01:52<00:22, 1.61s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 88%|████████▊ | 94/107 [01:54<00:20, 1.58s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 89%|████████▉ | 95/107 [01:59<00:30, 2.51s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 90%|████████▉ | 96/107 [02:00<00:22, 2.07s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 91%|█████████ | 97/107 [02:02<00:21, 2.12s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 92%|█████████▏| 98/107 [02:03<00:15, 1.72s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 93%|█████████▎| 99/107 [02:03<00:11, 1.46s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 93%|█████████▎| 100/107 [02:04<00:08, 1.24s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 94%|█████████▍| 101/107 [02:05<00:06, 1.08s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 95%|█████████▌| 102/107 [02:06<00:05, 1.16s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 96%|█████████▋| 103/107 [02:07<00:04, 1.04s/batch, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 97%|█████████▋| 104/107 [02:08<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 98%|█████████▊| 105/107 [02:09<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 99%|█████████▉| 106/107 [02:10<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 22: 100%|██████████| 107/107 [02:14<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.034, v_num=zf852wxn]\nEpoch 23: 51%|█████▏ | 55/107 [00:41<00:23, 2.23batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 23: 52%|█████▏ | 56/107 [00:52<02:57, 3.48s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 53%|█████▎ | 57/107 [00:55<02:46, 3.34s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 54%|█████▍ | 58/107 [00:57<02:16, 2.79s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 55%|█████▌ | 59/107 [00:59<02:08, 2.68s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 56%|█████▌ | 60/107 [01:00<01:48, 2.32s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 57%|█████▋ | 61/107 [01:02<01:29, 1.96s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 58%|█████▊ | 62/107 [01:04<01:29, 1.99s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 59%|█████▉ | 63/107 [01:05<01:14, 1.69s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 60%|█████▉ | 64/107 [01:06<01:09, 1.61s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 61%|██████ | 65/107 [01:07<01:02, 1.49s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 62%|██████▏ | 66/107 [01:09<01:05, 1.59s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 63%|██████▎ | 67/107 [01:10<00:57, 1.43s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 64%|██████▎ | 68/107 [01:11<00:49, 1.28s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 64%|██████▍ | 69/107 [01:12<00:48, 1.27s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 65%|██████▌ | 70/107 [01:13<00:44, 1.20s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 66%|██████▋ | 71/107 [01:15<00:48, 1.34s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 67%|██████▋ | 72/107 [01:16<00:45, 1.30s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 68%|██████▊ | 73/107 [01:18<00:44, 1.32s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 69%|██████▉ | 74/107 [01:19<00:42, 1.28s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 70%|███████ | 75/107 [01:20<00:36, 1.13s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 71%|███████ | 76/107 [01:20<00:32, 1.04s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 72%|███████▏ | 77/107 [01:21<00:30, 1.02s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 73%|███████▎ | 78/107 [01:23<00:36, 1.26s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 74%|███████▍ | 79/107 [01:24<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 75%|███████▍ | 80/107 [01:25<00:29, 1.11s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 76%|███████▌ | 81/107 [01:26<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 77%|███████▋ | 82/107 [01:26<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 78%|███████▊ | 83/107 [01:27<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 79%|███████▊ | 84/107 [01:28<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 79%|███████▉ | 85/107 [01:29<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 80%|████████ | 86/107 [01:30<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 81%|████████▏ | 87/107 [01:31<00:17, 1.18batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 82%|████████▏ | 88/107 [01:31<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 83%|████████▎ | 89/107 [01:32<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 84%|████████▍ | 90/107 [01:34<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 85%|████████▌ | 91/107 [01:34<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 86%|████████▌ | 92/107 [01:36<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 87%|████████▋ | 93/107 [01:37<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 88%|████████▊ | 94/107 [01:39<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 89%|████████▉ | 95/107 [01:43<00:26, 2.24s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 90%|████████▉ | 96/107 [01:44<00:20, 1.89s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 91%|█████████ | 97/107 [01:47<00:20, 2.01s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 92%|█████████▏| 98/107 [01:47<00:14, 1.65s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 93%|█████████▎| 99/107 [01:48<00:11, 1.42s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 93%|█████████▎| 100/107 [01:49<00:08, 1.23s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 94%|█████████▍| 101/107 [01:50<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 95%|█████████▌| 102/107 [01:51<00:05, 1.15s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 96%|█████████▋| 103/107 [01:52<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 97%|█████████▋| 104/107 [01:53<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 98%|█████████▊| 105/107 [01:53<00:01, 1.08batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 99%|█████████▉| 106/107 [01:55<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 23: 100%|██████████| 107/107 [01:59<00:00, 1.10batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 51%|█████▏ | 55/107 [00:43<00:28, 1.81batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 24: 52%|█████▏ | 56/107 [00:53<02:59, 3.52s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 53%|█████▎ | 57/107 [00:57<03:01, 3.64s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 54%|█████▍ | 58/107 [00:58<02:25, 2.97s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 55%|█████▌ | 59/107 [01:00<02:07, 2.66s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 56%|█████▌ | 60/107 [01:02<01:49, 2.34s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 57%|█████▋ | 61/107 [01:03<01:30, 1.97s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 58%|█████▊ | 62/107 [01:05<01:33, 2.08s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 59%|█████▉ | 63/107 [01:07<01:18, 1.79s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 60%|█████▉ | 64/107 [01:08<01:12, 1.70s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 61%|██████ | 65/107 [01:09<01:03, 1.52s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 62%|██████▏ | 66/107 [01:11<01:00, 1.49s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 63%|██████▎ | 67/107 [01:12<00:55, 1.38s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 64%|██████▎ | 68/107 [01:13<00:48, 1.24s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 64%|██████▍ | 69/107 [01:14<00:49, 1.30s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 65%|██████▌ | 70/107 [01:15<00:46, 1.25s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 66%|██████▋ | 71/107 [01:17<00:50, 1.39s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 67%|██████▋ | 72/107 [01:19<00:55, 1.58s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 68%|██████▊ | 73/107 [01:21<00:58, 1.72s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 69%|██████▉ | 74/107 [01:22<00:51, 1.57s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 70%|███████ | 75/107 [01:23<00:42, 1.33s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 71%|███████ | 76/107 [01:24<00:36, 1.18s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 72%|███████▏ | 77/107 [01:25<00:33, 1.12s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 73%|███████▎ | 78/107 [01:27<00:39, 1.35s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 74%|███████▍ | 79/107 [01:27<00:32, 1.16s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 75%|███████▍ | 80/107 [01:28<00:30, 1.14s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 76%|███████▌ | 81/107 [01:29<00:25, 1.02batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 77%|███████▋ | 82/107 [01:30<00:23, 1.08batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 78%|███████▊ | 83/107 [01:31<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 79%|███████▊ | 84/107 [01:32<00:20, 1.11batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 79%|███████▉ | 85/107 [01:32<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 80%|████████ | 86/107 [01:33<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 81%|████████▏ | 87/107 [01:34<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 82%|████████▏ | 88/107 [01:35<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 83%|████████▎ | 89/107 [01:36<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 84%|████████▍ | 90/107 [01:37<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 85%|████████▌ | 91/107 [01:38<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 86%|████████▌ | 92/107 [01:40<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 87%|████████▋ | 93/107 [01:40<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 88%|████████▊ | 94/107 [01:42<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 89%|████████▉ | 95/107 [01:47<00:27, 2.32s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 90%|████████▉ | 96/107 [01:48<00:21, 1.95s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 91%|█████████ | 97/107 [01:50<00:20, 2.05s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 92%|█████████▏| 98/107 [01:51<00:15, 1.68s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 93%|█████████▎| 99/107 [01:52<00:11, 1.43s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 93%|█████████▎| 100/107 [01:53<00:08, 1.23s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 94%|█████████▍| 101/107 [01:53<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 95%|█████████▌| 102/107 [01:55<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 96%|█████████▋| 103/107 [01:55<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 97%|█████████▋| 104/107 [01:56<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 98%|█████████▊| 105/107 [01:57<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 99%|█████████▉| 106/107 [01:58<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 24: 100%|██████████| 107/107 [02:02<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.033, v_num=zf852wxn]\nEpoch 25: 51%|█████▏ | 55/107 [00:42<00:28, 1.85batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 25: 52%|█████▏ | 56/107 [00:53<03:06, 3.66s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 53%|█████▎ | 57/107 [00:56<02:48, 3.36s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 54%|█████▍ | 58/107 [00:57<02:15, 2.77s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 55%|█████▌ | 59/107 [01:01<02:21, 2.95s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 56%|█████▌ | 60/107 [01:02<01:58, 2.51s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 57%|█████▋ | 61/107 [01:03<01:32, 2.01s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 58%|█████▊ | 62/107 [01:08<02:09, 2.87s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 59%|█████▉ | 63/107 [01:09<01:42, 2.34s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 60%|█████▉ | 64/107 [01:11<01:31, 2.13s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 61%|██████ | 65/107 [01:12<01:15, 1.80s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 62%|██████▏ | 66/107 [01:14<01:16, 1.87s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 63%|██████▎ | 67/107 [01:15<01:04, 1.62s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 64%|██████▎ | 68/107 [01:16<00:54, 1.40s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 64%|██████▍ | 69/107 [01:17<00:51, 1.36s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 65%|██████▌ | 70/107 [01:19<00:53, 1.44s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 66%|██████▋ | 71/107 [01:20<00:54, 1.51s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 67%|██████▋ | 72/107 [01:21<00:49, 1.43s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 68%|██████▊ | 73/107 [01:22<00:40, 1.19s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 69%|██████▉ | 74/107 [01:23<00:39, 1.19s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 70%|███████ | 75/107 [01:24<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 71%|███████ | 76/107 [01:25<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 72%|███████▏ | 77/107 [01:26<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 73%|███████▎ | 78/107 [01:28<00:36, 1.25s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 74%|███████▍ | 79/107 [01:29<00:31, 1.11s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 75%|███████▍ | 80/107 [01:30<00:29, 1.11s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 76%|███████▌ | 81/107 [01:30<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 77%|███████▋ | 82/107 [01:31<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 78%|███████▊ | 83/107 [01:32<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 79%|███████▊ | 84/107 [01:33<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 79%|███████▉ | 85/107 [01:34<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 80%|████████ | 86/107 [01:34<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 81%|████████▏ | 87/107 [01:35<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 82%|████████▏ | 88/107 [01:36<00:15, 1.25batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 83%|████████▎ | 89/107 [01:37<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 84%|████████▍ | 90/107 [01:38<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 85%|████████▌ | 91/107 [01:39<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 86%|████████▌ | 92/107 [01:41<00:18, 1.20s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 87%|████████▋ | 93/107 [01:42<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 88%|████████▊ | 94/107 [01:43<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 89%|████████▉ | 95/107 [01:48<00:27, 2.26s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 90%|████████▉ | 96/107 [01:49<00:20, 1.90s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 91%|█████████ | 97/107 [01:51<00:20, 2.02s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 92%|█████████▏| 98/107 [01:52<00:14, 1.65s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 93%|█████████▎| 99/107 [01:53<00:11, 1.41s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 93%|█████████▎| 100/107 [01:53<00:08, 1.20s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 94%|█████████▍| 101/107 [01:54<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 95%|█████████▌| 102/107 [01:56<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 96%|█████████▋| 103/107 [01:56<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 97%|█████████▋| 104/107 [01:57<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 98%|█████████▊| 105/107 [01:58<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 99%|█████████▉| 106/107 [01:59<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 25: 100%|██████████| 107/107 [02:03<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 51%|█████▏ | 55/107 [00:43<00:27, 1.89batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 26: 52%|█████▏ | 56/107 [00:54<03:06, 3.67s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 53%|█████▎ | 57/107 [00:57<02:57, 3.54s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 54%|█████▍ | 58/107 [00:59<02:21, 2.88s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 55%|█████▌ | 59/107 [01:01<02:16, 2.84s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 56%|█████▌ | 60/107 [01:03<01:54, 2.43s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 57%|█████▋ | 61/107 [01:04<01:34, 2.05s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 58%|█████▊ | 62/107 [01:06<01:31, 2.03s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 59%|█████▉ | 63/107 [01:07<01:17, 1.76s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 60%|█████▉ | 64/107 [01:09<01:14, 1.74s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 61%|██████ | 65/107 [01:10<01:02, 1.49s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 62%|██████▏ | 66/107 [01:12<01:05, 1.59s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 63%|██████▎ | 67/107 [01:13<00:59, 1.48s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 64%|██████▎ | 68/107 [01:14<00:51, 1.32s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 64%|██████▍ | 69/107 [01:15<00:49, 1.30s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 65%|██████▌ | 70/107 [01:16<00:45, 1.22s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 66%|██████▋ | 71/107 [01:18<00:48, 1.35s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 67%|██████▋ | 72/107 [01:19<00:45, 1.31s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 68%|██████▊ | 73/107 [01:20<00:37, 1.12s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 69%|██████▉ | 74/107 [01:21<00:44, 1.35s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 70%|███████ | 75/107 [01:22<00:37, 1.18s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 71%|███████ | 76/107 [01:23<00:33, 1.07s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 72%|███████▏ | 77/107 [01:24<00:31, 1.05s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 73%|███████▎ | 78/107 [01:26<00:37, 1.30s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 74%|███████▍ | 79/107 [01:27<00:31, 1.13s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 75%|███████▍ | 80/107 [01:28<00:30, 1.12s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 76%|███████▌ | 81/107 [01:28<00:25, 1.03batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 77%|███████▋ | 82/107 [01:29<00:23, 1.08batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 78%|███████▊ | 83/107 [01:30<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 79%|███████▊ | 84/107 [01:31<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 79%|███████▉ | 85/107 [01:32<00:19, 1.16batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 80%|████████ | 86/107 [01:33<00:18, 1.14batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 81%|████████▏ | 87/107 [01:33<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 82%|████████▏ | 88/107 [01:34<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 83%|████████▎ | 89/107 [01:35<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 84%|████████▍ | 90/107 [01:36<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 85%|████████▌ | 91/107 [01:37<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 86%|████████▌ | 92/107 [01:39<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 87%|████████▋ | 93/107 [01:40<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 88%|████████▊ | 94/107 [01:41<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 89%|████████▉ | 95/107 [01:46<00:27, 2.26s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 90%|████████▉ | 96/107 [01:47<00:20, 1.90s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 91%|█████████ | 97/107 [01:49<00:20, 2.02s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 92%|█████████▏| 98/107 [01:50<00:14, 1.65s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 93%|█████████▎| 99/107 [01:51<00:11, 1.42s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 93%|█████████▎| 100/107 [01:52<00:08, 1.21s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 94%|█████████▍| 101/107 [01:53<00:06, 1.06s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 95%|█████████▌| 102/107 [01:54<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 96%|█████████▋| 103/107 [01:55<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 97%|█████████▋| 104/107 [01:55<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 98%|█████████▊| 105/107 [02:15<00:13, 6.53s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 99%|█████████▉| 106/107 [02:16<00:04, 4.92s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 26: 100%|██████████| 107/107 [02:20<00:00, 3.65s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 51%|█████▏ | 55/107 [00:43<00:29, 1.76batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 27: 52%|█████▏ | 56/107 [00:55<03:19, 3.91s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 53%|█████▎ | 57/107 [00:59<03:15, 3.91s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 54%|█████▍ | 58/107 [01:00<02:32, 3.12s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 55%|█████▌ | 59/107 [01:03<02:32, 3.17s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 56%|█████▌ | 60/107 [01:05<02:02, 2.61s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 57%|█████▋ | 61/107 [01:06<01:37, 2.12s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 58%|█████▊ | 62/107 [01:09<01:54, 2.55s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 59%|█████▉ | 63/107 [01:10<01:35, 2.18s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 60%|█████▉ | 64/107 [01:12<01:26, 2.01s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 61%|██████ | 65/107 [01:13<01:13, 1.74s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 62%|██████▏ | 66/107 [01:15<01:14, 1.82s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 63%|██████▎ | 67/107 [01:16<01:03, 1.58s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 64%|██████▎ | 68/107 [01:17<00:53, 1.37s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 64%|██████▍ | 69/107 [01:18<00:50, 1.33s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 65%|██████▌ | 70/107 [01:20<00:55, 1.50s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 66%|██████▋ | 71/107 [01:22<00:56, 1.56s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 67%|██████▋ | 72/107 [01:23<00:51, 1.47s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 68%|██████▊ | 73/107 [01:24<00:41, 1.22s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 69%|██████▉ | 74/107 [01:25<00:40, 1.22s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 70%|███████ | 75/107 [01:26<00:35, 1.10s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 71%|███████ | 76/107 [01:27<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 72%|███████▏ | 77/107 [01:28<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 73%|███████▎ | 78/107 [01:30<00:36, 1.27s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 74%|███████▍ | 79/107 [01:30<00:31, 1.11s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 75%|███████▍ | 80/107 [01:31<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 76%|███████▌ | 81/107 [01:32<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 77%|███████▋ | 82/107 [01:33<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 78%|███████▊ | 83/107 [01:34<00:22, 1.04batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 79%|███████▊ | 84/107 [01:35<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 79%|███████▉ | 85/107 [01:35<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 80%|████████ | 86/107 [01:36<00:18, 1.14batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 81%|████████▏ | 87/107 [01:37<00:17, 1.16batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 82%|████████▏ | 88/107 [01:38<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 83%|████████▎ | 89/107 [01:39<00:15, 1.16batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 84%|████████▍ | 90/107 [01:40<00:16, 1.02batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 85%|████████▌ | 91/107 [01:41<00:14, 1.08batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 86%|████████▌ | 92/107 [01:43<00:18, 1.22s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 87%|████████▋ | 93/107 [01:44<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 88%|████████▊ | 94/107 [01:45<00:15, 1.22s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 89%|████████▉ | 95/107 [01:50<00:27, 2.27s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 90%|████████▉ | 96/107 [01:51<00:21, 1.93s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 91%|█████████ | 97/107 [01:53<00:20, 2.03s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 92%|█████████▏| 98/107 [01:54<00:15, 1.67s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 93%|█████████▎| 99/107 [01:55<00:11, 1.44s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 93%|█████████▎| 100/107 [01:56<00:08, 1.22s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 94%|█████████▍| 101/107 [01:56<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 95%|█████████▌| 102/107 [01:58<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 96%|█████████▋| 103/107 [01:59<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 97%|█████████▋| 104/107 [01:59<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 98%|█████████▊| 105/107 [02:00<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 99%|█████████▉| 106/107 [02:01<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 27: 100%|██████████| 107/107 [02:06<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 51%|█████▏ | 55/107 [00:44<00:27, 1.89batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 28: 52%|█████▏ | 56/107 [00:55<03:11, 3.76s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 53%|█████▎ | 57/107 [00:59<03:05, 3.70s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 54%|█████▍ | 58/107 [01:01<02:36, 3.19s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 55%|█████▌ | 59/107 [01:03<02:15, 2.82s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 56%|█████▌ | 60/107 [01:05<01:55, 2.46s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 57%|█████▋ | 61/107 [01:06<01:38, 2.13s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 58%|█████▊ | 62/107 [01:08<01:36, 2.13s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 59%|█████▉ | 63/107 [01:09<01:20, 1.84s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 60%|█████▉ | 64/107 [01:11<01:15, 1.75s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 61%|██████ | 65/107 [01:12<01:03, 1.51s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 62%|██████▏ | 66/107 [01:13<01:04, 1.57s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 63%|██████▎ | 67/107 [01:15<00:56, 1.42s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 64%|██████▎ | 68/107 [01:15<00:49, 1.26s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 64%|██████▍ | 69/107 [01:17<00:48, 1.27s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 65%|██████▌ | 70/107 [01:18<00:44, 1.20s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 66%|██████▋ | 71/107 [01:20<00:55, 1.55s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 67%|██████▋ | 72/107 [01:21<00:50, 1.45s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 68%|██████▊ | 73/107 [01:22<00:41, 1.21s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 69%|██████▉ | 74/107 [01:23<00:40, 1.22s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 70%|███████ | 75/107 [01:24<00:35, 1.11s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 71%|███████ | 76/107 [01:25<00:31, 1.03s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 72%|███████▏ | 77/107 [01:26<00:30, 1.02s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 73%|███████▎ | 78/107 [01:28<00:36, 1.27s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 74%|███████▍ | 79/107 [01:29<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 75%|███████▍ | 80/107 [01:30<00:29, 1.11s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 76%|███████▌ | 81/107 [01:30<00:25, 1.04batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 77%|███████▋ | 82/107 [01:31<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 78%|███████▊ | 83/107 [01:32<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 79%|███████▊ | 84/107 [01:33<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 79%|███████▉ | 85/107 [01:34<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 80%|████████ | 86/107 [01:35<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 81%|████████▏ | 87/107 [01:35<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 82%|████████▏ | 88/107 [01:36<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 83%|████████▎ | 89/107 [01:37<00:15, 1.16batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 84%|████████▍ | 90/107 [01:38<00:16, 1.01batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 85%|████████▌ | 91/107 [01:39<00:14, 1.08batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 86%|████████▌ | 92/107 [01:41<00:18, 1.22s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 87%|████████▋ | 93/107 [01:42<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 88%|████████▊ | 94/107 [01:43<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 89%|████████▉ | 95/107 [01:48<00:26, 2.25s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 90%|████████▉ | 96/107 [01:49<00:20, 1.89s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 91%|█████████ | 97/107 [01:51<00:20, 2.01s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 92%|█████████▏| 98/107 [01:52<00:14, 1.65s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 93%|█████████▎| 99/107 [01:53<00:11, 1.42s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 93%|█████████▎| 100/107 [01:54<00:08, 1.21s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 94%|█████████▍| 101/107 [01:54<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 95%|█████████▌| 102/107 [01:56<00:05, 1.15s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 96%|█████████▋| 103/107 [01:57<00:04, 1.04s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 97%|█████████▋| 104/107 [01:57<00:02, 1.05batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 98%|█████████▊| 105/107 [01:58<00:01, 1.08batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 99%|█████████▉| 106/107 [01:59<00:01, 1.01s/batch, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 28: 100%|██████████| 107/107 [02:04<00:00, 1.09batch/s, batch_idx=54, gpu=0, loss=0.032, v_num=zf852wxn]\nEpoch 29: 51%|█████▏ | 55/107 [00:44<00:22, 2.31batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 29: 52%|█████▏ | 56/107 [00:55<03:10, 3.74s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 53%|█████▎ | 57/107 [00:59<03:01, 3.62s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 54%|█████▍ | 58/107 [01:00<02:24, 2.94s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 55%|█████▌ | 59/107 [01:03<02:23, 3.00s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 56%|█████▌ | 60/107 [01:05<01:59, 2.53s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 57%|█████▋ | 61/107 [01:06<01:33, 2.03s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 58%|█████▊ | 62/107 [01:11<02:13, 2.96s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 59%|█████▉ | 63/107 [01:12<01:44, 2.38s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 60%|█████▉ | 64/107 [01:13<01:31, 2.13s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 61%|██████ | 65/107 [01:14<01:16, 1.82s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 62%|██████▏ | 66/107 [01:16<01:16, 1.87s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 63%|██████▎ | 67/107 [01:17<01:05, 1.63s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 64%|██████▎ | 68/107 [01:18<00:55, 1.42s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 64%|██████▍ | 69/107 [01:20<01:00, 1.60s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 65%|██████▌ | 70/107 [01:21<00:52, 1.43s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 66%|██████▋ | 71/107 [01:23<00:54, 1.51s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 67%|██████▋ | 72/107 [01:24<00:49, 1.42s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 68%|██████▊ | 73/107 [01:25<00:40, 1.19s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 69%|██████▉ | 74/107 [01:26<00:39, 1.19s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 70%|███████ | 75/107 [01:27<00:34, 1.07s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 71%|███████ | 76/107 [01:28<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 72%|███████▏ | 77/107 [01:29<00:29, 1.00batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 73%|███████▎ | 78/107 [01:31<00:36, 1.25s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 74%|███████▍ | 79/107 [01:31<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 75%|███████▍ | 80/107 [01:32<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 76%|███████▌ | 81/107 [01:33<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 77%|███████▋ | 82/107 [01:34<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 78%|███████▊ | 83/107 [01:35<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 79%|███████▊ | 84/107 [01:36<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 79%|███████▉ | 85/107 [01:36<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 80%|████████ | 86/107 [01:37<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 81%|████████▏ | 87/107 [01:38<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 82%|████████▏ | 88/107 [01:39<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 83%|████████▎ | 89/107 [01:40<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 84%|████████▍ | 90/107 [01:41<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 85%|████████▌ | 91/107 [01:42<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 86%|████████▌ | 92/107 [01:44<00:17, 1.20s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 87%|████████▋ | 93/107 [01:44<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 88%|████████▊ | 94/107 [01:46<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 89%|████████▉ | 95/107 [01:51<00:26, 2.25s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 90%|████████▉ | 96/107 [01:52<00:20, 1.89s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 91%|█████████ | 97/107 [01:54<00:20, 2.01s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 92%|█████████▏| 98/107 [01:55<00:14, 1.64s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 93%|█████████▎| 99/107 [01:56<00:11, 1.41s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 93%|█████████▎| 100/107 [01:56<00:08, 1.21s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 94%|█████████▍| 101/107 [01:57<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 95%|█████████▌| 102/107 [01:58<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 96%|█████████▋| 103/107 [01:59<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 97%|█████████▋| 104/107 [02:00<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 98%|█████████▊| 105/107 [02:01<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 99%|█████████▉| 106/107 [02:02<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 29: 100%|██████████| 107/107 [02:06<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 51%|█████▏ | 55/107 [00:43<00:26, 1.94batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 30: 52%|█████▏ | 56/107 [00:55<03:16, 3.86s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 53%|█████▎ | 57/107 [00:59<03:24, 4.09s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 54%|█████▍ | 58/107 [01:01<02:41, 3.30s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 55%|█████▌ | 59/107 [01:03<02:24, 3.00s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 56%|█████▌ | 60/107 [01:04<01:56, 2.49s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 57%|█████▋ | 61/107 [01:06<01:35, 2.08s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 58%|█████▊ | 62/107 [01:09<01:57, 2.62s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 59%|█████▉ | 63/107 [01:11<01:35, 2.16s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 60%|█████▉ | 64/107 [01:12<01:24, 1.98s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 61%|██████ | 65/107 [01:13<01:12, 1.74s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 62%|██████▏ | 66/107 [01:15<01:12, 1.76s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 63%|██████▎ | 67/107 [01:16<01:06, 1.66s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 64%|██████▎ | 68/107 [01:17<00:56, 1.44s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 64%|██████▍ | 69/107 [01:19<00:52, 1.39s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 65%|██████▌ | 70/107 [01:20<00:49, 1.33s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 66%|██████▋ | 71/107 [01:22<00:57, 1.60s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 67%|██████▋ | 72/107 [01:23<00:52, 1.49s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 68%|██████▊ | 73/107 [01:24<00:41, 1.23s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 69%|██████▉ | 74/107 [01:25<00:40, 1.23s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 70%|███████ | 75/107 [01:26<00:35, 1.11s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 71%|███████ | 76/107 [01:27<00:31, 1.03s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 72%|███████▏ | 77/107 [01:28<00:30, 1.02s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 73%|███████▎ | 78/107 [01:30<00:36, 1.25s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 74%|███████▍ | 79/107 [01:30<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 75%|███████▍ | 80/107 [01:32<00:29, 1.11s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 76%|███████▌ | 81/107 [01:32<00:25, 1.03batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 77%|███████▋ | 82/107 [01:33<00:23, 1.09batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 78%|███████▊ | 83/107 [01:34<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 79%|███████▊ | 84/107 [01:35<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 79%|███████▉ | 85/107 [01:36<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 80%|████████ | 86/107 [01:36<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 81%|████████▏ | 87/107 [01:37<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 82%|████████▏ | 88/107 [01:38<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 83%|████████▎ | 89/107 [01:39<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 84%|████████▍ | 90/107 [01:40<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 85%|████████▌ | 91/107 [01:41<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 86%|████████▌ | 92/107 [01:43<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 87%|████████▋ | 93/107 [01:44<00:15, 1.10s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 88%|████████▊ | 94/107 [01:45<00:16, 1.23s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 89%|████████▉ | 95/107 [01:50<00:27, 2.28s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 90%|████████▉ | 96/107 [01:51<00:21, 1.92s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 91%|█████████ | 97/107 [01:53<00:20, 2.02s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 92%|█████████▏| 98/107 [01:54<00:14, 1.66s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 93%|█████████▎| 99/107 [01:55<00:11, 1.42s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 93%|█████████▎| 100/107 [01:56<00:08, 1.22s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 94%|█████████▍| 101/107 [01:56<00:06, 1.09s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 95%|█████████▌| 102/107 [01:58<00:05, 1.15s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 96%|█████████▋| 103/107 [01:59<00:04, 1.04s/batch, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 97%|█████████▋| 104/107 [01:59<00:02, 1.04batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 98%|█████████▊| 105/107 [02:00<00:01, 1.08batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 99%|█████████▉| 106/107 [02:01<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 30: 100%|██████████| 107/107 [02:06<00:00, 1.10batch/s, batch_idx=54, gpu=0, loss=0.031, v_num=zf852wxn]\nEpoch 31: 51%|█████▏ | 55/107 [00:45<00:25, 2.02batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 31: 52%|█████▏ | 56/107 [00:57<03:23, 3.99s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 53%|█████▎ | 57/107 [01:01<03:11, 3.82s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 54%|█████▍ | 58/107 [01:02<02:28, 3.03s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 55%|█████▌ | 59/107 [01:05<02:31, 3.15s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 56%|█████▌ | 60/107 [01:07<02:05, 2.67s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 57%|█████▋ | 61/107 [01:08<01:38, 2.14s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 58%|█████▊ | 62/107 [01:12<02:01, 2.69s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 59%|█████▉ | 63/107 [01:13<01:36, 2.19s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 60%|█████▉ | 64/107 [01:14<01:24, 1.97s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 61%|██████ | 65/107 [01:15<01:14, 1.77s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 62%|██████▏ | 66/107 [01:17<01:13, 1.79s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 63%|██████▎ | 67/107 [01:18<01:04, 1.61s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 64%|██████▎ | 68/107 [01:19<00:54, 1.41s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 64%|██████▍ | 69/107 [01:21<00:51, 1.36s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 65%|██████▌ | 70/107 [01:22<00:46, 1.26s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 66%|██████▋ | 71/107 [01:24<00:55, 1.54s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 67%|██████▋ | 72/107 [01:25<00:50, 1.44s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 68%|██████▊ | 73/107 [01:26<00:40, 1.20s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 69%|██████▉ | 74/107 [01:27<00:39, 1.20s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 70%|███████ | 75/107 [01:28<00:35, 1.10s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 71%|███████ | 76/107 [01:29<00:31, 1.02s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 72%|███████▏ | 77/107 [01:30<00:30, 1.00s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 73%|███████▎ | 78/107 [01:31<00:36, 1.26s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 74%|███████▍ | 79/107 [01:32<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 75%|███████▍ | 80/107 [01:33<00:30, 1.13s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 76%|███████▌ | 81/107 [01:34<00:25, 1.02batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 77%|███████▋ | 82/107 [01:35<00:23, 1.08batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 78%|███████▊ | 83/107 [01:36<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 79%|███████▊ | 84/107 [01:37<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 79%|███████▉ | 85/107 [01:37<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 80%|████████ | 86/107 [01:38<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 81%|████████▏ | 87/107 [01:39<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 82%|████████▏ | 88/107 [01:40<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 83%|████████▎ | 89/107 [01:41<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 84%|████████▍ | 90/107 [01:42<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 85%|████████▌ | 91/107 [01:43<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 86%|████████▌ | 92/107 [01:45<00:17, 1.20s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 87%|████████▋ | 93/107 [01:45<00:15, 1.07s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 88%|████████▊ | 94/107 [01:47<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 89%|████████▉ | 95/107 [01:52<00:27, 2.25s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 90%|████████▉ | 96/107 [01:53<00:20, 1.89s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 91%|█████████ | 97/107 [01:55<00:20, 2.00s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 92%|█████████▏| 98/107 [01:56<00:14, 1.64s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 93%|█████████▎| 99/107 [01:56<00:11, 1.41s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 93%|█████████▎| 100/107 [01:57<00:08, 1.21s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 94%|█████████▍| 101/107 [01:58<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 95%|█████████▌| 102/107 [01:59<00:05, 1.15s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 96%|█████████▋| 103/107 [02:00<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 97%|█████████▋| 104/107 [02:01<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 98%|█████████▊| 105/107 [02:02<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 99%|█████████▉| 106/107 [02:03<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 31: 100%|██████████| 107/107 [02:07<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 51%|█████▏ | 55/107 [00:46<00:37, 1.39batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 32: 52%|█████▏ | 56/107 [00:59<03:40, 4.33s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 53%|█████▎ | 57/107 [01:41<13:12, 15.84s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 54%|█████▍ | 58/107 [01:43<09:24, 11.51s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 55%|█████▌ | 59/107 [01:44<06:51, 8.58s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 56%|█████▌ | 60/107 [01:46<05:02, 6.44s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 57%|█████▋ | 61/107 [01:47<03:38, 4.75s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 58%|█████▊ | 62/107 [01:49<03:02, 4.06s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 59%|█████▉ | 63/107 [01:50<02:20, 3.20s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 60%|█████▉ | 64/107 [01:52<01:55, 2.70s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 61%|██████ | 65/107 [01:53<01:32, 2.21s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 62%|██████▏ | 66/107 [01:55<01:27, 2.13s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 63%|██████▎ | 67/107 [01:56<01:12, 1.82s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 64%|██████▎ | 68/107 [01:57<00:59, 1.53s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 64%|██████▍ | 69/107 [01:58<00:56, 1.50s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 65%|██████▌ | 70/107 [01:59<00:50, 1.36s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 66%|██████▋ | 71/107 [02:01<00:52, 1.46s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 67%|██████▋ | 72/107 [02:02<00:48, 1.39s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 68%|██████▊ | 73/107 [02:03<00:39, 1.17s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 69%|██████▉ | 74/107 [02:04<00:38, 1.18s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 70%|███████ | 75/107 [02:05<00:33, 1.06s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 71%|███████ | 76/107 [02:06<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 72%|███████▏ | 77/107 [02:07<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 73%|███████▎ | 78/107 [02:08<00:35, 1.24s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 74%|███████▍ | 79/107 [02:09<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 75%|███████▍ | 80/107 [02:10<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 76%|███████▌ | 81/107 [02:11<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 77%|███████▋ | 82/107 [02:12<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 78%|███████▊ | 83/107 [02:13<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 79%|███████▊ | 84/107 [02:13<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 79%|███████▉ | 85/107 [02:14<00:19, 1.16batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 80%|████████ | 86/107 [02:15<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 81%|████████▏ | 87/107 [02:16<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 82%|████████▏ | 88/107 [02:17<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 83%|████████▎ | 89/107 [02:18<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 84%|████████▍ | 90/107 [02:19<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 85%|████████▌ | 91/107 [02:20<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 86%|████████▌ | 92/107 [02:22<00:17, 1.20s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 87%|████████▋ | 93/107 [02:22<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 88%|████████▊ | 94/107 [02:24<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 89%|████████▉ | 95/107 [02:29<00:27, 2.25s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 90%|████████▉ | 96/107 [02:30<00:20, 1.90s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 91%|█████████ | 97/107 [02:32<00:20, 2.01s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 92%|█████████▏| 98/107 [02:33<00:14, 1.65s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 93%|█████████▎| 99/107 [02:34<00:11, 1.42s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 93%|█████████▎| 100/107 [02:34<00:08, 1.21s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 94%|█████████▍| 101/107 [02:35<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 95%|█████████▌| 102/107 [02:36<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 96%|█████████▋| 103/107 [02:37<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 97%|█████████▋| 104/107 [02:38<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 98%|█████████▊| 105/107 [02:39<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 99%|█████████▉| 106/107 [02:40<00:00, 1.00batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 32: 100%|██████████| 107/107 [02:44<00:00, 1.09batch/s, batch_idx=54, gpu=0, loss=0.030, v_num=zf852wxn]\nEpoch 33: 51%|█████▏ | 55/107 [00:46<00:23, 2.24batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 33: 52%|█████▏ | 56/107 [00:59<03:34, 4.21s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 53%|█████▎ | 57/107 [01:04<03:36, 4.34s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 54%|█████▍ | 58/107 [01:05<02:46, 3.39s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 55%|█████▌ | 59/107 [01:08<02:43, 3.40s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 56%|█████▌ | 60/107 [01:09<02:08, 2.72s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 57%|█████▋ | 61/107 [01:10<01:42, 2.23s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 58%|█████▊ | 62/107 [01:14<01:56, 2.58s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 59%|█████▉ | 63/107 [01:15<01:33, 2.13s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 60%|█████▉ | 64/107 [01:16<01:24, 1.96s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 61%|██████ | 65/107 [01:18<01:12, 1.73s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 62%|██████▏ | 66/107 [01:20<01:13, 1.78s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 63%|██████▎ | 67/107 [01:21<01:03, 1.59s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 64%|██████▎ | 68/107 [01:22<00:54, 1.38s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 64%|██████▍ | 69/107 [01:23<00:52, 1.38s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 65%|██████▌ | 70/107 [01:24<00:48, 1.30s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 66%|██████▋ | 71/107 [01:26<00:57, 1.61s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 67%|██████▋ | 72/107 [01:28<00:52, 1.50s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 68%|██████▊ | 73/107 [01:28<00:42, 1.24s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 69%|██████▉ | 74/107 [01:29<00:40, 1.24s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 70%|███████ | 75/107 [01:30<00:35, 1.10s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 71%|███████ | 76/107 [01:31<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 72%|███████▏ | 77/107 [01:32<00:30, 1.00s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 73%|███████▎ | 78/107 [01:34<00:36, 1.26s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 74%|███████▍ | 79/107 [01:35<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 75%|███████▍ | 80/107 [01:36<00:29, 1.11s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 76%|███████▌ | 81/107 [01:36<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 77%|███████▋ | 82/107 [01:37<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 78%|███████▊ | 83/107 [01:38<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 79%|███████▊ | 84/107 [01:39<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 79%|███████▉ | 85/107 [01:40<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 80%|████████ | 86/107 [01:41<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 81%|████████▏ | 87/107 [01:41<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 82%|████████▏ | 88/107 [01:42<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 83%|████████▎ | 89/107 [01:43<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 84%|████████▍ | 90/107 [01:44<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 85%|████████▌ | 91/107 [01:45<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 86%|████████▌ | 92/107 [01:47<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 87%|████████▋ | 93/107 [01:48<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 88%|████████▊ | 94/107 [01:49<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 89%|████████▉ | 95/107 [01:54<00:27, 2.29s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 90%|████████▉ | 96/107 [01:55<00:21, 1.92s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 91%|█████████ | 97/107 [01:58<00:20, 2.03s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 92%|█████████▏| 98/107 [01:58<00:14, 1.66s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 93%|█████████▎| 99/107 [01:59<00:11, 1.43s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 93%|█████████▎| 100/107 [02:00<00:08, 1.22s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 94%|█████████▍| 101/107 [02:01<00:06, 1.08s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 95%|█████████▌| 102/107 [02:02<00:05, 1.15s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 96%|█████████▋| 103/107 [02:03<00:04, 1.04s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 97%|█████████▋| 104/107 [02:04<00:02, 1.05batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 98%|█████████▊| 105/107 [02:04<00:01, 1.08batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 99%|█████████▉| 106/107 [02:06<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 33: 100%|██████████| 107/107 [02:10<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 51%|█████▏ | 55/107 [00:45<00:31, 1.67batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 34: 52%|█████▏ | 56/107 [00:59<03:52, 4.55s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 53%|█████▎ | 57/107 [01:02<03:28, 4.16s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 54%|█████▍ | 58/107 [01:04<02:45, 3.37s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 55%|█████▌ | 59/107 [01:06<02:28, 3.10s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 56%|█████▌ | 60/107 [01:08<02:04, 2.65s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 57%|█████▋ | 61/107 [01:09<01:41, 2.21s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 58%|█████▊ | 62/107 [01:11<01:35, 2.13s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 59%|█████▉ | 63/107 [01:12<01:20, 1.83s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 60%|█████▉ | 64/107 [01:14<01:14, 1.73s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 61%|██████ | 65/107 [01:15<01:05, 1.57s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 62%|██████▏ | 66/107 [01:17<01:08, 1.66s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 63%|██████▎ | 67/107 [01:18<01:00, 1.51s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 64%|██████▎ | 68/107 [01:19<00:51, 1.32s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 64%|██████▍ | 69/107 [01:20<00:49, 1.29s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 65%|██████▌ | 70/107 [01:21<00:45, 1.22s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 66%|██████▋ | 71/107 [01:23<00:48, 1.34s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 67%|██████▋ | 72/107 [01:25<00:53, 1.53s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 68%|██████▊ | 73/107 [01:25<00:42, 1.26s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 69%|██████▉ | 74/107 [01:26<00:40, 1.23s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 70%|███████ | 75/107 [01:27<00:34, 1.09s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 71%|███████ | 76/107 [01:28<00:31, 1.02s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 72%|███████▏ | 77/107 [01:29<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 73%|███████▎ | 78/107 [01:31<00:36, 1.26s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 74%|███████▍ | 79/107 [01:32<00:31, 1.11s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 75%|███████▍ | 80/107 [01:33<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 76%|███████▌ | 81/107 [01:33<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 77%|███████▋ | 82/107 [01:34<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 78%|███████▊ | 83/107 [01:35<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 79%|███████▊ | 84/107 [01:36<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 79%|███████▉ | 85/107 [01:37<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 80%|████████ | 86/107 [01:37<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 81%|████████▏ | 87/107 [01:38<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 82%|████████▏ | 88/107 [01:39<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 83%|████████▎ | 89/107 [01:40<00:15, 1.16batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 84%|████████▍ | 90/107 [01:41<00:16, 1.01batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 85%|████████▌ | 91/107 [01:42<00:14, 1.08batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 86%|████████▌ | 92/107 [01:44<00:18, 1.22s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 87%|████████▋ | 93/107 [01:45<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 88%|████████▊ | 94/107 [01:46<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 89%|████████▉ | 95/107 [01:51<00:27, 2.27s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 90%|████████▉ | 96/107 [01:52<00:21, 1.92s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 91%|█████████ | 97/107 [01:54<00:20, 2.03s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 92%|█████████▏| 98/107 [01:55<00:15, 1.67s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 93%|█████████▎| 99/107 [01:56<00:11, 1.45s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 93%|█████████▎| 100/107 [01:57<00:08, 1.23s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 94%|█████████▍| 101/107 [01:58<00:06, 1.08s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 95%|█████████▌| 102/107 [01:59<00:05, 1.15s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 96%|█████████▋| 103/107 [02:00<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 97%|█████████▋| 104/107 [02:00<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 98%|█████████▊| 105/107 [02:01<00:01, 1.08batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 99%|█████████▉| 106/107 [02:02<00:01, 1.02s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 34: 100%|██████████| 107/107 [02:07<00:00, 1.08batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 51%|█████▏ | 55/107 [00:47<00:29, 1.75batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 35: 52%|█████▏ | 56/107 [01:01<03:52, 4.56s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 53%|█████▎ | 57/107 [01:04<03:30, 4.21s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 54%|█████▍ | 58/107 [01:06<02:43, 3.33s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 55%|█████▌ | 59/107 [01:08<02:29, 3.11s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 56%|█████▌ | 60/107 [01:10<02:03, 2.63s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 57%|█████▋ | 61/107 [01:11<01:40, 2.18s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 58%|█████▊ | 62/107 [01:13<01:37, 2.16s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 59%|█████▉ | 63/107 [01:14<01:23, 1.89s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 60%|█████▉ | 64/107 [01:16<01:18, 1.82s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 61%|██████ | 65/107 [01:17<01:08, 1.64s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 62%|██████▏ | 66/107 [01:19<01:07, 1.65s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 63%|██████▎ | 67/107 [01:20<00:59, 1.48s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 64%|██████▎ | 68/107 [01:21<00:50, 1.31s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 64%|██████▍ | 69/107 [01:22<00:49, 1.30s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 65%|██████▌ | 70/107 [01:23<00:45, 1.23s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 66%|██████▋ | 71/107 [01:25<00:49, 1.38s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 67%|██████▋ | 72/107 [01:26<00:47, 1.35s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 68%|██████▊ | 73/107 [01:27<00:39, 1.15s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 69%|██████▉ | 74/107 [01:29<00:44, 1.35s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 70%|███████ | 75/107 [01:29<00:37, 1.19s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 71%|███████ | 76/107 [01:30<00:33, 1.07s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 72%|███████▏ | 77/107 [01:31<00:31, 1.05s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 73%|███████▎ | 78/107 [01:33<00:37, 1.28s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 74%|███████▍ | 79/107 [01:34<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 75%|███████▍ | 80/107 [01:35<00:30, 1.12s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 76%|███████▌ | 81/107 [01:36<00:25, 1.03batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 77%|███████▋ | 82/107 [01:36<00:23, 1.07batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 78%|███████▊ | 83/107 [01:37<00:23, 1.03batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 79%|███████▊ | 84/107 [01:38<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 79%|███████▉ | 85/107 [01:39<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 80%|████████ | 86/107 [01:40<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 81%|████████▏ | 87/107 [01:41<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 82%|████████▏ | 88/107 [01:41<00:15, 1.21batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 83%|████████▎ | 89/107 [01:42<00:15, 1.16batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 84%|████████▍ | 90/107 [01:44<00:16, 1.00batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 85%|████████▌ | 91/107 [01:45<00:15, 1.06batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 86%|████████▌ | 92/107 [01:46<00:18, 1.23s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 87%|████████▋ | 93/107 [01:47<00:15, 1.11s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 88%|████████▊ | 94/107 [01:49<00:15, 1.23s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 89%|████████▉ | 95/107 [01:54<00:27, 2.29s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 90%|████████▉ | 96/107 [01:55<00:21, 1.92s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 91%|█████████ | 97/107 [01:57<00:20, 2.03s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 92%|█████████▏| 98/107 [01:58<00:14, 1.66s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 93%|█████████▎| 99/107 [01:59<00:11, 1.43s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 93%|█████████▎| 100/107 [01:59<00:08, 1.22s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 94%|█████████▍| 101/107 [02:00<00:06, 1.08s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 95%|█████████▌| 102/107 [02:01<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 96%|█████████▋| 103/107 [02:02<00:04, 1.04s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 97%|█████████▋| 104/107 [02:03<00:02, 1.05batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 98%|█████████▊| 105/107 [02:04<00:01, 1.08batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 99%|█████████▉| 106/107 [02:05<00:01, 1.00s/batch, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 35: 100%|██████████| 107/107 [02:09<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.029, v_num=zf852wxn]\nEpoch 36: 51%|█████▏ | 55/107 [00:48<00:20, 2.54batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 36: 52%|█████▏ | 56/107 [01:02<03:54, 4.59s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 53%|█████▎ | 57/107 [01:05<03:16, 3.93s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 54%|█████▍ | 58/107 [01:06<02:36, 3.18s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 55%|█████▌ | 59/107 [01:10<02:49, 3.54s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 56%|█████▌ | 60/107 [01:12<02:15, 2.89s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 57%|█████▋ | 61/107 [01:13<01:48, 2.35s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 58%|█████▊ | 62/107 [01:17<02:03, 2.74s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 59%|█████▉ | 63/107 [01:18<01:38, 2.24s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 60%|█████▉ | 64/107 [01:19<01:28, 2.05s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 61%|██████ | 65/107 [01:20<01:15, 1.80s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 62%|██████▏ | 66/107 [01:22<01:13, 1.79s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 63%|██████▎ | 67/107 [01:24<01:05, 1.64s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 64%|██████▎ | 68/107 [01:24<00:55, 1.42s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 64%|██████▍ | 69/107 [01:26<00:53, 1.40s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 65%|██████▌ | 70/107 [01:27<00:48, 1.31s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 66%|██████▋ | 71/107 [01:29<00:58, 1.61s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 67%|██████▋ | 72/107 [01:30<00:52, 1.50s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 68%|██████▊ | 73/107 [01:31<00:42, 1.24s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 69%|██████▉ | 74/107 [01:32<00:41, 1.24s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 70%|███████ | 75/107 [01:33<00:35, 1.11s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 71%|███████ | 76/107 [01:34<00:31, 1.02s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 72%|███████▏ | 77/107 [01:35<00:30, 1.02s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 73%|███████▎ | 78/107 [01:37<00:37, 1.28s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 74%|███████▍ | 79/107 [01:38<00:31, 1.13s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 75%|███████▍ | 80/107 [01:39<00:30, 1.12s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 76%|███████▌ | 81/107 [01:39<00:25, 1.03batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 77%|███████▋ | 82/107 [01:40<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 78%|███████▊ | 83/107 [01:41<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 79%|███████▊ | 84/107 [01:42<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 79%|███████▉ | 85/107 [01:43<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 80%|████████ | 86/107 [01:44<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 81%|████████▏ | 87/107 [01:44<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 82%|████████▏ | 88/107 [01:45<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 83%|████████▎ | 89/107 [01:46<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 84%|████████▍ | 90/107 [01:47<00:16, 1.02batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 85%|████████▌ | 91/107 [01:48<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 86%|████████▌ | 92/107 [01:50<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 87%|████████▋ | 93/107 [01:51<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 88%|████████▊ | 94/107 [01:52<00:15, 1.22s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 89%|████████▉ | 95/107 [01:57<00:27, 2.30s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 90%|████████▉ | 96/107 [01:58<00:21, 1.93s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 91%|█████████ | 97/107 [02:01<00:20, 2.05s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 92%|█████████▏| 98/107 [02:01<00:15, 1.69s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 93%|█████████▎| 99/107 [02:02<00:11, 1.44s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 93%|█████████▎| 100/107 [02:03<00:08, 1.23s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 94%|█████████▍| 101/107 [02:04<00:06, 1.08s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 95%|█████████▌| 102/107 [02:05<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 96%|█████████▋| 103/107 [02:06<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 97%|█████████▋| 104/107 [02:07<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 98%|█████████▊| 105/107 [02:07<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 99%|█████████▉| 106/107 [02:09<00:00, 1.00batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 36: 100%|██████████| 107/107 [02:13<00:00, 1.10batch/s, batch_idx=54, gpu=0, loss=0.028, v_num=zf852wxn]\nEpoch 37: 51%|█████▏ | 55/107 [00:49<00:32, 1.60batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 37: 52%|█████▏ | 56/107 [01:03<03:58, 4.69s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 53%|█████▎ | 57/107 [01:05<03:21, 4.02s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 54%|█████▍ | 58/107 [01:07<02:36, 3.20s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 55%|█████▌ | 59/107 [01:11<02:43, 3.41s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 56%|█████▌ | 60/107 [01:12<02:13, 2.84s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 57%|█████▋ | 61/107 [01:13<01:43, 2.24s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 58%|█████▊ | 62/107 [01:18<02:15, 3.01s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 59%|█████▉ | 63/107 [01:19<01:50, 2.51s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 60%|█████▉ | 64/107 [01:20<01:32, 2.14s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 61%|██████ | 65/107 [01:22<01:17, 1.86s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 62%|██████▏ | 66/107 [01:23<01:16, 1.86s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 63%|██████▎ | 67/107 [01:25<01:07, 1.70s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 64%|██████▎ | 68/107 [01:26<00:56, 1.45s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 64%|██████▍ | 69/107 [01:27<00:52, 1.39s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 65%|██████▌ | 70/107 [01:28<00:48, 1.30s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 66%|██████▋ | 71/107 [01:30<00:59, 1.64s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 67%|██████▋ | 72/107 [01:32<00:53, 1.54s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 68%|██████▊ | 73/107 [01:32<00:43, 1.27s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 69%|██████▉ | 74/107 [01:34<00:41, 1.27s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 70%|███████ | 75/107 [01:34<00:35, 1.12s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 71%|███████ | 76/107 [01:35<00:32, 1.03s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 72%|███████▏ | 77/107 [01:36<00:30, 1.02s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 73%|███████▎ | 78/107 [01:38<00:36, 1.27s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 74%|███████▍ | 79/107 [01:39<00:31, 1.11s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 75%|███████▍ | 80/107 [01:40<00:29, 1.11s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 76%|███████▌ | 81/107 [01:41<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 77%|███████▋ | 82/107 [01:41<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 78%|███████▊ | 83/107 [01:42<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 79%|███████▊ | 84/107 [01:43<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 79%|███████▉ | 85/107 [01:44<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 80%|████████ | 86/107 [01:45<00:18, 1.14batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 81%|████████▏ | 87/107 [01:46<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 82%|████████▏ | 88/107 [01:46<00:15, 1.21batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 83%|████████▎ | 89/107 [01:47<00:15, 1.16batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 84%|████████▍ | 90/107 [01:49<00:16, 1.01batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 85%|████████▌ | 91/107 [01:49<00:14, 1.08batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 86%|████████▌ | 92/107 [01:51<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 87%|████████▋ | 93/107 [01:52<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 88%|████████▊ | 94/107 [01:54<00:15, 1.23s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 89%|████████▉ | 95/107 [01:58<00:27, 2.29s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 90%|████████▉ | 96/107 [01:59<00:21, 1.92s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 91%|█████████ | 97/107 [02:02<00:20, 2.04s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 92%|█████████▏| 98/107 [02:03<00:15, 1.68s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 93%|█████████▎| 99/107 [02:04<00:11, 1.44s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 93%|█████████▎| 100/107 [02:04<00:08, 1.22s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 94%|█████████▍| 101/107 [02:05<00:06, 1.08s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 95%|█████████▌| 102/107 [02:06<00:05, 1.15s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 96%|█████████▋| 103/107 [02:07<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 97%|█████████▋| 104/107 [02:08<00:02, 1.05batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 98%|█████████▊| 105/107 [02:09<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 99%|█████████▉| 106/107 [02:10<00:01, 1.00s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 37: 100%|██████████| 107/107 [02:14<00:00, 1.10batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 51%|█████▏ | 55/107 [00:48<00:24, 2.10batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 38: 52%|█████▏ | 56/107 [01:03<03:56, 4.64s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 53%|█████▎ | 57/107 [01:07<03:49, 4.60s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 54%|█████▍ | 58/107 [01:08<02:54, 3.57s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 55%|█████▌ | 59/107 [02:11<17:00, 21.25s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 56%|█████▌ | 60/107 [02:12<11:57, 15.26s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 57%|█████▋ | 61/107 [02:13<08:24, 10.97s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 58%|█████▊ | 62/107 [02:15<06:12, 8.29s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 59%|█████▉ | 63/107 [02:16<04:28, 6.10s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 60%|█████▉ | 64/107 [02:18<03:24, 4.76s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 61%|██████ | 65/107 [02:19<02:34, 3.68s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 62%|██████▏ | 66/107 [02:21<02:08, 3.14s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 63%|██████▎ | 67/107 [02:22<01:46, 2.66s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 64%|██████▎ | 68/107 [02:23<01:22, 2.12s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 64%|██████▍ | 69/107 [02:24<01:10, 1.85s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 65%|██████▌ | 70/107 [02:25<00:59, 1.60s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 66%|██████▋ | 71/107 [02:27<00:58, 1.62s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 67%|██████▋ | 72/107 [02:28<00:52, 1.49s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 68%|██████▊ | 73/107 [02:29<00:42, 1.24s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 69%|██████▉ | 74/107 [02:30<00:40, 1.23s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 70%|███████ | 75/107 [02:31<00:35, 1.10s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 71%|███████ | 76/107 [02:32<00:31, 1.03s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 72%|███████▏ | 77/107 [02:33<00:30, 1.03s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 73%|███████▎ | 78/107 [02:35<00:37, 1.30s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 74%|███████▍ | 79/107 [02:36<00:31, 1.14s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 75%|███████▍ | 80/107 [02:37<00:30, 1.14s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 76%|███████▌ | 81/107 [02:37<00:25, 1.00batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 77%|███████▋ | 82/107 [02:38<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 78%|███████▊ | 83/107 [02:39<00:23, 1.01batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 79%|███████▊ | 84/107 [02:40<00:21, 1.09batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 79%|███████▉ | 85/107 [02:41<00:19, 1.15batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 80%|████████ | 86/107 [02:42<00:18, 1.14batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 81%|████████▏ | 87/107 [02:43<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 82%|████████▏ | 88/107 [02:43<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 83%|████████▎ | 89/107 [02:44<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 84%|████████▍ | 90/107 [02:45<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 85%|████████▌ | 91/107 [02:46<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 86%|████████▌ | 92/107 [02:48<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 87%|████████▋ | 93/107 [02:49<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 88%|████████▊ | 94/107 [02:50<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 89%|████████▉ | 95/107 [02:55<00:25, 2.14s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 90%|████████▉ | 96/107 [02:56<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 91%|█████████ | 97/107 [02:58<00:19, 1.93s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 92%|█████████▏| 98/107 [02:59<00:14, 1.60s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 93%|█████████▎| 99/107 [03:00<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 93%|█████████▎| 100/107 [03:00<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 94%|█████████▍| 101/107 [03:01<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 95%|█████████▌| 102/107 [03:02<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 96%|█████████▋| 103/107 [03:03<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 97%|█████████▋| 104/107 [03:04<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 98%|█████████▊| 105/107 [03:05<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 99%|█████████▉| 106/107 [03:06<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 38: 100%|██████████| 107/107 [03:10<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.027, v_num=zf852wxn]\nEpoch 39: 51%|█████▏ | 55/107 [00:48<00:27, 1.87batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 39: 52%|█████▏ | 56/107 [01:03<04:04, 4.79s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 53%|█████▎ | 57/107 [01:06<03:32, 4.25s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 54%|█████▍ | 58/107 [01:07<02:42, 3.31s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 55%|█████▌ | 59/107 [01:11<02:49, 3.54s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 56%|█████▌ | 60/107 [01:12<02:15, 2.88s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 57%|█████▋ | 61/107 [01:13<01:46, 2.31s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 58%|█████▊ | 62/107 [01:17<02:02, 2.72s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 59%|█████▉ | 63/107 [01:18<01:37, 2.21s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 60%|█████▉ | 64/107 [01:20<01:25, 2.00s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 61%|██████ | 65/107 [01:21<01:12, 1.72s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 62%|██████▏ | 66/107 [01:22<01:10, 1.73s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 63%|██████▎ | 67/107 [01:24<01:04, 1.62s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 64%|██████▎ | 68/107 [01:25<00:54, 1.40s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 64%|██████▍ | 69/107 [01:26<00:50, 1.32s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 65%|██████▌ | 70/107 [01:28<00:53, 1.45s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 66%|██████▋ | 71/107 [01:29<00:54, 1.51s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 67%|██████▋ | 72/107 [01:30<00:49, 1.43s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 68%|██████▊ | 73/107 [01:31<00:40, 1.20s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 69%|██████▉ | 74/107 [01:32<00:39, 1.19s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 70%|███████ | 75/107 [01:33<00:34, 1.07s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 71%|███████ | 76/107 [01:34<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 72%|███████▏ | 77/107 [01:35<00:30, 1.00s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 73%|███████▎ | 78/107 [01:37<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 74%|███████▍ | 79/107 [01:37<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 75%|███████▍ | 80/107 [01:39<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 76%|███████▌ | 81/107 [01:39<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 77%|███████▋ | 82/107 [01:40<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 78%|███████▊ | 83/107 [01:41<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 79%|███████▊ | 84/107 [01:42<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 79%|███████▉ | 85/107 [01:42<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 80%|████████ | 86/107 [01:43<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 81%|████████▏ | 87/107 [01:44<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 82%|████████▏ | 88/107 [01:45<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 83%|████████▎ | 89/107 [01:46<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 84%|████████▍ | 90/107 [01:47<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 85%|████████▌ | 91/107 [01:48<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 86%|████████▌ | 92/107 [01:50<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 87%|████████▋ | 93/107 [01:50<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 88%|████████▊ | 94/107 [01:52<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 89%|████████▉ | 95/107 [01:56<00:25, 2.13s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 90%|████████▉ | 96/107 [01:57<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 91%|█████████ | 97/107 [01:59<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 92%|█████████▏| 98/107 [02:00<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 93%|█████████▎| 99/107 [02:01<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 93%|█████████▎| 100/107 [02:02<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 94%|█████████▍| 101/107 [02:03<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 95%|█████████▌| 102/107 [02:04<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 96%|█████████▋| 103/107 [02:05<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 97%|█████████▋| 104/107 [02:05<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 98%|█████████▊| 105/107 [02:06<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 99%|█████████▉| 106/107 [02:07<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 39: 100%|██████████| 107/107 [02:12<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 51%|█████▏ | 55/107 [00:49<00:19, 2.70batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 40: 52%|█████▏ | 56/107 [01:05<04:08, 4.87s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 53%|█████▎ | 57/107 [01:07<03:33, 4.27s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 54%|█████▍ | 58/107 [01:09<02:43, 3.35s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 55%|█████▌ | 59/107 [01:12<02:45, 3.45s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 56%|█████▌ | 60/107 [01:14<02:11, 2.80s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 57%|█████▋ | 61/107 [01:15<01:44, 2.26s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 58%|█████▊ | 62/107 [01:19<02:06, 2.81s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 59%|█████▉ | 63/107 [01:20<01:42, 2.32s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 60%|█████▉ | 64/107 [01:21<01:26, 2.00s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 61%|██████ | 65/107 [01:22<01:13, 1.75s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 62%|██████▏ | 66/107 [01:24<01:12, 1.77s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 63%|██████▎ | 67/107 [01:25<01:04, 1.61s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 64%|██████▎ | 68/107 [01:26<00:54, 1.40s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 64%|██████▍ | 69/107 [01:28<00:52, 1.38s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 65%|██████▌ | 70/107 [01:29<00:48, 1.31s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 66%|██████▋ | 71/107 [01:31<00:57, 1.61s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 67%|██████▋ | 72/107 [01:32<00:51, 1.48s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 68%|██████▊ | 73/107 [01:33<00:42, 1.24s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 69%|██████▉ | 74/107 [01:34<00:40, 1.22s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 70%|███████ | 75/107 [01:35<00:34, 1.09s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 71%|███████ | 76/107 [01:36<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 72%|███████▏ | 77/107 [01:37<00:30, 1.00s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 73%|███████▎ | 78/107 [01:38<00:35, 1.24s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 74%|███████▍ | 79/107 [01:39<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 75%|███████▍ | 80/107 [01:40<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 76%|███████▌ | 81/107 [01:41<00:25, 1.04batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 77%|███████▋ | 82/107 [01:42<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 78%|███████▊ | 83/107 [01:43<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 79%|███████▊ | 84/107 [01:43<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 79%|███████▉ | 85/107 [01:44<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 80%|████████ | 86/107 [01:45<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 81%|████████▏ | 87/107 [01:46<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 82%|████████▏ | 88/107 [01:47<00:15, 1.25batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 83%|████████▎ | 89/107 [01:48<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 84%|████████▍ | 90/107 [01:49<00:16, 1.06batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 85%|████████▌ | 91/107 [01:50<00:14, 1.12batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 86%|████████▌ | 92/107 [01:51<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 87%|████████▋ | 93/107 [01:52<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 88%|████████▊ | 94/107 [01:54<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 89%|████████▉ | 95/107 [01:58<00:25, 2.13s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 90%|████████▉ | 96/107 [01:59<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 91%|█████████ | 97/107 [02:01<00:19, 1.91s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 92%|█████████▏| 98/107 [02:02<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 93%|█████████▎| 99/107 [02:03<00:10, 1.36s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 93%|█████████▎| 100/107 [02:04<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 94%|█████████▍| 101/107 [02:04<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 95%|█████████▌| 102/107 [02:06<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 96%|█████████▋| 103/107 [02:06<00:04, 1.00s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 97%|█████████▋| 104/107 [02:07<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 98%|█████████▊| 105/107 [02:08<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 99%|█████████▉| 106/107 [02:09<00:00, 1.04batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 40: 100%|██████████| 107/107 [02:14<00:00, 1.14batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 51%|█████▏ | 55/107 [00:49<00:31, 1.64batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 41: 52%|█████▏ | 56/107 [01:04<04:21, 5.13s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 53%|█████▎ | 57/107 [01:07<03:33, 4.27s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 54%|█████▍ | 58/107 [01:08<02:47, 3.42s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 55%|█████▌ | 59/107 [01:12<02:48, 3.50s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 56%|█████▌ | 60/107 [01:13<02:15, 2.89s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 57%|█████▋ | 61/107 [01:14<01:47, 2.34s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 58%|█████▊ | 62/107 [01:18<02:07, 2.82s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 59%|█████▉ | 63/107 [01:19<01:42, 2.33s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 60%|█████▉ | 64/107 [01:21<01:30, 2.11s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 61%|██████ | 65/107 [01:22<01:14, 1.78s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 62%|██████▏ | 66/107 [01:24<01:13, 1.79s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 63%|██████▎ | 67/107 [01:25<01:03, 1.59s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 64%|██████▎ | 68/107 [01:26<00:53, 1.38s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 64%|██████▍ | 69/107 [01:27<00:51, 1.35s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 65%|██████▌ | 70/107 [01:28<00:48, 1.31s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 66%|██████▋ | 71/107 [01:31<00:58, 1.62s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 67%|██████▋ | 72/107 [01:32<00:52, 1.49s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 68%|██████▊ | 73/107 [01:32<00:41, 1.23s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 69%|██████▉ | 74/107 [01:34<00:40, 1.22s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 70%|███████ | 75/107 [01:34<00:34, 1.09s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 71%|███████ | 76/107 [01:35<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 72%|███████▏ | 77/107 [01:36<00:29, 1.00batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 73%|███████▎ | 78/107 [01:38<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 74%|███████▍ | 79/107 [01:39<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 75%|███████▍ | 80/107 [01:40<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 76%|███████▌ | 81/107 [01:40<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 77%|███████▋ | 82/107 [01:41<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 78%|███████▊ | 83/107 [01:42<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 79%|███████▊ | 84/107 [01:43<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 79%|███████▉ | 85/107 [01:44<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 80%|████████ | 86/107 [01:45<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 81%|████████▏ | 87/107 [01:45<00:16, 1.20batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 82%|████████▏ | 88/107 [01:46<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 83%|████████▎ | 89/107 [01:47<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 84%|████████▍ | 90/107 [01:48<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 85%|████████▌ | 91/107 [01:49<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 86%|████████▌ | 92/107 [01:51<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 87%|████████▋ | 93/107 [01:52<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 88%|████████▊ | 94/107 [01:53<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 89%|████████▉ | 95/107 [01:57<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 90%|████████▉ | 96/107 [01:59<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 91%|█████████ | 97/107 [02:01<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 92%|█████████▏| 98/107 [02:02<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 93%|█████████▎| 99/107 [02:02<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 93%|█████████▎| 100/107 [02:03<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 94%|█████████▍| 101/107 [02:04<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 95%|█████████▌| 102/107 [02:05<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 96%|█████████▋| 103/107 [02:06<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 97%|█████████▋| 104/107 [02:07<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 98%|█████████▊| 105/107 [02:08<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 99%|█████████▉| 106/107 [02:09<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 41: 100%|██████████| 107/107 [02:13<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.026, v_num=zf852wxn]\nEpoch 42: 51%|█████▏ | 55/107 [00:49<00:25, 2.05batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 42: 52%|█████▏ | 56/107 [01:04<04:13, 4.98s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 53%|█████▎ | 57/107 [01:08<03:42, 4.44s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 54%|█████▍ | 58/107 [01:09<02:50, 3.47s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 55%|█████▌ | 59/107 [01:13<02:57, 3.71s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 56%|█████▌ | 60/107 [01:14<02:21, 3.00s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 57%|█████▋ | 61/107 [01:15<01:49, 2.39s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 58%|█████▊ | 62/107 [01:18<01:52, 2.50s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 59%|█████▉ | 63/107 [01:19<01:34, 2.14s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 60%|█████▉ | 64/107 [01:21<01:22, 1.91s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 61%|██████ | 65/107 [01:22<01:11, 1.71s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 62%|██████▏ | 66/107 [01:24<01:10, 1.72s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 63%|██████▎ | 67/107 [01:25<01:03, 1.59s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 64%|██████▎ | 68/107 [01:26<00:53, 1.37s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 64%|██████▍ | 69/107 [01:27<00:51, 1.35s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 65%|██████▌ | 70/107 [01:28<00:47, 1.28s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 66%|██████▋ | 71/107 [01:31<00:56, 1.56s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 67%|██████▋ | 72/107 [01:32<00:51, 1.47s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 68%|██████▊ | 73/107 [01:33<00:41, 1.23s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 69%|██████▉ | 74/107 [01:34<00:40, 1.23s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 70%|███████ | 75/107 [01:35<00:35, 1.11s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 71%|███████ | 76/107 [01:35<00:31, 1.03s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 72%|███████▏ | 77/107 [01:36<00:30, 1.03s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 73%|███████▎ | 78/107 [01:38<00:36, 1.26s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 74%|███████▍ | 79/107 [01:39<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 75%|███████▍ | 80/107 [01:40<00:30, 1.12s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 76%|███████▌ | 81/107 [01:41<00:25, 1.02batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 77%|███████▋ | 82/107 [01:42<00:23, 1.06batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 78%|███████▊ | 83/107 [01:43<00:23, 1.03batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 79%|███████▊ | 84/107 [01:43<00:21, 1.08batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 79%|███████▉ | 85/107 [01:44<00:19, 1.12batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 80%|████████ | 86/107 [01:45<00:18, 1.11batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 81%|████████▏ | 87/107 [01:46<00:17, 1.14batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 82%|████████▏ | 88/107 [01:47<00:16, 1.17batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 83%|████████▎ | 89/107 [01:48<00:16, 1.12batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 84%|████████▍ | 90/107 [01:49<00:17, 1.01s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 85%|████████▌ | 91/107 [01:50<00:15, 1.04batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 86%|████████▌ | 92/107 [01:52<00:18, 1.23s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 87%|████████▋ | 93/107 [01:53<00:15, 1.11s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 88%|████████▊ | 94/107 [01:54<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 89%|████████▉ | 95/107 [01:59<00:26, 2.17s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 90%|████████▉ | 96/107 [02:00<00:20, 1.83s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 91%|█████████ | 97/107 [02:02<00:19, 1.94s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 92%|█████████▏| 98/107 [02:03<00:14, 1.61s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 93%|█████████▎| 99/107 [02:03<00:11, 1.40s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 93%|█████████▎| 100/107 [02:04<00:08, 1.21s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 94%|█████████▍| 101/107 [02:05<00:06, 1.08s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 95%|█████████▌| 102/107 [02:06<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 96%|█████████▋| 103/107 [02:07<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 97%|█████████▋| 104/107 [02:08<00:02, 1.05batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 98%|█████████▊| 105/107 [02:09<00:01, 1.07batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 99%|█████████▉| 106/107 [02:10<00:01, 1.01s/batch, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 42: 100%|██████████| 107/107 [02:14<00:00, 1.08batch/s, batch_idx=54, gpu=0, loss=0.025, v_num=zf852wxn]\nEpoch 43: 51%|█████▏ | 55/107 [00:49<00:31, 1.67batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 43: 52%|█████▏ | 56/107 [01:05<04:19, 5.09s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 53%|█████▎ | 57/107 [01:08<03:45, 4.50s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 54%|█████▍ | 58/107 [01:09<02:54, 3.56s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 55%|█████▌ | 59/107 [01:13<02:54, 3.63s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 56%|█████▌ | 60/107 [01:14<02:18, 2.95s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 57%|█████▋ | 61/107 [01:16<01:50, 2.40s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 58%|█████▊ | 62/107 [01:20<02:10, 2.89s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 59%|█████▉ | 63/107 [01:21<01:44, 2.38s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 60%|█████▉ | 64/107 [01:22<01:31, 2.13s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 61%|██████ | 65/107 [01:23<01:15, 1.80s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 62%|██████▏ | 66/107 [01:25<01:13, 1.79s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 63%|██████▎ | 67/107 [01:26<01:02, 1.55s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 64%|██████▎ | 68/107 [01:27<00:53, 1.37s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 64%|██████▍ | 69/107 [01:29<00:57, 1.51s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 65%|██████▌ | 70/107 [01:30<00:50, 1.37s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 66%|██████▋ | 71/107 [01:32<00:51, 1.44s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 67%|██████▋ | 72/107 [01:33<00:47, 1.36s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 68%|██████▊ | 73/107 [01:33<00:38, 1.15s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 69%|██████▉ | 74/107 [01:35<00:38, 1.15s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 70%|███████ | 75/107 [01:35<00:33, 1.05s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 71%|███████ | 76/107 [01:36<00:30, 1.03batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 72%|███████▏ | 77/107 [01:37<00:29, 1.03batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 73%|███████▎ | 78/107 [01:39<00:34, 1.20s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 74%|███████▍ | 79/107 [01:40<00:30, 1.07s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 75%|███████▍ | 80/107 [01:41<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 76%|███████▌ | 81/107 [01:41<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 77%|███████▋ | 82/107 [01:42<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 78%|███████▊ | 83/107 [01:43<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 79%|███████▊ | 84/107 [01:44<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 79%|███████▉ | 85/107 [01:45<00:18, 1.19batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 80%|████████ | 86/107 [01:46<00:17, 1.18batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 81%|████████▏ | 87/107 [01:46<00:16, 1.20batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 82%|████████▏ | 88/107 [01:47<00:15, 1.25batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 83%|████████▎ | 89/107 [01:48<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 84%|████████▍ | 90/107 [01:49<00:16, 1.05batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 85%|████████▌ | 91/107 [01:50<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 86%|████████▌ | 92/107 [01:52<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 87%|████████▋ | 93/107 [01:53<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 88%|████████▊ | 94/107 [01:54<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 89%|████████▉ | 95/107 [01:58<00:25, 2.11s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 90%|████████▉ | 96/107 [01:59<00:19, 1.79s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 91%|█████████ | 97/107 [02:02<00:18, 1.90s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 92%|█████████▏| 98/107 [02:02<00:14, 1.56s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 93%|█████████▎| 99/107 [02:03<00:10, 1.35s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 93%|█████████▎| 100/107 [02:04<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 94%|█████████▍| 101/107 [02:05<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 95%|█████████▌| 102/107 [02:06<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 96%|█████████▋| 103/107 [02:07<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 97%|█████████▋| 104/107 [02:07<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 98%|█████████▊| 105/107 [02:08<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 99%|█████████▉| 106/107 [02:09<00:00, 1.04batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 43: 100%|██████████| 107/107 [02:14<00:00, 1.14batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 51%|█████▏ | 55/107 [00:49<00:28, 1.81batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 44: 52%|█████▏ | 56/107 [01:06<04:39, 5.48s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 53%|█████▎ | 57/107 [01:10<04:00, 4.81s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 54%|█████▍ | 58/107 [01:11<03:02, 3.73s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 55%|█████▌ | 59/107 [01:15<03:00, 3.76s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 56%|█████▌ | 60/107 [01:16<02:21, 3.02s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 57%|█████▋ | 61/107 [01:17<01:49, 2.38s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 58%|█████▊ | 62/107 [01:20<01:54, 2.55s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 59%|█████▉ | 63/107 [01:21<01:33, 2.12s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 60%|█████▉ | 64/107 [01:22<01:22, 1.92s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 61%|██████ | 65/107 [01:23<01:08, 1.63s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 62%|██████▏ | 66/107 [01:25<01:10, 1.71s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 63%|██████▎ | 67/107 [01:26<01:00, 1.50s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 64%|██████▎ | 68/107 [01:27<00:51, 1.33s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 64%|██████▍ | 69/107 [01:28<00:49, 1.31s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 65%|██████▌ | 70/107 [01:30<00:45, 1.23s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 66%|██████▋ | 71/107 [01:32<00:55, 1.55s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 67%|██████▋ | 72/107 [01:33<00:50, 1.45s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 68%|██████▊ | 73/107 [01:34<00:41, 1.22s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 69%|██████▉ | 74/107 [01:35<00:40, 1.21s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 70%|███████ | 75/107 [01:36<00:35, 1.09s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 71%|███████ | 76/107 [01:37<00:31, 1.02s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 72%|███████▏ | 77/107 [01:38<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 73%|███████▎ | 78/107 [01:39<00:36, 1.25s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 74%|███████▍ | 79/107 [01:40<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 75%|███████▍ | 80/107 [01:41<00:30, 1.12s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 76%|███████▌ | 81/107 [01:42<00:25, 1.03batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 77%|███████▋ | 82/107 [01:43<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 78%|███████▊ | 83/107 [01:44<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 79%|███████▊ | 84/107 [01:45<00:20, 1.11batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 79%|███████▉ | 85/107 [01:45<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 80%|████████ | 86/107 [01:46<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 81%|████████▏ | 87/107 [01:47<00:17, 1.18batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 82%|████████▏ | 88/107 [01:48<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 83%|████████▎ | 89/107 [01:49<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 84%|████████▍ | 90/107 [01:50<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 85%|████████▌ | 91/107 [01:51<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 86%|████████▌ | 92/107 [01:52<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 87%|████████▋ | 93/107 [01:53<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 88%|████████▊ | 94/107 [01:55<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 89%|████████▉ | 95/107 [01:59<00:25, 2.13s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 90%|████████▉ | 96/107 [02:00<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 91%|█████████ | 97/107 [02:02<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 92%|█████████▏| 98/107 [02:03<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 93%|█████████▎| 99/107 [02:04<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 93%|█████████▎| 100/107 [02:05<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 94%|█████████▍| 101/107 [02:05<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 95%|█████████▌| 102/107 [02:07<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 96%|█████████▋| 103/107 [02:07<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 97%|█████████▋| 104/107 [02:08<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 98%|█████████▊| 105/107 [02:09<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 99%|█████████▉| 106/107 [02:10<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 44: 100%|██████████| 107/107 [02:15<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 51%|█████▏ | 55/107 [00:50<00:22, 2.32batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 45: 52%|█████▏ | 56/107 [01:07<04:30, 5.31s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 53%|█████▎ | 57/107 [01:10<03:45, 4.51s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 54%|█████▍ | 58/107 [01:11<02:53, 3.53s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 55%|█████▌ | 59/107 [01:15<02:49, 3.52s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 56%|█████▌ | 60/107 [01:16<02:17, 2.93s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 57%|█████▋ | 61/107 [01:17<01:45, 2.30s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 58%|█████▊ | 62/107 [01:21<02:06, 2.81s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 59%|█████▉ | 63/107 [01:22<01:42, 2.32s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 60%|█████▉ | 64/107 [01:24<01:28, 2.05s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 61%|██████ | 65/107 [01:25<01:13, 1.75s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 62%|██████▏ | 66/107 [01:26<01:11, 1.74s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 63%|██████▎ | 67/107 [02:03<08:14, 12.36s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 64%|██████▎ | 68/107 [02:05<05:57, 9.15s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 64%|██████▍ | 69/107 [02:06<04:17, 6.77s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 65%|██████▌ | 70/107 [02:07<03:06, 5.05s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 66%|██████▋ | 71/107 [02:09<02:24, 4.02s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 67%|██████▋ | 72/107 [02:10<01:51, 3.18s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 68%|██████▊ | 73/107 [02:11<01:22, 2.42s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 69%|██████▉ | 74/107 [02:12<01:07, 2.05s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 70%|███████ | 75/107 [02:13<00:53, 1.67s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 71%|███████ | 76/107 [02:14<00:43, 1.41s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 72%|███████▏ | 77/107 [02:15<00:38, 1.28s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 73%|███████▎ | 78/107 [02:17<00:43, 1.48s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 74%|███████▍ | 79/107 [02:17<00:37, 1.32s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 75%|███████▍ | 80/107 [02:19<00:35, 1.31s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 76%|███████▌ | 81/107 [02:20<00:29, 1.14s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 77%|███████▋ | 82/107 [02:20<00:27, 1.08s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 78%|███████▊ | 83/107 [02:22<00:26, 1.12s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 79%|███████▊ | 84/107 [02:23<00:23, 1.04s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 79%|███████▉ | 85/107 [02:23<00:21, 1.00batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 80%|████████ | 86/107 [02:24<00:21, 1.01s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 81%|████████▏ | 87/107 [02:25<00:19, 1.01batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 82%|████████▏ | 88/107 [02:26<00:18, 1.03batch/s, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 83%|████████▎ | 89/107 [02:27<00:18, 1.01s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 84%|████████▍ | 90/107 [02:29<00:19, 1.13s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 85%|████████▌ | 91/107 [02:30<00:17, 1.08s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 86%|████████▌ | 92/107 [02:32<00:20, 1.37s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 87%|████████▋ | 93/107 [02:33<00:17, 1.25s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 88%|████████▊ | 94/107 [02:34<00:17, 1.38s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 89%|████████▉ | 95/107 [02:39<00:28, 2.40s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 90%|████████▉ | 96/107 [02:41<00:22, 2.05s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 91%|█████████ | 97/107 [02:43<00:21, 2.17s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 92%|█████████▏| 98/107 [02:44<00:16, 1.80s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 93%|█████████▎| 99/107 [02:45<00:12, 1.57s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 93%|█████████▎| 100/107 [02:46<00:09, 1.35s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 94%|█████████▍| 101/107 [02:47<00:07, 1.20s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 95%|█████████▌| 102/107 [02:48<00:06, 1.28s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 96%|█████████▋| 103/107 [02:49<00:04, 1.17s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 97%|█████████▋| 104/107 [02:50<00:03, 1.07s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 98%|█████████▊| 105/107 [02:51<00:02, 1.05s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 99%|█████████▉| 106/107 [02:52<00:01, 1.14s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 45: 100%|██████████| 107/107 [02:57<00:00, 1.04s/batch, batch_idx=54, gpu=0, loss=0.024, v_num=zf852wxn]\nEpoch 46: 51%|█████▏ | 55/107 [00:51<00:23, 2.18batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 46: 52%|█████▏ | 56/107 [01:07<04:31, 5.32s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 53%|█████▎ | 57/107 [01:11<03:53, 4.67s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 54%|█████▍ | 58/107 [01:12<02:57, 3.62s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 55%|█████▌ | 59/107 [01:16<02:55, 3.66s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 56%|█████▌ | 60/107 [01:17<02:18, 2.95s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 57%|█████▋ | 61/107 [01:18<01:46, 2.31s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 58%|█████▊ | 62/107 [01:22<02:10, 2.90s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 59%|█████▉ | 63/107 [01:23<01:43, 2.34s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 60%|█████▉ | 64/107 [01:25<01:30, 2.10s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 61%|██████ | 65/107 [01:26<01:16, 1.82s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 62%|██████▏ | 66/107 [01:27<01:13, 1.80s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 63%|██████▎ | 67/107 [01:29<01:06, 1.66s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 64%|██████▎ | 68/107 [01:30<00:55, 1.43s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 64%|██████▍ | 69/107 [01:31<00:52, 1.37s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 65%|██████▌ | 70/107 [01:33<00:53, 1.45s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 66%|██████▋ | 71/107 [01:34<00:54, 1.50s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 67%|██████▋ | 72/107 [01:35<00:49, 1.41s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 68%|██████▊ | 73/107 [01:36<00:40, 1.18s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 69%|██████▉ | 74/107 [01:37<00:38, 1.18s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 70%|███████ | 75/107 [01:38<00:33, 1.06s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 71%|███████ | 76/107 [01:39<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 72%|███████▏ | 77/107 [01:40<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 73%|███████▎ | 78/107 [01:42<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 74%|███████▍ | 79/107 [01:42<00:30, 1.07s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 75%|███████▍ | 80/107 [01:43<00:28, 1.07s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 76%|███████▌ | 81/107 [01:44<00:24, 1.07batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 77%|███████▋ | 82/107 [01:45<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 78%|███████▊ | 83/107 [01:46<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 79%|███████▊ | 84/107 [01:47<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 79%|███████▉ | 85/107 [01:47<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 80%|████████ | 86/107 [01:48<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 81%|████████▏ | 87/107 [01:49<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 82%|████████▏ | 88/107 [01:50<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 83%|████████▎ | 89/107 [01:51<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 84%|████████▍ | 90/107 [01:52<00:16, 1.05batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 85%|████████▌ | 91/107 [01:53<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 86%|████████▌ | 92/107 [01:54<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 87%|████████▋ | 93/107 [01:55<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 88%|████████▊ | 94/107 [01:57<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 89%|████████▉ | 95/107 [02:01<00:25, 2.15s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 90%|████████▉ | 96/107 [02:02<00:20, 1.83s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 91%|█████████ | 97/107 [02:04<00:19, 1.94s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 92%|█████████▏| 98/107 [02:05<00:14, 1.61s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 93%|█████████▎| 99/107 [02:06<00:11, 1.39s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 93%|█████████▎| 100/107 [02:07<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 94%|█████████▍| 101/107 [02:08<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 95%|█████████▌| 102/107 [02:09<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 96%|█████████▋| 103/107 [02:10<00:04, 1.00s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 97%|█████████▋| 104/107 [02:10<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 98%|█████████▊| 105/107 [02:11<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 99%|█████████▉| 106/107 [02:12<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 46: 100%|██████████| 107/107 [02:17<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 51%|█████▏ | 55/107 [00:51<00:16, 3.19batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 47: 52%|█████▏ | 56/107 [01:07<04:27, 5.24s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 53%|█████▎ | 57/107 [01:10<03:47, 4.55s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 54%|█████▍ | 58/107 [01:12<02:58, 3.63s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 55%|█████▌ | 59/107 [01:15<02:41, 3.36s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 56%|█████▌ | 60/107 [01:16<02:15, 2.88s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 57%|█████▋ | 61/107 [01:17<01:47, 2.35s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 58%|█████▊ | 62/107 [01:19<01:39, 2.21s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 59%|█████▉ | 63/107 [01:21<01:24, 1.92s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 60%|█████▉ | 64/107 [01:22<01:18, 1.83s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 61%|██████ | 65/107 [01:23<01:05, 1.57s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 62%|██████▏ | 66/107 [01:25<01:07, 1.64s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 63%|██████▎ | 67/107 [01:26<00:59, 1.48s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 64%|██████▎ | 68/107 [01:27<00:50, 1.29s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 64%|██████▍ | 69/107 [01:28<00:49, 1.30s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 65%|██████▌ | 70/107 [01:29<00:45, 1.22s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 66%|██████▋ | 71/107 [01:31<00:54, 1.52s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 67%|██████▋ | 72/107 [01:33<00:50, 1.43s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 68%|██████▊ | 73/107 [01:33<00:40, 1.20s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 69%|██████▉ | 74/107 [01:35<00:39, 1.19s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 70%|███████ | 75/107 [01:35<00:34, 1.07s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 71%|███████ | 76/107 [01:36<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 72%|███████▏ | 77/107 [01:37<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 73%|███████▎ | 78/107 [01:39<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 74%|███████▍ | 79/107 [01:40<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 75%|███████▍ | 80/107 [01:41<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 76%|███████▌ | 81/107 [01:41<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 77%|███████▋ | 82/107 [01:42<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 78%|███████▊ | 83/107 [01:43<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 79%|███████▊ | 84/107 [01:44<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 79%|███████▉ | 85/107 [01:45<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 80%|████████ | 86/107 [01:46<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 81%|████████▏ | 87/107 [01:46<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 82%|████████▏ | 88/107 [01:47<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 83%|████████▎ | 89/107 [01:48<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 84%|████████▍ | 90/107 [01:49<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 85%|████████▌ | 91/107 [01:50<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 86%|████████▌ | 92/107 [01:52<00:17, 1.16s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 87%|████████▋ | 93/107 [01:53<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 88%|████████▊ | 94/107 [01:54<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 89%|████████▉ | 95/107 [01:58<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 90%|████████▉ | 96/107 [01:59<00:19, 1.79s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 91%|█████████ | 97/107 [02:02<00:19, 1.90s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 92%|█████████▏| 98/107 [02:02<00:14, 1.57s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 93%|█████████▎| 99/107 [02:03<00:10, 1.36s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 93%|█████████▎| 100/107 [02:04<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 94%|█████████▍| 101/107 [02:05<00:06, 1.03s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 95%|█████████▌| 102/107 [02:06<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 96%|█████████▋| 103/107 [02:07<00:03, 1.00batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 97%|█████████▋| 104/107 [02:07<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 98%|█████████▊| 105/107 [02:08<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 99%|█████████▉| 106/107 [02:09<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 47: 100%|██████████| 107/107 [02:14<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 51%|█████▏ | 55/107 [00:51<00:18, 2.78batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 48: 52%|█████▏ | 56/107 [01:08<04:28, 5.27s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 53%|█████▎ | 57/107 [01:11<03:54, 4.70s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 54%|█████▍ | 58/107 [01:13<03:02, 3.73s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 55%|█████▌ | 59/107 [01:15<02:36, 3.26s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 56%|█████▌ | 60/107 [01:17<02:10, 2.78s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 57%|█████▋ | 61/107 [01:18<01:44, 2.27s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 58%|█████▊ | 62/107 [01:20<01:37, 2.18s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 59%|█████▉ | 63/107 [01:21<01:22, 1.87s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 60%|█████▉ | 64/107 [01:22<01:14, 1.73s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 61%|██████ | 65/107 [01:24<01:09, 1.65s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 62%|██████▏ | 66/107 [01:25<01:05, 1.61s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 63%|██████▎ | 67/107 [01:26<00:57, 1.45s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 64%|██████▎ | 68/107 [01:27<00:50, 1.30s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 64%|██████▍ | 69/107 [01:28<00:49, 1.29s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 65%|██████▌ | 70/107 [01:29<00:44, 1.22s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 66%|██████▋ | 71/107 [01:31<00:48, 1.34s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 67%|██████▋ | 72/107 [01:32<00:45, 1.30s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 68%|██████▊ | 73/107 [01:34<00:44, 1.30s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 69%|██████▉ | 74/107 [01:35<00:42, 1.27s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 70%|███████ | 75/107 [01:36<00:35, 1.12s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 71%|███████ | 76/107 [01:36<00:32, 1.04s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 72%|███████▏ | 77/107 [01:37<00:30, 1.02s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 73%|███████▎ | 78/107 [01:39<00:36, 1.25s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 74%|███████▍ | 79/107 [01:40<00:31, 1.11s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 75%|███████▍ | 80/107 [01:41<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 76%|███████▌ | 81/107 [01:42<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 77%|███████▋ | 82/107 [01:42<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 78%|███████▊ | 83/107 [01:44<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 79%|███████▊ | 84/107 [01:44<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 79%|███████▉ | 85/107 [01:45<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 80%|████████ | 86/107 [01:46<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 81%|████████▏ | 87/107 [01:47<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 82%|████████▏ | 88/107 [01:48<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 83%|████████▎ | 89/107 [01:48<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 84%|████████▍ | 90/107 [01:50<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 85%|████████▌ | 91/107 [01:50<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 86%|████████▌ | 92/107 [01:52<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 87%|████████▋ | 93/107 [01:53<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 88%|████████▊ | 94/107 [01:54<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 89%|████████▉ | 95/107 [01:59<00:25, 2.13s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 90%|████████▉ | 96/107 [02:00<00:20, 1.83s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 91%|█████████ | 97/107 [02:02<00:19, 1.93s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 92%|█████████▏| 98/107 [02:03<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 93%|█████████▎| 99/107 [02:04<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 93%|█████████▎| 100/107 [02:05<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 94%|█████████▍| 101/107 [02:05<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 95%|█████████▌| 102/107 [02:07<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 96%|█████████▋| 103/107 [02:07<00:03, 1.00batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 97%|█████████▋| 104/107 [02:08<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 98%|█████████▊| 105/107 [02:09<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 99%|█████████▉| 106/107 [02:10<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 48: 100%|██████████| 107/107 [02:15<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 51%|█████▏ | 55/107 [00:52<00:24, 2.10batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 49: 52%|█████▏ | 56/107 [01:09<04:37, 5.44s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 53%|█████▎ | 57/107 [01:12<03:51, 4.63s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 54%|█████▍ | 58/107 [01:13<02:58, 3.64s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 55%|█████▌ | 59/107 [01:17<03:00, 3.75s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 56%|█████▌ | 60/107 [01:18<02:24, 3.07s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 57%|█████▋ | 61/107 [01:19<01:52, 2.45s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 58%|█████▊ | 62/107 [01:23<02:08, 2.86s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 59%|█████▉ | 63/107 [01:24<01:42, 2.33s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 60%|█████▉ | 64/107 [01:26<01:30, 2.09s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 61%|██████ | 65/107 [01:27<01:17, 1.84s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 62%|██████▏ | 66/107 [01:29<01:13, 1.80s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 63%|██████▎ | 67/107 [01:30<01:06, 1.67s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 64%|██████▎ | 68/107 [01:31<00:56, 1.46s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 64%|██████▍ | 69/107 [01:33<01:02, 1.64s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 65%|██████▌ | 70/107 [01:34<00:54, 1.46s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 66%|██████▋ | 71/107 [01:36<00:55, 1.53s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 67%|██████▋ | 72/107 [01:37<00:50, 1.43s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 68%|██████▊ | 73/107 [01:38<00:41, 1.21s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 69%|██████▉ | 74/107 [01:39<00:40, 1.22s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 70%|███████ | 75/107 [01:40<00:35, 1.10s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 71%|███████ | 76/107 [01:41<00:31, 1.02s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 72%|███████▏ | 77/107 [01:42<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 73%|███████▎ | 78/107 [01:44<00:35, 1.24s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 74%|███████▍ | 79/107 [01:44<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 75%|███████▍ | 80/107 [01:45<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 76%|███████▌ | 81/107 [01:46<00:25, 1.04batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 77%|███████▋ | 82/107 [01:47<00:23, 1.09batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 78%|███████▊ | 83/107 [01:48<00:23, 1.03batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 79%|███████▊ | 84/107 [01:49<00:20, 1.11batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 79%|███████▉ | 85/107 [01:50<00:19, 1.14batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 80%|████████ | 86/107 [01:50<00:18, 1.12batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 81%|████████▏ | 87/107 [01:51<00:17, 1.13batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 82%|████████▏ | 88/107 [01:52<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 83%|████████▎ | 89/107 [01:53<00:15, 1.14batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 84%|████████▍ | 90/107 [01:54<00:16, 1.01batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 85%|████████▌ | 91/107 [01:55<00:14, 1.07batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 86%|████████▌ | 92/107 [01:57<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 87%|████████▋ | 93/107 [01:58<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 88%|████████▊ | 94/107 [01:59<00:15, 1.22s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 89%|████████▉ | 95/107 [02:04<00:25, 2.16s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 90%|████████▉ | 96/107 [02:05<00:20, 1.84s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 91%|█████████ | 97/107 [02:07<00:19, 1.94s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 92%|█████████▏| 98/107 [02:08<00:14, 1.61s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 93%|█████████▎| 99/107 [02:09<00:11, 1.40s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 93%|█████████▎| 100/107 [02:09<00:08, 1.21s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 94%|█████████▍| 101/107 [02:10<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 95%|█████████▌| 102/107 [02:11<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 96%|█████████▋| 103/107 [02:12<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 97%|█████████▋| 104/107 [02:13<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 98%|█████████▊| 105/107 [02:14<00:01, 1.08batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 99%|█████████▉| 106/107 [02:15<00:00, 1.00batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 49: 100%|██████████| 107/107 [02:20<00:00, 1.09batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 51%|█████▏ | 55/107 [00:51<00:18, 2.88batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 50: 52%|█████▏ | 56/107 [01:08<04:33, 5.36s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 53%|█████▎ | 57/107 [01:11<03:52, 4.64s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 54%|█████▍ | 58/107 [01:12<02:59, 3.67s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 55%|█████▌ | 59/107 [01:16<02:58, 3.71s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 56%|█████▌ | 60/107 [01:17<02:19, 2.97s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 57%|█████▋ | 61/107 [01:18<01:48, 2.36s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 58%|█████▊ | 62/107 [01:23<02:14, 2.99s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 59%|█████▉ | 63/107 [01:24<01:46, 2.43s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 60%|█████▉ | 64/107 [01:25<01:30, 2.11s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 61%|██████ | 65/107 [01:26<01:16, 1.83s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 62%|██████▏ | 66/107 [01:28<01:13, 1.79s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 63%|██████▎ | 67/107 [01:29<01:05, 1.63s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 64%|██████▎ | 68/107 [01:30<00:54, 1.39s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 64%|██████▍ | 69/107 [01:31<00:50, 1.34s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 65%|██████▌ | 70/107 [01:32<00:47, 1.27s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 66%|██████▋ | 71/107 [01:35<00:56, 1.57s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 67%|██████▋ | 72/107 [01:36<00:51, 1.48s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 68%|██████▊ | 73/107 [01:37<00:42, 1.24s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 69%|██████▉ | 74/107 [01:38<00:41, 1.26s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 70%|███████ | 75/107 [01:39<00:36, 1.14s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 71%|███████ | 76/107 [01:40<00:32, 1.06s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 72%|███████▏ | 77/107 [01:41<00:31, 1.05s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 73%|███████▎ | 78/107 [01:43<00:37, 1.28s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 74%|███████▍ | 79/107 [01:43<00:31, 1.13s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 75%|███████▍ | 80/107 [01:44<00:30, 1.13s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 76%|███████▌ | 81/107 [01:45<00:25, 1.01batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 77%|███████▋ | 82/107 [01:46<00:23, 1.05batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 78%|███████▊ | 83/107 [01:47<00:23, 1.00batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 79%|███████▊ | 84/107 [01:48<00:21, 1.07batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 79%|███████▉ | 85/107 [01:49<00:19, 1.12batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 80%|████████ | 86/107 [01:50<00:18, 1.11batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 81%|████████▏ | 87/107 [01:50<00:17, 1.14batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 82%|████████▏ | 88/107 [01:51<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 83%|████████▎ | 89/107 [01:52<00:15, 1.13batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 84%|████████▍ | 90/107 [01:53<00:16, 1.01batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 85%|████████▌ | 91/107 [01:54<00:15, 1.06batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 86%|████████▌ | 92/107 [01:56<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 87%|████████▋ | 93/107 [01:57<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 88%|████████▊ | 94/107 [01:58<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 89%|████████▉ | 95/107 [02:03<00:25, 2.16s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 90%|████████▉ | 96/107 [02:04<00:20, 1.84s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 91%|█████████ | 97/107 [02:06<00:19, 1.95s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 92%|█████████▏| 98/107 [02:07<00:14, 1.62s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 93%|█████████▎| 99/107 [02:08<00:11, 1.40s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 93%|█████████▎| 100/107 [02:09<00:08, 1.21s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 94%|█████████▍| 101/107 [02:09<00:06, 1.08s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 95%|█████████▌| 102/107 [02:11<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 96%|█████████▋| 103/107 [02:11<00:04, 1.05s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 97%|█████████▋| 104/107 [02:12<00:02, 1.03batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 98%|█████████▊| 105/107 [02:13<00:01, 1.06batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 99%|█████████▉| 106/107 [02:14<00:01, 1.02s/batch, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 50: 100%|██████████| 107/107 [02:19<00:00, 1.06batch/s, batch_idx=54, gpu=0, loss=0.023, v_num=zf852wxn]\nEpoch 51: 51%|█████▏ | 55/107 [00:51<00:24, 2.11batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 51: 52%|█████▏ | 56/107 [01:09<04:43, 5.56s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 53%|█████▎ | 57/107 [01:12<04:05, 4.91s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 54%|█████▍ | 58/107 [01:13<03:04, 3.76s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 55%|█████▌ | 59/107 [01:17<02:58, 3.71s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 56%|█████▌ | 60/107 [01:18<02:24, 3.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 57%|█████▋ | 61/107 [01:19<01:52, 2.45s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 58%|█████▊ | 62/107 [01:23<02:07, 2.82s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 59%|█████▉ | 63/107 [01:24<01:42, 2.32s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 60%|█████▉ | 64/107 [01:26<01:30, 2.10s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 61%|██████ | 65/107 [01:27<01:15, 1.81s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 62%|██████▏ | 66/107 [01:29<01:14, 1.83s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 63%|██████▎ | 67/107 [01:30<01:04, 1.60s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 64%|██████▎ | 68/107 [01:31<00:54, 1.40s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 64%|██████▍ | 69/107 [01:32<00:51, 1.36s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 65%|██████▌ | 70/107 [01:33<00:47, 1.29s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 66%|██████▋ | 71/107 [01:35<00:50, 1.41s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 67%|██████▋ | 72/107 [01:37<00:54, 1.56s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 68%|██████▊ | 73/107 [01:38<00:43, 1.29s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 69%|██████▉ | 74/107 [01:39<00:41, 1.27s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 70%|███████ | 75/107 [01:40<00:35, 1.12s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 71%|███████ | 76/107 [01:40<00:31, 1.03s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 72%|███████▏ | 77/107 [01:41<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 73%|███████▎ | 78/107 [01:43<00:35, 1.24s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 74%|███████▍ | 79/107 [01:44<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 75%|███████▍ | 80/107 [01:45<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 76%|███████▌ | 81/107 [01:46<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 77%|███████▋ | 82/107 [01:46<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 78%|███████▊ | 83/107 [01:47<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 79%|███████▊ | 84/107 [01:48<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 79%|███████▉ | 85/107 [01:49<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 80%|████████ | 86/107 [01:50<00:18, 1.14batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 81%|████████▏ | 87/107 [01:51<00:17, 1.16batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 82%|████████▏ | 88/107 [01:51<00:15, 1.21batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 83%|████████▎ | 89/107 [01:52<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 84%|████████▍ | 90/107 [01:54<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 85%|████████▌ | 91/107 [01:54<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 86%|████████▌ | 92/107 [01:56<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 87%|████████▋ | 93/107 [01:57<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 88%|████████▊ | 94/107 [01:58<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 89%|████████▉ | 95/107 [02:03<00:25, 2.14s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 90%|████████▉ | 96/107 [02:04<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 91%|█████████ | 97/107 [02:06<00:19, 1.91s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 92%|█████████▏| 98/107 [02:07<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 93%|█████████▎| 99/107 [02:08<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 93%|█████████▎| 100/107 [02:08<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 94%|█████████▍| 101/107 [02:09<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 95%|█████████▌| 102/107 [02:10<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 96%|█████████▋| 103/107 [02:11<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 97%|█████████▋| 104/107 [02:12<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 98%|█████████▊| 105/107 [02:13<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 99%|█████████▉| 106/107 [02:14<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 51: 100%|██████████| 107/107 [02:19<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 51%|█████▏ | 55/107 [00:53<00:30, 1.71batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 52: 52%|█████▏ | 56/107 [01:11<04:59, 5.87s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 53%|█████▎ | 57/107 [01:15<04:13, 5.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 54%|█████▍ | 58/107 [01:16<03:12, 3.92s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 55%|█████▌ | 59/107 [01:19<03:03, 3.82s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 56%|█████▌ | 60/107 [01:21<02:25, 3.09s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 57%|█████▋ | 61/107 [01:22<01:51, 2.42s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 58%|█████▊ | 62/107 [01:26<02:13, 2.96s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 59%|█████▉ | 63/107 [01:27<01:47, 2.45s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 60%|█████▉ | 64/107 [01:29<01:33, 2.19s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 61%|██████ | 65/107 [01:30<01:19, 1.90s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 62%|██████▏ | 66/107 [01:32<01:16, 1.87s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 63%|██████▎ | 67/107 [01:34<01:14, 1.85s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 64%|██████▎ | 68/107 [01:35<01:10, 1.80s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 64%|██████▍ | 69/107 [01:37<01:03, 1.68s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 65%|██████▌ | 70/107 [01:39<01:04, 1.74s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 66%|██████▋ | 71/107 [01:40<01:03, 1.76s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 67%|██████▋ | 72/107 [01:42<00:56, 1.62s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 68%|██████▊ | 73/107 [01:42<00:45, 1.35s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 69%|██████▉ | 74/107 [01:44<00:43, 1.31s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 70%|███████ | 75/107 [01:44<00:37, 1.16s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 71%|███████ | 76/107 [01:45<00:32, 1.06s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 72%|███████▏ | 77/107 [01:46<00:31, 1.04s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 73%|███████▎ | 78/107 [01:48<00:36, 1.27s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 74%|███████▍ | 79/107 [01:49<00:36, 1.30s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 75%|███████▍ | 80/107 [01:51<00:36, 1.36s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 76%|███████▌ | 81/107 [01:52<00:30, 1.17s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 77%|███████▋ | 82/107 [01:52<00:26, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 78%|███████▊ | 83/107 [01:54<00:25, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 79%|███████▊ | 84/107 [01:54<00:22, 1.02batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 79%|███████▉ | 85/107 [01:55<00:20, 1.08batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 80%|████████ | 86/107 [01:56<00:19, 1.10batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 81%|████████▏ | 87/107 [01:57<00:17, 1.14batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 82%|████████▏ | 88/107 [01:58<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 83%|████████▎ | 89/107 [01:58<00:15, 1.16batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 84%|████████▍ | 90/107 [02:00<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 85%|████████▌ | 91/107 [02:00<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 86%|████████▌ | 92/107 [02:02<00:18, 1.20s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 87%|████████▋ | 93/107 [02:03<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 88%|████████▊ | 94/107 [02:05<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 89%|████████▉ | 95/107 [02:09<00:25, 2.15s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 90%|████████▉ | 96/107 [02:10<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 91%|█████████ | 97/107 [02:12<00:20, 2.01s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 92%|█████████▏| 98/107 [02:14<00:15, 1.73s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 93%|█████████▎| 99/107 [02:15<00:12, 1.53s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 93%|█████████▎| 100/107 [02:16<00:09, 1.35s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 94%|█████████▍| 101/107 [02:16<00:07, 1.23s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 95%|█████████▌| 102/107 [02:18<00:06, 1.32s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 96%|█████████▋| 103/107 [02:19<00:04, 1.19s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 97%|█████████▋| 104/107 [02:20<00:03, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 98%|█████████▊| 105/107 [02:21<00:02, 1.03s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 99%|█████████▉| 106/107 [02:22<00:01, 1.10s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 52: 100%|██████████| 107/107 [02:27<00:00, 1.00s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 51%|█████▏ | 55/107 [00:54<00:23, 2.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 53: 52%|█████▏ | 56/107 [01:12<04:54, 5.78s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 53%|█████▎ | 57/107 [01:16<04:13, 5.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 54%|█████▍ | 58/107 [01:17<03:14, 3.97s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 55%|█████▌ | 59/107 [01:19<02:49, 3.52s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 56%|█████▌ | 60/107 [01:21<02:18, 2.94s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 57%|█████▋ | 61/107 [01:22<01:51, 2.43s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 58%|█████▊ | 62/107 [01:24<01:43, 2.30s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 59%|█████▉ | 63/107 [01:25<01:26, 1.96s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 60%|█████▉ | 64/107 [01:27<01:18, 1.83s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 61%|██████ | 65/107 [01:28<01:05, 1.56s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 62%|██████▏ | 66/107 [01:30<01:08, 1.67s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 63%|██████▎ | 67/107 [01:31<00:59, 1.48s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 64%|██████▎ | 68/107 [01:32<00:50, 1.30s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 64%|██████▍ | 69/107 [01:33<00:48, 1.28s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 65%|██████▌ | 70/107 [01:34<00:44, 1.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 66%|██████▋ | 71/107 [01:36<00:48, 1.34s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 67%|██████▋ | 72/107 [01:37<00:45, 1.31s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 68%|██████▊ | 73/107 [01:38<00:44, 1.30s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 69%|██████▉ | 74/107 [01:39<00:42, 1.28s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 70%|███████ | 75/107 [01:40<00:36, 1.14s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 71%|███████ | 76/107 [01:41<00:32, 1.05s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 72%|███████▏ | 77/107 [01:42<00:31, 1.04s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 73%|███████▎ | 78/107 [01:44<00:36, 1.27s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 74%|███████▍ | 79/107 [01:45<00:31, 1.13s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 75%|███████▍ | 80/107 [01:46<00:30, 1.12s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 76%|███████▌ | 81/107 [01:46<00:25, 1.02batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 77%|███████▋ | 82/107 [01:47<00:23, 1.06batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 78%|███████▊ | 83/107 [01:48<00:23, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 79%|███████▊ | 84/107 [01:49<00:21, 1.09batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 79%|███████▉ | 85/107 [01:50<00:19, 1.13batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 80%|████████ | 86/107 [01:51<00:18, 1.12batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 81%|████████▏ | 87/107 [01:52<00:17, 1.15batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 82%|████████▏ | 88/107 [01:52<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 83%|████████▎ | 89/107 [01:53<00:15, 1.14batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 84%|████████▍ | 90/107 [01:55<00:16, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 85%|████████▌ | 91/107 [01:55<00:14, 1.07batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 86%|████████▌ | 92/107 [01:57<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 87%|████████▋ | 93/107 [01:58<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 88%|████████▊ | 94/107 [02:00<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 89%|████████▉ | 95/107 [02:04<00:26, 2.17s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 90%|████████▉ | 96/107 [02:05<00:20, 1.84s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 91%|█████████ | 97/107 [02:07<00:19, 1.95s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 92%|█████████▏| 98/107 [02:08<00:14, 1.62s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 93%|█████████▎| 99/107 [02:09<00:11, 1.40s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 93%|█████████▎| 100/107 [02:10<00:08, 1.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 94%|█████████▍| 101/107 [02:11<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 95%|█████████▌| 102/107 [02:12<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 96%|█████████▋| 103/107 [02:13<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 97%|█████████▋| 104/107 [02:13<00:02, 1.05batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 98%|█████████▊| 105/107 [02:14<00:01, 1.07batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 99%|█████████▉| 106/107 [02:55<00:12, 12.76s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 53: 100%|██████████| 107/107 [02:59<00:00, 9.15s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 51%|█████▏ | 55/107 [00:53<00:26, 1.98batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 54: 52%|█████▏ | 56/107 [01:12<05:01, 5.91s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 53%|█████▎ | 57/107 [01:16<04:32, 5.46s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 54%|█████▍ | 58/107 [01:18<03:36, 4.43s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 55%|█████▌ | 59/107 [01:22<03:21, 4.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 56%|█████▌ | 60/107 [01:23<02:36, 3.34s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 57%|█████▋ | 61/107 [01:24<02:00, 2.62s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 58%|█████▊ | 62/107 [01:29<02:29, 3.33s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 59%|█████▉ | 63/107 [01:30<01:56, 2.65s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 60%|█████▉ | 64/107 [01:32<01:43, 2.40s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 61%|██████ | 65/107 [01:34<01:35, 2.27s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 62%|██████▏ | 66/107 [01:36<01:27, 2.14s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 63%|██████▎ | 67/107 [01:37<01:13, 1.84s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 64%|██████▎ | 68/107 [01:38<01:00, 1.55s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 64%|██████▍ | 69/107 [01:39<00:55, 1.45s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 65%|██████▌ | 70/107 [01:40<00:48, 1.32s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 66%|██████▋ | 71/107 [01:42<00:58, 1.63s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 67%|██████▋ | 72/107 [01:44<00:52, 1.49s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 68%|██████▊ | 73/107 [01:44<00:41, 1.23s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 69%|██████▉ | 74/107 [01:45<00:39, 1.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 70%|███████ | 75/107 [01:46<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 71%|███████ | 76/107 [01:47<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 72%|███████▏ | 77/107 [01:48<00:29, 1.02batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 73%|███████▎ | 78/107 [01:50<00:35, 1.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 74%|███████▍ | 79/107 [01:50<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 75%|███████▍ | 80/107 [01:52<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 76%|███████▌ | 81/107 [01:52<00:24, 1.07batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 77%|███████▋ | 82/107 [01:53<00:22, 1.12batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 78%|███████▊ | 83/107 [01:54<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 79%|███████▊ | 84/107 [01:55<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 79%|███████▉ | 85/107 [01:56<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 80%|████████ | 86/107 [01:57<00:19, 1.07batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 81%|████████▏ | 87/107 [01:58<00:19, 1.04batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 82%|████████▏ | 88/107 [01:59<00:17, 1.06batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 83%|████████▎ | 89/107 [02:00<00:17, 1.03batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 84%|████████▍ | 90/107 [02:01<00:18, 1.11s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 85%|████████▌ | 91/107 [02:02<00:16, 1.03s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 86%|████████▌ | 92/107 [02:04<00:19, 1.29s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 87%|████████▋ | 93/107 [02:05<00:16, 1.14s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 88%|████████▊ | 94/107 [02:06<00:16, 1.24s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 89%|████████▉ | 95/107 [02:10<00:25, 2.16s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 90%|████████▉ | 96/107 [02:11<00:20, 1.83s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 91%|█████████ | 97/107 [02:14<00:19, 1.93s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 92%|█████████▏| 98/107 [02:14<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 93%|█████████▎| 99/107 [02:15<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 93%|█████████▎| 100/107 [02:16<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 94%|█████████▍| 101/107 [02:17<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 95%|█████████▌| 102/107 [02:18<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 96%|█████████▋| 103/107 [02:19<00:03, 1.00batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 97%|█████████▋| 104/107 [02:19<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 98%|█████████▊| 105/107 [02:20<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 99%|█████████▉| 106/107 [02:21<00:00, 1.04batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 54: 100%|██████████| 107/107 [02:26<00:00, 1.14batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 51%|█████▏ | 55/107 [00:54<00:24, 2.14batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 55: 52%|█████▏ | 56/107 [01:13<05:15, 6.19s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 53%|█████▎ | 57/107 [01:17<04:34, 5.49s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 54%|█████▍ | 58/107 [01:18<03:27, 4.24s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 55%|█████▌ | 59/107 [01:24<03:47, 4.74s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 56%|█████▌ | 60/107 [01:26<02:59, 3.82s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 57%|█████▋ | 61/107 [01:27<02:18, 3.01s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 58%|█████▊ | 62/107 [01:29<02:03, 2.75s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 59%|█████▉ | 63/107 [01:30<01:39, 2.27s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 60%|█████▉ | 64/107 [01:31<01:22, 1.92s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 61%|██████ | 65/107 [01:32<01:07, 1.61s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 62%|██████▏ | 66/107 [01:34<01:04, 1.57s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 63%|██████▎ | 67/107 [01:35<00:57, 1.44s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 64%|██████▎ | 68/107 [01:36<00:50, 1.29s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 64%|██████▍ | 69/107 [01:37<00:48, 1.26s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 65%|██████▌ | 70/107 [01:38<00:46, 1.25s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 66%|██████▋ | 71/107 [01:40<00:49, 1.38s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 67%|██████▋ | 72/107 [01:42<00:51, 1.48s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 68%|██████▊ | 73/107 [01:42<00:41, 1.22s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 69%|██████▉ | 74/107 [01:43<00:39, 1.20s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 70%|███████ | 75/107 [01:44<00:34, 1.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 71%|███████ | 76/107 [01:45<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 72%|███████▏ | 77/107 [01:46<00:29, 1.02batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 73%|███████▎ | 78/107 [01:48<00:35, 1.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 74%|███████▍ | 79/107 [01:48<00:29, 1.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 75%|███████▍ | 80/107 [01:49<00:28, 1.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 76%|███████▌ | 81/107 [01:50<00:24, 1.08batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 77%|███████▋ | 82/107 [01:51<00:22, 1.13batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 78%|███████▊ | 83/107 [01:52<00:22, 1.08batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 79%|███████▊ | 84/107 [01:53<00:19, 1.16batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 79%|███████▉ | 85/107 [01:53<00:18, 1.20batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 80%|████████ | 86/107 [01:54<00:17, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 81%|████████▏ | 87/107 [01:55<00:16, 1.20batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 82%|████████▏ | 88/107 [01:56<00:15, 1.25batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 83%|████████▎ | 89/107 [01:57<00:14, 1.20batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 84%|████████▍ | 90/107 [01:58<00:16, 1.06batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 85%|████████▌ | 91/107 [01:59<00:14, 1.12batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 86%|████████▌ | 92/107 [02:00<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 87%|████████▋ | 93/107 [02:01<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 88%|████████▊ | 94/107 [02:03<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 89%|████████▉ | 95/107 [02:07<00:26, 2.17s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 90%|████████▉ | 96/107 [02:08<00:20, 1.86s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 91%|█████████ | 97/107 [02:11<00:20, 2.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 92%|█████████▏| 98/107 [02:12<00:15, 1.72s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 93%|█████████▎| 99/107 [02:13<00:11, 1.47s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 93%|█████████▎| 100/107 [02:13<00:08, 1.25s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 94%|█████████▍| 101/107 [02:14<00:06, 1.10s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 95%|█████████▌| 102/107 [02:16<00:05, 1.15s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 96%|█████████▋| 103/107 [02:16<00:04, 1.05s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 97%|█████████▋| 104/107 [02:17<00:02, 1.05batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 98%|█████████▊| 105/107 [02:18<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 99%|█████████▉| 106/107 [02:19<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 55: 100%|██████████| 107/107 [02:24<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 51%|█████▏ | 55/107 [00:56<00:31, 1.65batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 56: 52%|█████▏ | 56/107 [01:15<05:20, 6.29s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 53%|█████▎ | 57/107 [01:20<04:50, 5.80s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 54%|█████▍ | 58/107 [01:21<03:37, 4.43s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 55%|█████▌ | 59/107 [01:24<03:05, 3.87s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 56%|█████▌ | 60/107 [01:25<02:24, 3.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 57%|█████▋ | 61/107 [01:26<01:53, 2.47s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 58%|█████▊ | 62/107 [01:30<02:08, 2.85s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 59%|█████▉ | 63/107 [01:31<01:44, 2.37s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 60%|█████▉ | 64/107 [01:33<01:34, 2.20s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 61%|██████ | 65/107 [01:35<01:28, 2.11s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 62%|██████▏ | 66/107 [01:37<01:24, 2.05s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 63%|██████▎ | 67/107 [01:38<01:11, 1.79s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 64%|██████▎ | 68/107 [01:39<01:00, 1.55s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 64%|██████▍ | 69/107 [01:40<00:57, 1.50s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 65%|██████▌ | 70/107 [01:42<00:58, 1.58s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 66%|██████▋ | 71/107 [01:43<00:57, 1.59s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 67%|██████▋ | 72/107 [01:45<00:51, 1.46s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 68%|██████▊ | 73/107 [01:45<00:41, 1.22s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 69%|██████▉ | 74/107 [01:46<00:39, 1.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 70%|███████ | 75/107 [01:47<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 71%|███████ | 76/107 [01:48<00:30, 1.00batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 72%|███████▏ | 77/107 [01:49<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 73%|███████▎ | 78/107 [01:51<00:35, 1.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 74%|███████▍ | 79/107 [01:52<00:30, 1.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 75%|███████▍ | 80/107 [01:53<00:28, 1.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 76%|███████▌ | 81/107 [01:53<00:24, 1.08batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 77%|███████▋ | 82/107 [01:54<00:22, 1.13batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 78%|███████▊ | 83/107 [01:55<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 79%|███████▊ | 84/107 [01:56<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 79%|███████▉ | 85/107 [01:56<00:18, 1.20batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 80%|████████ | 86/107 [01:57<00:17, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 81%|████████▏ | 87/107 [01:58<00:16, 1.21batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 82%|████████▏ | 88/107 [01:59<00:15, 1.26batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 83%|████████▎ | 89/107 [02:00<00:14, 1.21batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 84%|████████▍ | 90/107 [02:01<00:16, 1.06batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 85%|████████▌ | 91/107 [02:02<00:14, 1.13batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 86%|████████▌ | 92/107 [02:03<00:17, 1.15s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 87%|████████▋ | 93/107 [02:04<00:14, 1.04s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 88%|████████▊ | 94/107 [02:06<00:14, 1.15s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 89%|████████▉ | 95/107 [02:10<00:25, 2.09s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 90%|████████▉ | 96/107 [02:11<00:19, 1.78s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 91%|█████████ | 97/107 [02:13<00:18, 1.89s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 92%|█████████▏| 98/107 [02:14<00:14, 1.57s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 93%|█████████▎| 99/107 [02:15<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 93%|█████████▎| 100/107 [02:16<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 94%|█████████▍| 101/107 [02:16<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 95%|█████████▌| 102/107 [02:18<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 96%|█████████▋| 103/107 [02:18<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 97%|█████████▋| 104/107 [02:19<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 98%|█████████▊| 105/107 [02:20<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 99%|█████████▉| 106/107 [02:21<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 56: 100%|██████████| 107/107 [02:26<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 51%|█████▏ | 55/107 [00:56<00:27, 1.89batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 57: 52%|█████▏ | 56/107 [01:16<05:23, 6.35s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 53%|█████▎ | 57/107 [01:19<04:24, 5.29s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 54%|█████▍ | 58/107 [01:20<03:18, 4.06s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 55%|█████▌ | 59/107 [01:23<02:50, 3.55s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 56%|█████▌ | 60/107 [01:24<02:14, 2.86s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 57%|█████▋ | 61/107 [01:25<01:44, 2.26s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 58%|█████▊ | 62/107 [01:29<02:12, 2.93s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 59%|█████▉ | 63/107 [01:30<01:46, 2.42s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 60%|█████▉ | 64/107 [01:32<01:32, 2.15s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 61%|██████ | 65/107 [01:33<01:14, 1.77s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 62%|██████▏ | 66/107 [01:35<01:21, 1.98s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 63%|██████▎ | 67/107 [01:37<01:17, 1.94s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 64%|██████▎ | 68/107 [01:38<01:04, 1.64s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 64%|██████▍ | 69/107 [01:39<00:58, 1.55s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 65%|██████▌ | 70/107 [01:40<00:51, 1.39s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 66%|██████▋ | 71/107 [01:43<01:00, 1.69s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 67%|██████▋ | 72/107 [01:44<00:53, 1.53s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 68%|██████▊ | 73/107 [01:45<00:42, 1.26s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 69%|██████▉ | 74/107 [01:46<00:40, 1.23s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 70%|███████ | 75/107 [01:47<00:35, 1.10s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 71%|███████ | 76/107 [01:47<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 72%|███████▏ | 77/107 [01:48<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 73%|███████▎ | 78/107 [01:50<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 74%|███████▍ | 79/107 [01:51<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 75%|███████▍ | 80/107 [01:52<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 76%|███████▌ | 81/107 [01:53<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 77%|███████▋ | 82/107 [01:53<00:22, 1.12batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 78%|███████▊ | 83/107 [01:54<00:22, 1.08batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 79%|███████▊ | 84/107 [01:55<00:19, 1.15batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 79%|███████▉ | 85/107 [01:56<00:18, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 80%|████████ | 86/107 [01:57<00:17, 1.18batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 81%|████████▏ | 87/107 [01:58<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 82%|████████▏ | 88/107 [01:58<00:16, 1.16batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 83%|████████▎ | 89/107 [02:00<00:16, 1.06batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 84%|████████▍ | 90/107 [02:01<00:18, 1.10s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 85%|████████▌ | 91/107 [02:02<00:16, 1.05s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 86%|████████▌ | 92/107 [02:04<00:19, 1.32s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 87%|████████▋ | 93/107 [02:05<00:16, 1.19s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 88%|████████▊ | 94/107 [02:06<00:16, 1.27s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 89%|████████▉ | 95/107 [02:11<00:26, 2.17s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 90%|████████▉ | 96/107 [02:12<00:20, 1.84s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 91%|█████████ | 97/107 [02:14<00:19, 1.95s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 92%|█████████▏| 98/107 [02:15<00:14, 1.61s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 93%|█████████▎| 99/107 [02:15<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 93%|█████████▎| 100/107 [02:16<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 94%|█████████▍| 101/107 [02:17<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 95%|█████████▌| 102/107 [02:18<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 96%|█████████▋| 103/107 [02:19<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 97%|█████████▋| 104/107 [02:20<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 98%|█████████▊| 105/107 [02:20<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 99%|█████████▉| 106/107 [02:22<00:00, 1.04batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 57: 100%|██████████| 107/107 [02:26<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 51%|█████▏ | 55/107 [00:54<00:19, 2.60batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 58: 52%|█████▏ | 56/107 [01:14<05:18, 6.24s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 53%|█████▎ | 57/107 [01:19<04:45, 5.71s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 54%|█████▍ | 58/107 [01:20<03:36, 4.42s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 55%|█████▌ | 59/107 [01:22<02:50, 3.55s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 56%|█████▌ | 60/107 [01:24<02:24, 3.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 57%|█████▋ | 61/107 [01:25<01:54, 2.50s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 58%|█████▊ | 62/107 [01:29<02:20, 3.13s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 59%|█████▉ | 63/107 [01:30<01:50, 2.51s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 60%|█████▉ | 64/107 [01:32<01:33, 2.18s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 61%|██████ | 65/107 [01:33<01:17, 1.85s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 62%|██████▏ | 66/107 [01:35<01:16, 1.86s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 63%|██████▎ | 67/107 [01:36<01:04, 1.61s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 64%|██████▎ | 68/107 [01:37<00:54, 1.40s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 64%|██████▍ | 69/107 [01:38<00:52, 1.38s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 65%|██████▌ | 70/107 [01:40<00:55, 1.49s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 66%|██████▋ | 71/107 [01:41<00:54, 1.52s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 67%|██████▋ | 72/107 [01:43<00:49, 1.42s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 68%|██████▊ | 73/107 [01:43<00:40, 1.19s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 69%|██████▉ | 74/107 [01:44<00:39, 1.18s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 70%|███████ | 75/107 [01:45<00:33, 1.06s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 71%|███████ | 76/107 [01:46<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 72%|███████▏ | 77/107 [01:47<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 73%|███████▎ | 78/107 [01:49<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 74%|███████▍ | 79/107 [01:50<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 75%|███████▍ | 80/107 [01:51<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 76%|███████▌ | 81/107 [01:51<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 77%|███████▋ | 82/107 [01:52<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 78%|███████▊ | 83/107 [01:53<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 79%|███████▊ | 84/107 [01:54<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 79%|███████▉ | 85/107 [01:55<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 80%|████████ | 86/107 [01:55<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 81%|████████▏ | 87/107 [01:56<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 82%|████████▏ | 88/107 [01:57<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 83%|████████▎ | 89/107 [01:58<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 84%|████████▍ | 90/107 [01:59<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 85%|████████▌ | 91/107 [02:00<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 86%|████████▌ | 92/107 [02:02<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 87%|████████▋ | 93/107 [02:03<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 88%|████████▊ | 94/107 [02:04<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 89%|████████▉ | 95/107 [02:08<00:25, 2.13s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 90%|████████▉ | 96/107 [02:09<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 91%|█████████ | 97/107 [02:12<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 92%|█████████▏| 98/107 [02:12<00:14, 1.60s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 93%|█████████▎| 99/107 [02:13<00:11, 1.39s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 93%|█████████▎| 100/107 [02:14<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 94%|█████████▍| 101/107 [02:15<00:06, 1.06s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 95%|█████████▌| 102/107 [02:16<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 96%|█████████▋| 103/107 [02:17<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 97%|█████████▋| 104/107 [02:18<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 98%|█████████▊| 105/107 [02:18<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 99%|█████████▉| 106/107 [02:20<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 58: 100%|██████████| 107/107 [02:25<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 51%|█████▏ | 55/107 [00:56<00:32, 1.61batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 59: 52%|█████▏ | 56/107 [01:15<05:16, 6.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 53%|█████▎ | 57/107 [01:19<04:26, 5.33s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 54%|█████▍ | 58/107 [01:21<03:38, 4.46s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 55%|█████▌ | 59/107 [01:23<02:59, 3.74s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 56%|█████▌ | 60/107 [01:25<02:24, 3.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 57%|█████▋ | 61/107 [01:26<01:54, 2.49s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 58%|█████▊ | 62/107 [01:28<01:44, 2.33s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 59%|█████▉ | 63/107 [01:29<01:27, 1.98s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 60%|█████▉ | 64/107 [01:30<01:18, 1.82s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 61%|██████ | 65/107 [01:32<01:08, 1.63s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 62%|██████▏ | 66/107 [01:33<01:09, 1.69s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 63%|██████▎ | 67/107 [01:34<01:00, 1.50s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 64%|██████▎ | 68/107 [01:35<00:51, 1.32s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 64%|██████▍ | 69/107 [01:37<00:49, 1.30s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 65%|██████▌ | 70/107 [01:38<00:45, 1.23s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 66%|██████▋ | 71/107 [01:40<00:54, 1.51s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 67%|██████▋ | 72/107 [01:41<00:49, 1.41s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 68%|██████▊ | 73/107 [01:42<00:40, 1.18s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 69%|██████▉ | 74/107 [01:43<00:38, 1.18s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 70%|███████ | 75/107 [01:44<00:33, 1.06s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 71%|███████ | 76/107 [01:44<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 72%|███████▏ | 77/107 [01:45<00:29, 1.00batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 73%|███████▎ | 78/107 [01:47<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 74%|███████▍ | 79/107 [01:48<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 75%|███████▍ | 80/107 [01:49<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 76%|███████▌ | 81/107 [01:50<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 77%|███████▋ | 82/107 [01:50<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 78%|███████▊ | 83/107 [01:52<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 79%|███████▊ | 84/107 [01:52<00:20, 1.11batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 79%|███████▉ | 85/107 [01:53<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 80%|████████ | 86/107 [01:54<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 81%|████████▏ | 87/107 [01:55<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 82%|████████▏ | 88/107 [01:55<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 83%|████████▎ | 89/107 [01:56<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 84%|████████▍ | 90/107 [01:58<00:16, 1.05batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 85%|████████▌ | 91/107 [01:58<00:14, 1.12batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 86%|████████▌ | 92/107 [02:00<00:17, 1.16s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 87%|████████▋ | 93/107 [02:01<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 88%|████████▊ | 94/107 [02:02<00:15, 1.16s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 89%|████████▉ | 95/107 [02:07<00:25, 2.11s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 90%|████████▉ | 96/107 [02:08<00:19, 1.79s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 91%|█████████ | 97/107 [02:10<00:18, 1.90s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 92%|█████████▏| 98/107 [02:11<00:14, 1.57s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 93%|█████████▎| 99/107 [02:12<00:10, 1.35s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 93%|█████████▎| 100/107 [02:12<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 94%|█████████▍| 101/107 [02:13<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 95%|█████████▌| 102/107 [02:14<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 96%|█████████▋| 103/107 [02:15<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 97%|█████████▋| 104/107 [02:16<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 98%|█████████▊| 105/107 [02:17<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 99%|█████████▉| 106/107 [02:18<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 59: 100%|██████████| 107/107 [02:23<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 51%|█████▏ | 55/107 [00:56<00:20, 2.52batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 60: 52%|█████▏ | 56/107 [01:16<05:19, 6.27s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 53%|█████▎ | 57/107 [01:20<04:30, 5.41s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 54%|█████▍ | 58/107 [01:21<03:31, 4.32s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 55%|█████▌ | 59/107 [01:24<02:59, 3.74s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 56%|█████▌ | 60/107 [01:25<02:22, 3.03s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 57%|█████▋ | 61/107 [01:26<01:49, 2.38s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 58%|█████▊ | 62/107 [01:30<02:16, 3.03s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 59%|█████▉ | 63/107 [01:32<01:48, 2.48s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 60%|█████▉ | 64/107 [01:33<01:34, 2.20s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 61%|██████ | 65/107 [01:34<01:18, 1.86s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 62%|██████▏ | 66/107 [01:36<01:14, 1.82s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 63%|██████▎ | 67/107 [01:37<01:06, 1.66s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 64%|██████▎ | 68/107 [01:38<00:55, 1.43s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 64%|██████▍ | 69/107 [01:40<00:53, 1.40s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 65%|██████▌ | 70/107 [01:41<00:48, 1.32s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 66%|██████▋ | 71/107 [01:43<00:58, 1.62s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 67%|██████▋ | 72/107 [01:44<00:52, 1.49s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 68%|██████▊ | 73/107 [01:45<00:41, 1.23s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 69%|██████▉ | 74/107 [01:46<00:40, 1.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 70%|███████ | 75/107 [01:47<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 71%|███████ | 76/107 [01:48<00:31, 1.00s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 72%|███████▏ | 77/107 [01:49<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 73%|███████▎ | 78/107 [01:50<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 74%|███████▍ | 79/107 [01:51<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 75%|███████▍ | 80/107 [01:52<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 76%|███████▌ | 81/107 [01:53<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 77%|███████▋ | 82/107 [01:54<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 78%|███████▊ | 83/107 [01:55<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 79%|███████▊ | 84/107 [01:55<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 79%|███████▉ | 85/107 [01:56<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 80%|████████ | 86/107 [01:57<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 81%|████████▏ | 87/107 [01:58<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 82%|████████▏ | 88/107 [01:59<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 83%|████████▎ | 89/107 [01:59<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 84%|████████▍ | 90/107 [02:01<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 85%|████████▌ | 91/107 [02:02<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 86%|████████▌ | 92/107 [02:03<00:18, 1.20s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 87%|████████▋ | 93/107 [02:04<00:15, 1.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 88%|████████▊ | 94/107 [02:06<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 89%|████████▉ | 95/107 [02:10<00:25, 2.16s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 90%|████████▉ | 96/107 [02:11<00:20, 1.84s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 91%|█████████ | 97/107 [02:13<00:19, 1.96s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 92%|█████████▏| 98/107 [02:14<00:14, 1.62s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 93%|█████████▎| 99/107 [02:15<00:11, 1.40s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 93%|█████████▎| 100/107 [02:16<00:08, 1.20s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 94%|█████████▍| 101/107 [02:17<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 95%|█████████▌| 102/107 [02:18<00:05, 1.15s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 96%|█████████▋| 103/107 [02:19<00:04, 1.04s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 97%|█████████▋| 104/107 [02:19<00:02, 1.05batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 98%|█████████▊| 105/107 [02:20<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 99%|█████████▉| 106/107 [02:21<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 60: 100%|██████████| 107/107 [02:26<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 51%|█████▏ | 55/107 [00:57<00:25, 2.03batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 61: 52%|█████▏ | 56/107 [01:18<05:39, 6.66s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 53%|█████▎ | 57/107 [01:21<04:35, 5.52s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 54%|█████▍ | 58/107 [01:22<03:29, 4.28s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 55%|█████▌ | 59/107 [01:25<03:01, 3.79s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 56%|█████▌ | 60/107 [01:27<02:27, 3.15s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 57%|█████▋ | 61/107 [01:28<01:57, 2.55s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 58%|█████▊ | 62/107 [01:30<01:49, 2.43s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 59%|█████▉ | 63/107 [01:31<01:28, 2.02s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 60%|█████▉ | 64/107 [01:33<01:22, 1.91s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 61%|██████ | 65/107 [01:34<01:11, 1.70s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 62%|██████▏ | 66/107 [01:35<01:07, 1.64s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 63%|██████▎ | 67/107 [01:37<01:00, 1.52s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 64%|██████▎ | 68/107 [01:38<00:51, 1.33s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 64%|██████▍ | 69/107 [01:39<00:49, 1.32s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 65%|██████▌ | 70/107 [01:40<00:45, 1.24s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 66%|██████▋ | 71/107 [01:42<00:50, 1.40s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 67%|██████▋ | 72/107 [01:44<00:53, 1.52s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 68%|██████▊ | 73/107 [01:44<00:43, 1.26s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 69%|██████▉ | 74/107 [01:45<00:40, 1.24s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 70%|███████ | 75/107 [01:46<00:35, 1.11s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 71%|███████ | 76/107 [01:47<00:31, 1.03s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 72%|███████▏ | 77/107 [01:48<00:30, 1.02s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 73%|███████▎ | 78/107 [01:50<00:36, 1.24s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 74%|███████▍ | 79/107 [01:51<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 75%|███████▍ | 80/107 [01:52<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 76%|███████▌ | 81/107 [01:52<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 77%|███████▋ | 82/107 [01:53<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 78%|███████▊ | 83/107 [01:54<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 79%|███████▊ | 84/107 [01:55<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 79%|███████▉ | 85/107 [01:56<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 80%|████████ | 86/107 [01:56<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 81%|████████▏ | 87/107 [01:57<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 82%|████████▏ | 88/107 [01:58<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 83%|████████▎ | 89/107 [01:59<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 84%|████████▍ | 90/107 [02:00<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 85%|████████▌ | 91/107 [02:01<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 86%|████████▌ | 92/107 [02:03<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 87%|████████▋ | 93/107 [02:04<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 88%|████████▊ | 94/107 [02:05<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 89%|████████▉ | 95/107 [02:10<00:25, 2.16s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 90%|████████▉ | 96/107 [02:11<00:20, 1.82s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 91%|█████████ | 97/107 [02:13<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 92%|█████████▏| 98/107 [02:14<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 93%|█████████▎| 99/107 [02:14<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 93%|█████████▎| 100/107 [02:15<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 94%|█████████▍| 101/107 [02:16<00:06, 1.03s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 95%|█████████▌| 102/107 [02:17<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 96%|█████████▋| 103/107 [02:18<00:03, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 97%|█████████▋| 104/107 [02:19<00:02, 1.10batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 98%|█████████▊| 105/107 [02:19<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 99%|█████████▉| 106/107 [02:21<00:00, 1.04batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 61: 100%|██████████| 107/107 [02:26<00:00, 1.14batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 51%|█████▏ | 55/107 [00:56<00:29, 1.77batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 62: 52%|█████▏ | 56/107 [01:17<05:32, 6.52s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 53%|█████▎ | 57/107 [01:20<04:37, 5.54s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 54%|█████▍ | 58/107 [01:21<03:29, 4.27s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 55%|█████▌ | 59/107 [01:24<03:04, 3.84s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 56%|█████▌ | 60/107 [01:26<02:27, 3.13s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 57%|█████▋ | 61/107 [01:27<01:55, 2.51s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 58%|█████▊ | 62/107 [01:29<01:45, 2.34s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 59%|█████▉ | 63/107 [01:30<01:26, 1.98s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 60%|█████▉ | 64/107 [01:32<01:19, 1.86s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 61%|██████ | 65/107 [01:33<01:09, 1.65s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 62%|██████▏ | 66/107 [01:34<01:09, 1.70s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 63%|██████▎ | 67/107 [01:35<00:59, 1.49s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 64%|██████▎ | 68/107 [01:36<00:51, 1.31s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 64%|██████▍ | 69/107 [01:38<00:48, 1.28s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 65%|██████▌ | 70/107 [01:39<00:44, 1.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 66%|██████▋ | 71/107 [01:41<00:54, 1.52s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 67%|██████▋ | 72/107 [01:42<00:49, 1.42s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 68%|██████▊ | 73/107 [01:43<00:40, 1.19s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 69%|██████▉ | 74/107 [01:44<00:39, 1.19s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 70%|███████ | 75/107 [01:45<00:33, 1.06s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 71%|███████ | 76/107 [01:45<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 72%|███████▏ | 77/107 [01:46<00:29, 1.02batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 73%|███████▎ | 78/107 [01:48<00:35, 1.21s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 74%|███████▍ | 79/107 [01:49<00:30, 1.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 75%|███████▍ | 80/107 [01:50<00:28, 1.07s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 76%|███████▌ | 81/107 [01:51<00:24, 1.07batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 77%|███████▋ | 82/107 [01:51<00:22, 1.13batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 78%|███████▊ | 83/107 [01:52<00:22, 1.08batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 79%|███████▊ | 84/107 [01:53<00:19, 1.15batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 79%|███████▉ | 85/107 [01:54<00:18, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 80%|████████ | 86/107 [01:55<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 81%|████████▏ | 87/107 [01:56<00:16, 1.20batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 82%|████████▏ | 88/107 [01:56<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 83%|████████▎ | 89/107 [01:57<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 84%|████████▍ | 90/107 [01:59<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 85%|████████▌ | 91/107 [01:59<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 86%|████████▌ | 92/107 [02:01<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 87%|████████▋ | 93/107 [02:02<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 88%|████████▊ | 94/107 [02:03<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 89%|████████▉ | 95/107 [02:08<00:25, 2.10s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 90%|████████▉ | 96/107 [02:09<00:19, 1.78s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 91%|█████████ | 97/107 [02:11<00:18, 1.90s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 92%|█████████▏| 98/107 [02:12<00:14, 1.57s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 93%|█████████▎| 99/107 [02:12<00:10, 1.36s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 93%|█████████▎| 100/107 [02:13<00:08, 1.16s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 94%|█████████▍| 101/107 [02:14<00:06, 1.03s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 95%|█████████▌| 102/107 [02:15<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 96%|█████████▋| 103/107 [02:16<00:03, 1.01batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 97%|█████████▋| 104/107 [02:17<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 98%|█████████▊| 105/107 [02:17<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 99%|█████████▉| 106/107 [02:19<00:00, 1.04batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 62: 100%|██████████| 107/107 [02:24<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.022, v_num=zf852wxn]\nEpoch 63: 51%|█████▏ | 55/107 [00:56<00:29, 1.79batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 63: 52%|█████▏ | 56/107 [01:17<05:44, 6.76s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 53%|█████▎ | 57/107 [01:20<04:40, 5.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 54%|█████▍ | 58/107 [01:21<03:32, 4.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 55%|█████▌ | 59/107 [01:24<03:06, 3.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 56%|█████▌ | 60/107 [01:26<02:29, 3.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 57%|█████▋ | 61/107 [01:27<01:58, 2.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 58%|█████▊ | 62/107 [01:29<01:48, 2.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 59%|█████▉ | 63/107 [01:30<01:28, 2.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 60%|█████▉ | 64/107 [01:32<01:23, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 61%|██████ | 65/107 [01:32<01:07, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 62%|██████▏ | 66/107 [01:34<01:06, 1.63s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 63%|██████▎ | 67/107 [01:35<00:58, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 64%|██████▎ | 68/107 [01:36<00:50, 1.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 64%|██████▍ | 69/107 [01:37<00:49, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 65%|██████▌ | 70/107 [01:39<00:45, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 66%|██████▋ | 71/107 [01:40<00:48, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 67%|██████▋ | 72/107 [01:42<00:51, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 68%|██████▊ | 73/107 [01:43<00:41, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 69%|██████▉ | 74/107 [01:44<00:40, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 70%|███████ | 75/107 [01:45<00:35, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 71%|███████ | 76/107 [01:45<00:31, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 72%|███████▏ | 77/107 [01:46<00:30, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 73%|███████▎ | 78/107 [01:48<00:36, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 74%|███████▍ | 79/107 [01:49<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 75%|███████▍ | 80/107 [01:50<00:30, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 76%|███████▌ | 81/107 [01:51<00:25, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 77%|███████▋ | 82/107 [01:52<00:23, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 78%|███████▊ | 83/107 [01:53<00:23, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 79%|███████▊ | 84/107 [01:54<00:21, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 79%|███████▉ | 85/107 [01:54<00:19, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 80%|████████ | 86/107 [01:55<00:18, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 81%|████████▏ | 87/107 [01:56<00:17, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 82%|████████▏ | 88/107 [01:57<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 83%|████████▎ | 89/107 [01:58<00:15, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 84%|████████▍ | 90/107 [01:59<00:16, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 85%|████████▌ | 91/107 [02:00<00:15, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 86%|████████▌ | 92/107 [02:02<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 87%|████████▋ | 93/107 [02:03<00:15, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 88%|████████▊ | 94/107 [02:04<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 89%|████████▉ | 95/107 [02:08<00:25, 2.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 90%|████████▉ | 96/107 [02:09<00:20, 1.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 91%|█████████ | 97/107 [02:12<00:19, 1.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 92%|█████████▏| 98/107 [02:13<00:14, 1.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 93%|█████████▎| 99/107 [02:13<00:11, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 93%|█████████▎| 100/107 [02:14<00:08, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 94%|█████████▍| 101/107 [02:15<00:06, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 95%|█████████▌| 102/107 [02:16<00:05, 1.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 96%|█████████▋| 103/107 [02:17<00:04, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 97%|█████████▋| 104/107 [02:18<00:02, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 98%|█████████▊| 105/107 [02:19<00:01, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 99%|█████████▉| 106/107 [02:20<00:01, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 63: 100%|██████████| 107/107 [02:25<00:00, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 51%|█████▏ | 55/107 [00:57<00:23, 2.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 64: 52%|█████▏ | 56/107 [01:18<05:38, 6.63s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 53%|█████▎ | 57/107 [01:21<04:45, 5.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 54%|█████▍ | 58/107 [01:23<03:36, 4.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 55%|█████▌ | 59/107 [01:25<02:58, 3.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 56%|█████▌ | 60/107 [01:26<02:26, 3.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 57%|█████▋ | 61/107 [01:28<01:56, 2.53s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 58%|█████▊ | 62/107 [01:30<01:47, 2.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 59%|█████▉ | 63/107 [01:31<01:30, 2.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 60%|█████▉ | 64/107 [02:22<12:00, 16.75s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 61%|██████ | 65/107 [02:23<08:24, 12.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 62%|██████▏ | 66/107 [02:25<06:03, 8.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 63%|██████▎ | 67/107 [02:26<04:27, 6.69s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 64%|██████▎ | 68/107 [02:27<03:12, 4.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 64%|██████▍ | 69/107 [02:28<02:24, 3.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 65%|██████▌ | 70/107 [02:29<01:49, 2.97s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 66%|██████▋ | 71/107 [02:31<01:32, 2.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 67%|██████▋ | 72/107 [02:32<01:15, 2.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 68%|██████▊ | 73/107 [02:33<00:58, 1.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 69%|██████▉ | 74/107 [02:34<00:51, 1.56s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 70%|███████ | 75/107 [02:35<00:42, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 71%|███████ | 76/107 [02:36<00:36, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 72%|███████▏ | 77/107 [02:36<00:33, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 73%|███████▎ | 78/107 [02:38<00:37, 1.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 74%|███████▍ | 79/107 [02:39<00:31, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 75%|███████▍ | 80/107 [02:40<00:30, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 76%|███████▌ | 81/107 [02:41<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 77%|███████▋ | 82/107 [02:41<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 78%|███████▊ | 83/107 [02:42<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 79%|███████▊ | 84/107 [02:43<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 79%|███████▉ | 85/107 [02:44<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 80%|████████ | 86/107 [02:45<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 81%|████████▏ | 87/107 [02:46<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 82%|████████▏ | 88/107 [02:46<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 83%|████████▎ | 89/107 [02:47<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 84%|████████▍ | 90/107 [02:49<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 85%|████████▌ | 91/107 [02:49<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 86%|████████▌ | 92/107 [02:51<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 87%|████████▋ | 93/107 [02:52<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 88%|████████▊ | 94/107 [02:53<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 89%|████████▉ | 95/107 [02:58<00:25, 2.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 90%|████████▉ | 96/107 [02:59<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 91%|█████████ | 97/107 [03:01<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 92%|█████████▏| 98/107 [03:02<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 93%|█████████▎| 99/107 [03:03<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 93%|█████████▎| 100/107 [03:03<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 94%|█████████▍| 101/107 [03:04<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 95%|█████████▌| 102/107 [03:05<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 96%|█████████▋| 103/107 [03:06<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 97%|█████████▋| 104/107 [03:07<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 98%|█████████▊| 105/107 [03:08<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 99%|█████████▉| 106/107 [03:09<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 64: 100%|██████████| 107/107 [03:14<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 51%|█████▏ | 55/107 [00:58<00:33, 1.54batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 65: 52%|█████▏ | 56/107 [01:19<05:43, 6.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 53%|█████▎ | 57/107 [01:22<04:46, 5.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 54%|█████▍ | 58/107 [01:23<03:33, 4.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 55%|█████▌ | 59/107 [01:27<03:23, 4.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 56%|█████▌ | 60/107 [01:29<02:37, 3.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 57%|█████▋ | 61/107 [01:30<02:01, 2.65s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 58%|█████▊ | 62/107 [01:33<02:13, 2.96s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 59%|█████▉ | 63/107 [01:34<01:44, 2.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 60%|█████▉ | 64/107 [01:36<01:31, 2.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 61%|██████ | 65/107 [01:37<01:17, 1.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 62%|██████▏ | 66/107 [01:39<01:14, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 63%|██████▎ | 67/107 [01:40<01:07, 1.69s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 64%|██████▎ | 68/107 [01:41<00:56, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 64%|██████▍ | 69/107 [01:43<00:59, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 65%|██████▌ | 70/107 [01:44<00:52, 1.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 66%|██████▋ | 71/107 [01:46<00:53, 1.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 67%|██████▋ | 72/107 [01:47<00:49, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 68%|██████▊ | 73/107 [01:47<00:39, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 69%|██████▉ | 74/107 [01:49<00:38, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 70%|███████ | 75/107 [01:49<00:33, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 71%|███████ | 76/107 [01:50<00:30, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 72%|███████▏ | 77/107 [01:51<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 73%|███████▎ | 78/107 [01:53<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 74%|███████▍ | 79/107 [01:54<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 75%|███████▍ | 80/107 [01:55<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 76%|███████▌ | 81/107 [01:56<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 77%|███████▋ | 82/107 [01:56<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 78%|███████▊ | 83/107 [01:57<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 79%|███████▊ | 84/107 [01:58<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 79%|███████▉ | 85/107 [01:59<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 80%|████████ | 86/107 [02:00<00:18, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 81%|████████▏ | 87/107 [02:01<00:17, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 82%|████████▏ | 88/107 [02:01<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 83%|████████▎ | 89/107 [02:02<00:15, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 84%|████████▍ | 90/107 [02:04<00:16, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 85%|████████▌ | 91/107 [02:04<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 86%|████████▌ | 92/107 [02:06<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 87%|████████▋ | 93/107 [02:07<00:15, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 88%|████████▊ | 94/107 [02:08<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 89%|████████▉ | 95/107 [02:13<00:25, 2.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 90%|████████▉ | 96/107 [02:14<00:20, 1.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 91%|█████████ | 97/107 [02:16<00:19, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 92%|█████████▏| 98/107 [02:17<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 93%|█████████▎| 99/107 [02:18<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 93%|█████████▎| 100/107 [02:19<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 94%|█████████▍| 101/107 [02:19<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 95%|█████████▌| 102/107 [02:21<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 96%|█████████▋| 103/107 [02:21<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 97%|█████████▋| 104/107 [02:22<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 98%|█████████▊| 105/107 [02:23<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 99%|█████████▉| 106/107 [02:24<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 65: 100%|██████████| 107/107 [02:29<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 51%|█████▏ | 55/107 [00:57<00:24, 2.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 66: 52%|█████▏ | 56/107 [01:19<05:52, 6.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 53%|█████▎ | 57/107 [01:23<05:12, 6.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 54%|█████▍ | 58/107 [01:25<03:50, 4.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 55%|█████▌ | 59/107 [01:27<03:15, 4.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 56%|█████▌ | 60/107 [01:29<02:35, 3.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 57%|█████▋ | 61/107 [01:30<02:00, 2.63s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 58%|█████▊ | 62/107 [01:34<02:15, 3.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 59%|█████▉ | 63/107 [01:35<01:47, 2.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 60%|█████▉ | 64/107 [01:36<01:33, 2.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 61%|██████ | 65/107 [01:38<01:19, 1.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 62%|██████▏ | 66/107 [01:39<01:13, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 63%|██████▎ | 67/107 [01:40<01:04, 1.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 64%|██████▎ | 68/107 [01:41<00:54, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 64%|██████▍ | 69/107 [01:43<00:53, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 65%|██████▌ | 70/107 [01:44<00:53, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 66%|██████▋ | 71/107 [01:46<00:53, 1.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 67%|██████▋ | 72/107 [01:47<00:49, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 68%|██████▊ | 73/107 [01:48<00:40, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 69%|██████▉ | 74/107 [01:49<00:39, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 70%|███████ | 75/107 [01:50<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 71%|███████ | 76/107 [01:50<00:30, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 72%|███████▏ | 77/107 [01:51<00:29, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 73%|███████▎ | 78/107 [01:53<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 74%|███████▍ | 79/107 [01:54<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 75%|███████▍ | 80/107 [01:55<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 76%|███████▌ | 81/107 [01:56<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 77%|███████▋ | 82/107 [01:56<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 78%|███████▊ | 83/107 [01:58<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 79%|███████▊ | 84/107 [01:58<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 79%|███████▉ | 85/107 [01:59<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 80%|████████ | 86/107 [02:00<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 81%|████████▏ | 87/107 [02:01<00:17, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 82%|████████▏ | 88/107 [02:01<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 83%|████████▎ | 89/107 [02:02<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 84%|████████▍ | 90/107 [02:04<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 85%|████████▌ | 91/107 [02:04<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 86%|████████▌ | 92/107 [02:06<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 87%|████████▋ | 93/107 [02:07<00:15, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 88%|████████▊ | 94/107 [02:09<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 89%|████████▉ | 95/107 [02:13<00:25, 2.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 90%|████████▉ | 96/107 [02:14<00:19, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 91%|█████████ | 97/107 [02:16<00:19, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 92%|█████████▏| 98/107 [02:17<00:14, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 93%|█████████▎| 99/107 [02:18<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 93%|█████████▎| 100/107 [02:19<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 94%|█████████▍| 101/107 [02:19<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 95%|█████████▌| 102/107 [02:21<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 96%|█████████▋| 103/107 [02:21<00:04, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 97%|█████████▋| 104/107 [02:22<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 98%|█████████▊| 105/107 [02:23<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 99%|█████████▉| 106/107 [02:24<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 66: 100%|██████████| 107/107 [02:29<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 51%|█████▏ | 55/107 [01:00<00:36, 1.44batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 67: 52%|█████▏ | 56/107 [01:23<06:25, 7.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 53%|█████▎ | 57/107 [01:26<05:05, 6.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 54%|█████▍ | 58/107 [01:27<03:47, 4.64s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 55%|█████▌ | 59/107 [01:30<03:12, 4.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 56%|█████▌ | 60/107 [01:31<02:31, 3.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 57%|█████▋ | 61/107 [01:32<01:57, 2.56s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 58%|█████▊ | 62/107 [01:36<02:09, 2.87s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 59%|█████▉ | 63/107 [01:37<01:41, 2.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 60%|█████▉ | 64/107 [01:38<01:29, 2.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 61%|██████ | 65/107 [01:40<01:15, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 62%|██████▏ | 66/107 [01:41<01:13, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 63%|██████▎ | 67/107 [01:43<01:06, 1.65s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 64%|██████▎ | 68/107 [01:44<00:55, 1.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 64%|██████▍ | 69/107 [01:45<00:51, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 65%|██████▌ | 70/107 [01:46<00:54, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 66%|██████▋ | 71/107 [01:48<00:53, 1.49s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 67%|██████▋ | 72/107 [01:49<00:48, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 68%|██████▊ | 73/107 [01:50<00:40, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 69%|██████▉ | 74/107 [01:51<00:39, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 70%|███████ | 75/107 [01:52<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 71%|███████ | 76/107 [01:53<00:31, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 72%|███████▏ | 77/107 [01:54<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 73%|███████▎ | 78/107 [01:56<00:36, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 74%|███████▍ | 79/107 [01:56<00:31, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 75%|███████▍ | 80/107 [01:57<00:29, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 76%|███████▌ | 81/107 [01:58<00:25, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 77%|███████▋ | 82/107 [01:59<00:23, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 78%|███████▊ | 83/107 [02:00<00:23, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 79%|███████▊ | 84/107 [02:01<00:20, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 79%|███████▉ | 85/107 [02:02<00:19, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 80%|████████ | 86/107 [02:02<00:18, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 81%|████████▏ | 87/107 [02:03<00:17, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 82%|████████▏ | 88/107 [02:04<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 83%|████████▎ | 89/107 [02:05<00:15, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 84%|████████▍ | 90/107 [02:06<00:16, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 85%|████████▌ | 91/107 [02:07<00:14, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 86%|████████▌ | 92/107 [02:09<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 87%|████████▋ | 93/107 [02:10<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 88%|████████▊ | 94/107 [02:11<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 89%|████████▉ | 95/107 [02:16<00:25, 2.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 90%|████████▉ | 96/107 [02:17<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 91%|█████████ | 97/107 [02:19<00:19, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 92%|█████████▏| 98/107 [02:20<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 93%|█████████▎| 99/107 [02:20<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 93%|█████████▎| 100/107 [02:21<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 94%|█████████▍| 101/107 [02:22<00:06, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 95%|█████████▌| 102/107 [02:23<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 96%|█████████▋| 103/107 [02:24<00:03, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 97%|█████████▋| 104/107 [02:25<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 98%|█████████▊| 105/107 [02:25<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 99%|█████████▉| 106/107 [02:27<00:00, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 67: 100%|██████████| 107/107 [02:32<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 51%|█████▏ | 55/107 [00:58<00:26, 1.94batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 68: 52%|█████▏ | 56/107 [01:19<05:48, 6.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 53%|█████▎ | 57/107 [01:24<05:13, 6.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 54%|█████▍ | 58/107 [01:25<03:51, 4.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 55%|█████▌ | 59/107 [01:28<03:23, 4.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 56%|█████▌ | 60/107 [01:30<02:36, 3.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 57%|█████▋ | 61/107 [01:30<01:58, 2.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 58%|█████▊ | 62/107 [01:33<02:03, 2.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 59%|█████▉ | 63/107 [01:35<01:40, 2.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 60%|█████▉ | 64/107 [01:36<01:29, 2.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 61%|██████ | 65/107 [01:37<01:13, 1.76s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 62%|██████▏ | 66/107 [01:39<01:13, 1.78s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 63%|██████▎ | 67/107 [01:40<01:01, 1.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 64%|██████▎ | 68/107 [01:41<00:52, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 64%|██████▍ | 69/107 [01:42<00:50, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 65%|██████▌ | 70/107 [01:44<00:53, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 66%|██████▋ | 71/107 [01:46<00:53, 1.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 67%|██████▋ | 72/107 [01:47<00:49, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 68%|██████▊ | 73/107 [01:47<00:40, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 69%|██████▉ | 74/107 [01:49<00:38, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 70%|███████ | 75/107 [01:49<00:33, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 71%|███████ | 76/107 [01:50<00:30, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 72%|███████▏ | 77/107 [01:51<00:29, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 73%|███████▎ | 78/107 [01:53<00:35, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 74%|███████▍ | 79/107 [01:54<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 75%|███████▍ | 80/107 [01:55<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 76%|███████▌ | 81/107 [01:55<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 77%|███████▋ | 82/107 [01:56<00:22, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 78%|███████▊ | 83/107 [01:57<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 79%|███████▊ | 84/107 [01:58<00:19, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 79%|███████▉ | 85/107 [01:59<00:18, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 80%|████████ | 86/107 [02:00<00:17, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 81%|████████▏ | 87/107 [02:00<00:16, 1.21batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 82%|████████▏ | 88/107 [02:01<00:15, 1.25batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 83%|████████▎ | 89/107 [02:02<00:14, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 84%|████████▍ | 90/107 [02:03<00:16, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 85%|████████▌ | 91/107 [02:04<00:14, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 86%|████████▌ | 92/107 [02:06<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 87%|████████▋ | 93/107 [02:07<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 88%|████████▊ | 94/107 [02:08<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 89%|████████▉ | 95/107 [02:12<00:25, 2.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 90%|████████▉ | 96/107 [02:14<00:19, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 91%|█████████ | 97/107 [02:16<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 92%|█████████▏| 98/107 [02:16<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 93%|█████████▎| 99/107 [02:17<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 93%|█████████▎| 100/107 [02:18<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 94%|█████████▍| 101/107 [02:19<00:06, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 95%|█████████▌| 102/107 [02:20<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 96%|█████████▋| 103/107 [02:21<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 97%|█████████▋| 104/107 [02:22<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 98%|█████████▊| 105/107 [02:23<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 99%|█████████▉| 106/107 [02:24<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 68: 100%|██████████| 107/107 [02:29<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 51%|█████▏ | 55/107 [00:59<00:25, 2.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 69: 52%|█████▏ | 56/107 [01:22<05:56, 6.99s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 53%|█████▎ | 57/107 [01:25<04:54, 5.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 54%|█████▍ | 58/107 [01:27<03:56, 4.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 55%|█████▌ | 59/107 [01:29<03:05, 3.87s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 56%|█████▌ | 60/107 [01:30<02:29, 3.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 57%|█████▋ | 61/107 [01:31<01:54, 2.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 58%|█████▊ | 62/107 [01:36<02:19, 3.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 59%|█████▉ | 63/107 [01:37<01:51, 2.53s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 60%|█████▉ | 64/107 [01:38<01:33, 2.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 61%|██████ | 65/107 [01:40<01:18, 1.87s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 62%|██████▏ | 66/107 [01:41<01:13, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 63%|██████▎ | 67/107 [01:43<01:08, 1.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 64%|██████▎ | 68/107 [01:44<00:56, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 64%|██████▍ | 69/107 [01:45<00:53, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 65%|██████▌ | 70/107 [01:46<00:54, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 66%|██████▋ | 71/107 [01:48<00:54, 1.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 67%|██████▋ | 72/107 [01:49<00:50, 1.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 68%|██████▊ | 73/107 [01:50<00:40, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 69%|██████▉ | 74/107 [01:51<00:39, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 70%|███████ | 75/107 [01:52<00:33, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 71%|███████ | 76/107 [01:53<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 72%|███████▏ | 77/107 [01:54<00:29, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 73%|███████▎ | 78/107 [01:55<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 74%|███████▍ | 79/107 [01:56<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 75%|███████▍ | 80/107 [01:57<00:28, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 76%|███████▌ | 81/107 [01:58<00:24, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 77%|███████▋ | 82/107 [01:59<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 78%|███████▊ | 83/107 [02:00<00:22, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 79%|███████▊ | 84/107 [02:01<00:20, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 79%|███████▉ | 85/107 [02:01<00:19, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 80%|████████ | 86/107 [02:02<00:18, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 81%|████████▏ | 87/107 [02:03<00:17, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 82%|████████▏ | 88/107 [02:04<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 83%|████████▎ | 89/107 [02:05<00:15, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 84%|████████▍ | 90/107 [02:06<00:17, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 85%|████████▌ | 91/107 [02:07<00:15, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 86%|████████▌ | 92/107 [02:09<00:18, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 87%|████████▋ | 93/107 [02:10<00:15, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 88%|████████▊ | 94/107 [02:11<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 89%|████████▉ | 95/107 [02:15<00:25, 2.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 90%|████████▉ | 96/107 [02:17<00:19, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 91%|█████████ | 97/107 [02:19<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 92%|█████████▏| 98/107 [02:19<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 93%|█████████▎| 99/107 [02:20<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 93%|█████████▎| 100/107 [02:21<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 94%|█████████▍| 101/107 [02:22<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 95%|█████████▌| 102/107 [02:23<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 96%|█████████▋| 103/107 [02:24<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 97%|█████████▋| 104/107 [02:25<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 98%|█████████▊| 105/107 [02:25<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 99%|█████████▉| 106/107 [02:27<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 69: 100%|██████████| 107/107 [02:32<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 51%|█████▏ | 55/107 [01:01<00:27, 1.87batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 70: 52%|█████▏ | 56/107 [01:23<06:01, 7.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 53%|█████▎ | 57/107 [01:26<04:56, 5.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 54%|█████▍ | 58/107 [01:28<03:45, 4.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 55%|█████▌ | 59/107 [01:30<03:03, 3.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 56%|█████▌ | 60/107 [01:31<02:28, 3.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 57%|█████▋ | 61/107 [01:33<01:59, 2.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 58%|█████▊ | 62/107 [01:35<01:50, 2.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 59%|█████▉ | 63/107 [01:36<01:30, 2.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 60%|█████▉ | 64/107 [01:38<01:24, 1.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 61%|██████ | 65/107 [01:39<01:11, 1.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 62%|██████▏ | 66/107 [01:40<01:06, 1.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 63%|██████▎ | 67/107 [01:41<00:59, 1.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 64%|██████▎ | 68/107 [01:42<00:51, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 64%|██████▍ | 69/107 [01:43<00:49, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 65%|██████▌ | 70/107 [01:45<00:45, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 66%|██████▋ | 71/107 [01:46<00:49, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 67%|██████▋ | 72/107 [01:48<00:53, 1.53s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 68%|██████▊ | 73/107 [01:49<00:42, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 69%|██████▉ | 74/107 [01:50<00:40, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 70%|███████ | 75/107 [01:51<00:35, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 71%|███████ | 76/107 [01:52<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 72%|███████▏ | 77/107 [01:53<00:29, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 73%|███████▎ | 78/107 [01:54<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 74%|███████▍ | 79/107 [01:55<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 75%|███████▍ | 80/107 [01:56<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 76%|███████▌ | 81/107 [01:57<00:24, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 77%|███████▋ | 82/107 [01:58<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 78%|███████▊ | 83/107 [01:59<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 79%|███████▊ | 84/107 [01:59<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 79%|███████▉ | 85/107 [02:00<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 80%|████████ | 86/107 [02:01<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 81%|████████▏ | 87/107 [02:02<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 82%|████████▏ | 88/107 [02:02<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 83%|████████▎ | 89/107 [02:03<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 84%|████████▍ | 90/107 [02:05<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 85%|████████▌ | 91/107 [02:05<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 86%|████████▌ | 92/107 [02:07<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 87%|████████▋ | 93/107 [02:08<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 88%|████████▊ | 94/107 [02:09<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 89%|████████▉ | 95/107 [02:14<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 90%|████████▉ | 96/107 [02:15<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 91%|█████████ | 97/107 [02:17<00:19, 1.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 92%|█████████▏| 98/107 [02:18<00:14, 1.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 93%|█████████▎| 99/107 [02:19<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 93%|█████████▎| 100/107 [02:19<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 94%|█████████▍| 101/107 [02:20<00:06, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 95%|█████████▌| 102/107 [02:21<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 96%|█████████▋| 103/107 [02:22<00:03, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 97%|█████████▋| 104/107 [02:23<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 98%|█████████▊| 105/107 [02:24<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 99%|█████████▉| 106/107 [02:25<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 70: 100%|██████████| 107/107 [02:30<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 51%|█████▏ | 55/107 [00:59<00:24, 2.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 71: 52%|█████▏ | 56/107 [01:22<06:08, 7.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 53%|█████▎ | 57/107 [01:26<05:16, 6.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 54%|█████▍ | 58/107 [01:27<03:56, 4.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 55%|█████▌ | 59/107 [01:29<03:05, 3.86s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 56%|█████▌ | 60/107 [01:30<02:29, 3.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 57%|█████▋ | 61/107 [01:31<01:54, 2.49s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 58%|█████▊ | 62/107 [01:34<01:49, 2.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 59%|█████▉ | 63/107 [01:35<01:29, 2.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 60%|█████▉ | 64/107 [01:36<01:23, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 61%|██████ | 65/107 [01:37<01:10, 1.69s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 62%|██████▏ | 66/107 [01:39<01:12, 1.78s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 63%|██████▎ | 67/107 [01:41<01:07, 1.69s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 64%|██████▎ | 68/107 [01:42<01:01, 1.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 64%|██████▍ | 69/107 [01:43<00:54, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 65%|██████▌ | 70/107 [01:44<00:49, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 66%|██████▋ | 71/107 [01:47<00:57, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 67%|██████▋ | 72/107 [01:48<00:51, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 68%|██████▊ | 73/107 [01:48<00:41, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 69%|██████▉ | 74/107 [01:50<00:39, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 70%|███████ | 75/107 [01:50<00:34, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 71%|███████ | 76/107 [01:51<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 72%|███████▏ | 77/107 [01:52<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 73%|███████▎ | 78/107 [01:54<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 74%|███████▍ | 79/107 [01:55<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 75%|███████▍ | 80/107 [01:56<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 76%|███████▌ | 81/107 [01:56<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 77%|███████▋ | 82/107 [01:57<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 78%|███████▊ | 83/107 [01:58<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 79%|███████▊ | 84/107 [01:59<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 79%|███████▉ | 85/107 [02:00<00:18, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 80%|████████ | 86/107 [02:01<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 81%|████████▏ | 87/107 [02:01<00:16, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 82%|████████▏ | 88/107 [02:02<00:15, 1.25batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 83%|████████▎ | 89/107 [02:03<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 84%|████████▍ | 90/107 [02:04<00:16, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 85%|████████▌ | 91/107 [02:05<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 86%|████████▌ | 92/107 [02:07<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 87%|████████▋ | 93/107 [02:08<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 88%|████████▊ | 94/107 [02:09<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 89%|████████▉ | 95/107 [02:13<00:25, 2.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 90%|████████▉ | 96/107 [02:14<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 91%|█████████ | 97/107 [02:17<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 92%|█████████▏| 98/107 [02:18<00:14, 1.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 93%|█████████▎| 99/107 [02:19<00:11, 1.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 93%|█████████▎| 100/107 [02:19<00:08, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 94%|█████████▍| 101/107 [02:20<00:06, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 95%|█████████▌| 102/107 [02:22<00:06, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 96%|█████████▋| 103/107 [02:23<00:04, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 97%|█████████▋| 104/107 [02:23<00:03, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 98%|█████████▊| 105/107 [02:24<00:02, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 99%|█████████▉| 106/107 [02:26<00:01, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 71: 100%|██████████| 107/107 [02:31<00:00, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 51%|█████▏ | 55/107 [01:00<00:22, 2.28batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 72: 52%|█████▏ | 56/107 [01:23<06:08, 7.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 53%|█████▎ | 57/107 [01:27<05:20, 6.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 54%|█████▍ | 58/107 [01:28<03:57, 4.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 55%|█████▌ | 59/107 [01:32<03:28, 4.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 56%|█████▌ | 60/107 [01:33<02:37, 3.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 57%|█████▋ | 61/107 [01:34<02:01, 2.65s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 58%|█████▊ | 62/107 [01:38<02:17, 3.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 59%|█████▉ | 63/107 [01:39<01:48, 2.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 60%|█████▉ | 64/107 [01:40<01:34, 2.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 61%|██████ | 65/107 [01:42<01:19, 1.89s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 62%|██████▏ | 66/107 [01:43<01:14, 1.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 63%|██████▎ | 67/107 [01:44<01:02, 1.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 64%|██████▎ | 68/107 [01:45<00:53, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 64%|██████▍ | 69/107 [01:47<00:57, 1.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 65%|██████▌ | 70/107 [01:48<00:52, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 66%|██████▋ | 71/107 [01:50<00:53, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 67%|██████▋ | 72/107 [01:51<00:48, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 68%|██████▊ | 73/107 [01:52<00:39, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 69%|██████▉ | 74/107 [01:53<00:38, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 70%|███████ | 75/107 [01:54<00:33, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 71%|███████ | 76/107 [01:54<00:30, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 72%|███████▏ | 77/107 [01:55<00:29, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 73%|███████▎ | 78/107 [01:57<00:35, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 74%|███████▍ | 79/107 [01:58<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 75%|███████▍ | 80/107 [01:59<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 76%|███████▌ | 81/107 [02:00<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 77%|███████▋ | 82/107 [02:00<00:22, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 78%|███████▊ | 83/107 [02:01<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 79%|███████▊ | 84/107 [02:02<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 79%|███████▉ | 85/107 [02:03<00:18, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 80%|████████ | 86/107 [02:04<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 81%|████████▏ | 87/107 [02:05<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 82%|████████▏ | 88/107 [02:05<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 83%|████████▎ | 89/107 [02:06<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 84%|████████▍ | 90/107 [02:07<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 85%|████████▌ | 91/107 [02:08<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 86%|████████▌ | 92/107 [02:10<00:17, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 87%|████████▋ | 93/107 [02:11<00:15, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 88%|████████▊ | 94/107 [02:12<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 89%|████████▉ | 95/107 [02:17<00:25, 2.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 90%|████████▉ | 96/107 [02:18<00:19, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 91%|█████████ | 97/107 [02:20<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 92%|█████████▏| 98/107 [02:21<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 93%|█████████▎| 99/107 [02:22<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 93%|█████████▎| 100/107 [02:22<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 94%|█████████▍| 101/107 [02:23<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 95%|█████████▌| 102/107 [02:24<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 96%|█████████▋| 103/107 [02:25<00:03, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 97%|█████████▋| 104/107 [02:26<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 98%|█████████▊| 105/107 [02:27<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 99%|█████████▉| 106/107 [02:28<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 72: 100%|██████████| 107/107 [02:33<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 51%|█████▏ | 55/107 [01:02<00:30, 1.73batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 73: 52%|█████▏ | 56/107 [01:25<06:11, 7.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 53%|█████▎ | 57/107 [01:28<04:59, 6.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 54%|█████▍ | 58/107 [01:29<03:44, 4.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 55%|█████▌ | 59/107 [01:33<03:31, 4.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 56%|█████▌ | 60/107 [01:34<02:43, 3.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 57%|█████▋ | 61/107 [01:35<02:05, 2.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 58%|█████▊ | 62/107 [01:39<02:15, 3.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 59%|█████▉ | 63/107 [01:40<01:47, 2.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 60%|█████▉ | 64/107 [01:42<01:33, 2.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 61%|██████ | 65/107 [01:43<01:17, 1.86s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 62%|██████▏ | 66/107 [01:45<01:14, 1.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 63%|██████▎ | 67/107 [01:46<01:07, 1.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 64%|██████▎ | 68/107 [01:47<00:56, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 64%|██████▍ | 69/107 [01:48<00:53, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 65%|██████▌ | 70/107 [01:49<00:49, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 66%|██████▋ | 71/107 [01:52<00:59, 1.64s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 67%|██████▋ | 72/107 [01:53<00:52, 1.51s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 68%|██████▊ | 73/107 [01:53<00:42, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 69%|██████▉ | 74/107 [01:55<00:40, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 70%|███████ | 75/107 [01:55<00:35, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 71%|███████ | 76/107 [01:56<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 72%|███████▏ | 77/107 [01:57<00:29, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 73%|███████▎ | 78/107 [01:59<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 74%|███████▍ | 79/107 [02:00<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 75%|███████▍ | 80/107 [02:01<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 76%|███████▌ | 81/107 [02:01<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 77%|███████▋ | 82/107 [02:02<00:22, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 78%|███████▊ | 83/107 [02:03<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 79%|███████▊ | 84/107 [02:04<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 79%|███████▉ | 85/107 [02:05<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 80%|████████ | 86/107 [02:06<00:18, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 81%|████████▏ | 87/107 [02:07<00:17, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 82%|████████▏ | 88/107 [02:07<00:15, 1.21batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 83%|████████▎ | 89/107 [02:08<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 84%|████████▍ | 90/107 [02:09<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 85%|████████▌ | 91/107 [02:10<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 86%|████████▌ | 92/107 [02:12<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 87%|████████▋ | 93/107 [02:13<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 88%|████████▊ | 94/107 [02:14<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 89%|████████▉ | 95/107 [02:19<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 90%|████████▉ | 96/107 [02:20<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 91%|█████████ | 97/107 [02:22<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 92%|█████████▏| 98/107 [02:23<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 93%|█████████▎| 99/107 [02:24<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 93%|█████████▎| 100/107 [02:24<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 94%|█████████▍| 101/107 [02:25<00:06, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 95%|█████████▌| 102/107 [02:26<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 96%|█████████▋| 103/107 [02:27<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 97%|█████████▋| 104/107 [02:28<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 98%|█████████▊| 105/107 [02:29<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 99%|█████████▉| 106/107 [02:30<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 73: 100%|██████████| 107/107 [02:35<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 51%|█████▏ | 55/107 [01:01<00:27, 1.88batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 74: 52%|█████▏ | 56/107 [01:24<06:16, 7.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 53%|█████▎ | 57/107 [01:27<05:04, 6.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 54%|█████▍ | 58/107 [01:28<03:48, 4.67s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 55%|█████▌ | 59/107 [01:32<03:35, 4.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 56%|█████▌ | 60/107 [01:34<02:45, 3.53s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 57%|█████▋ | 61/107 [01:35<02:07, 2.77s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 58%|█████▊ | 62/107 [01:38<02:14, 3.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 59%|█████▉ | 63/107 [01:39<01:46, 2.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 60%|█████▉ | 64/107 [01:41<01:32, 2.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 61%|██████ | 65/107 [01:42<01:18, 1.86s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 62%|██████▏ | 66/107 [01:44<01:16, 1.86s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 63%|██████▎ | 67/107 [01:45<01:07, 1.68s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 64%|██████▎ | 68/107 [01:46<00:56, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 64%|██████▍ | 69/107 [01:47<00:53, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 65%|██████▌ | 70/107 [01:48<00:47, 1.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 66%|██████▋ | 71/107 [01:51<00:56, 1.56s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 67%|██████▋ | 72/107 [01:52<00:50, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 68%|██████▊ | 73/107 [01:52<00:41, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 69%|██████▉ | 74/107 [01:54<00:39, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 70%|███████ | 75/107 [01:54<00:34, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 71%|███████ | 76/107 [01:55<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 72%|███████▏ | 77/107 [01:56<00:29, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 73%|███████▎ | 78/107 [01:58<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 74%|███████▍ | 79/107 [01:59<00:30, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 75%|███████▍ | 80/107 [02:00<00:28, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 76%|███████▌ | 81/107 [02:00<00:24, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 77%|███████▋ | 82/107 [02:01<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 78%|███████▊ | 83/107 [02:02<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 79%|███████▊ | 84/107 [02:03<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 79%|███████▉ | 85/107 [02:04<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 80%|████████ | 86/107 [02:04<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 81%|████████▏ | 87/107 [02:05<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 82%|████████▏ | 88/107 [02:06<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 83%|████████▎ | 89/107 [02:07<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 84%|████████▍ | 90/107 [02:08<00:16, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 85%|████████▌ | 91/107 [02:09<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 86%|████████▌ | 92/107 [02:11<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 87%|████████▋ | 93/107 [02:12<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 88%|████████▊ | 94/107 [02:13<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 89%|████████▉ | 95/107 [02:17<00:25, 2.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 90%|████████▉ | 96/107 [02:18<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 91%|█████████ | 97/107 [02:21<00:19, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 92%|█████████▏| 98/107 [02:21<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 93%|█████████▎| 99/107 [02:22<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 93%|█████████▎| 100/107 [02:23<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 94%|█████████▍| 101/107 [02:24<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 95%|█████████▌| 102/107 [02:25<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 96%|█████████▋| 103/107 [02:26<00:04, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 97%|█████████▋| 104/107 [02:27<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 98%|█████████▊| 105/107 [02:27<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 99%|█████████▉| 106/107 [02:29<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 74: 100%|██████████| 107/107 [02:34<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 51%|█████▏ | 55/107 [01:01<00:23, 2.24batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 75: 52%|█████▏ | 56/107 [01:24<06:13, 7.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 53%|█████▎ | 57/107 [01:28<05:10, 6.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 54%|█████▍ | 58/107 [01:29<03:51, 4.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 55%|█████▌ | 59/107 [01:31<03:12, 4.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 56%|█████▌ | 60/107 [01:33<02:35, 3.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 57%|█████▋ | 61/107 [01:34<02:02, 2.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 58%|█████▊ | 62/107 [01:36<01:50, 2.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 59%|█████▉ | 63/107 [01:37<01:30, 2.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 60%|█████▉ | 64/107 [01:39<01:20, 1.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 61%|██████ | 65/107 [01:40<01:09, 1.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 62%|██████▏ | 66/107 [01:41<01:04, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 63%|██████▎ | 67/107 [01:42<00:57, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 64%|██████▎ | 68/107 [01:43<00:50, 1.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 64%|██████▍ | 69/107 [01:45<00:48, 1.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 65%|██████▌ | 70/107 [01:46<00:45, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 66%|██████▋ | 71/107 [01:47<00:48, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 67%|██████▋ | 72/107 [01:49<00:46, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 68%|██████▊ | 73/107 [01:49<00:38, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 69%|██████▉ | 74/107 [01:51<00:46, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 70%|███████ | 75/107 [01:52<00:39, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 71%|███████ | 76/107 [01:53<00:34, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 72%|███████▏ | 77/107 [01:54<00:31, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 73%|███████▎ | 78/107 [01:56<00:36, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 74%|███████▍ | 79/107 [01:56<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 75%|███████▍ | 80/107 [01:57<00:30, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 76%|███████▌ | 81/107 [01:58<00:25, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 77%|███████▋ | 82/107 [01:59<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 78%|███████▊ | 83/107 [02:00<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 79%|███████▊ | 84/107 [02:01<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 79%|███████▉ | 85/107 [02:01<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 80%|████████ | 86/107 [02:02<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 81%|████████▏ | 87/107 [02:03<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 82%|████████▏ | 88/107 [02:04<00:15, 1.21batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 83%|████████▎ | 89/107 [02:05<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 84%|████████▍ | 90/107 [02:06<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 85%|████████▌ | 91/107 [02:07<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 86%|████████▌ | 92/107 [02:09<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 87%|████████▋ | 93/107 [02:10<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 88%|████████▊ | 94/107 [02:11<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 89%|████████▉ | 95/107 [02:15<00:25, 2.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 90%|████████▉ | 96/107 [02:16<00:19, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 91%|█████████ | 97/107 [02:19<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 92%|█████████▏| 98/107 [02:19<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 93%|█████████▎| 99/107 [02:20<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 93%|█████████▎| 100/107 [02:21<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 94%|█████████▍| 101/107 [02:22<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 95%|█████████▌| 102/107 [02:23<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 96%|█████████▋| 103/107 [02:24<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 97%|█████████▋| 104/107 [02:24<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 98%|█████████▊| 105/107 [02:25<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 99%|█████████▉| 106/107 [02:26<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 75: 100%|██████████| 107/107 [02:32<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 51%|█████▏ | 55/107 [01:01<00:30, 1.73batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 76: 52%|█████▏ | 56/107 [01:25<06:22, 7.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 53%|█████▎ | 57/107 [01:28<05:09, 6.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 54%|█████▍ | 58/107 [01:29<03:51, 4.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 55%|█████▌ | 59/107 [01:33<03:32, 4.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 56%|█████▌ | 60/107 [01:34<02:41, 3.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 57%|█████▋ | 61/107 [01:35<02:02, 2.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 58%|█████▊ | 62/107 [01:40<02:31, 3.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 59%|█████▉ | 63/107 [01:41<01:59, 2.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 60%|█████▉ | 64/107 [01:43<01:38, 2.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 61%|██████ | 65/107 [01:44<01:19, 1.89s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 62%|██████▏ | 66/107 [01:45<01:16, 1.86s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 63%|██████▎ | 67/107 [01:47<01:08, 1.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 64%|██████▎ | 68/107 [01:48<00:56, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 64%|██████▍ | 69/107 [01:49<00:57, 1.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 65%|██████▌ | 70/107 [01:50<00:52, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 66%|██████▋ | 71/107 [01:52<00:52, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 67%|██████▋ | 72/107 [01:53<00:48, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 68%|██████▊ | 73/107 [01:54<00:39, 1.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 69%|██████▉ | 74/107 [01:55<00:38, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 70%|███████ | 75/107 [01:56<00:33, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 71%|███████ | 76/107 [01:57<00:30, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 72%|███████▏ | 77/107 [01:58<00:29, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 73%|███████▎ | 78/107 [01:59<00:34, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 74%|███████▍ | 79/107 [02:57<08:26, 18.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 75%|███████▍ | 80/107 [02:58<05:50, 12.99s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 76%|███████▌ | 81/107 [02:58<04:01, 9.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 77%|███████▋ | 82/107 [02:59<02:48, 6.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 78%|███████▊ | 83/107 [03:00<02:00, 5.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 79%|███████▊ | 84/107 [03:01<01:26, 3.74s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 79%|███████▉ | 85/107 [03:02<01:02, 2.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 80%|████████ | 86/107 [03:03<00:47, 2.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 81%|████████▏ | 87/107 [03:03<00:36, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 82%|████████▏ | 88/107 [03:04<00:28, 1.49s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 83%|████████▎ | 89/107 [03:05<00:23, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 84%|████████▍ | 90/107 [03:06<00:21, 1.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 85%|████████▌ | 91/107 [03:07<00:18, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 86%|████████▌ | 92/107 [03:09<00:20, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 87%|████████▋ | 93/107 [03:10<00:16, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 88%|████████▊ | 94/107 [03:11<00:16, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 89%|████████▉ | 95/107 [03:16<00:26, 2.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 90%|████████▉ | 96/107 [03:17<00:20, 1.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 91%|█████████ | 97/107 [03:19<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 92%|█████████▏| 98/107 [03:19<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 93%|█████████▎| 99/107 [03:20<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 93%|█████████▎| 100/107 [03:21<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 94%|█████████▍| 101/107 [03:22<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 95%|█████████▌| 102/107 [03:23<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 96%|█████████▋| 103/107 [03:24<00:04, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 97%|█████████▋| 104/107 [03:25<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 98%|█████████▊| 105/107 [03:25<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 99%|█████████▉| 106/107 [03:27<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 76: 100%|██████████| 107/107 [03:32<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 51%|█████▏ | 55/107 [01:03<00:24, 2.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 77: 52%|█████▏ | 56/107 [01:28<06:32, 7.69s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 53%|█████▎ | 57/107 [01:31<05:14, 6.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 54%|█████▍ | 58/107 [01:32<03:57, 4.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 55%|█████▌ | 59/107 [01:34<03:10, 3.97s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 56%|█████▌ | 60/107 [01:35<02:32, 3.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 57%|█████▋ | 61/107 [01:37<02:00, 2.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 58%|█████▊ | 62/107 [01:39<01:49, 2.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 59%|█████▉ | 63/107 [01:40<01:33, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 60%|█████▉ | 64/107 [01:41<01:21, 1.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 61%|██████ | 65/107 [01:42<01:06, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 62%|██████▏ | 66/107 [01:44<01:07, 1.64s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 63%|██████▎ | 67/107 [01:45<00:59, 1.49s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 64%|██████▎ | 68/107 [01:46<00:51, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 64%|██████▍ | 69/107 [01:47<00:49, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 65%|██████▌ | 70/107 [01:48<00:45, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 66%|██████▋ | 71/107 [01:51<00:54, 1.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 67%|██████▋ | 72/107 [01:52<00:49, 1.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 68%|██████▊ | 73/107 [01:52<00:40, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 69%|██████▉ | 74/107 [01:54<00:39, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 70%|███████ | 75/107 [01:54<00:34, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 71%|███████ | 76/107 [01:55<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 72%|███████▏ | 77/107 [01:56<00:29, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 73%|███████▎ | 78/107 [01:58<00:35, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 74%|███████▍ | 79/107 [01:59<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 75%|███████▍ | 80/107 [02:00<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 76%|███████▌ | 81/107 [02:00<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 77%|███████▋ | 82/107 [02:01<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 78%|███████▊ | 83/107 [02:02<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 79%|███████▊ | 84/107 [02:03<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 79%|███████▉ | 85/107 [02:04<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 80%|████████ | 86/107 [02:05<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 81%|████████▏ | 87/107 [02:06<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 82%|████████▏ | 88/107 [02:06<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 83%|████████▎ | 89/107 [02:07<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 84%|████████▍ | 90/107 [02:08<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 85%|████████▌ | 91/107 [02:09<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 86%|████████▌ | 92/107 [02:11<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 87%|████████▋ | 93/107 [02:12<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 88%|████████▊ | 94/107 [02:13<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 89%|████████▉ | 95/107 [02:18<00:25, 2.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 90%|████████▉ | 96/107 [02:19<00:20, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 91%|█████████ | 97/107 [02:21<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 92%|█████████▏| 98/107 [02:22<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 93%|█████████▎| 99/107 [02:23<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 93%|█████████▎| 100/107 [02:23<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 94%|█████████▍| 101/107 [02:24<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 95%|█████████▌| 102/107 [02:25<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 96%|█████████▋| 103/107 [02:26<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 97%|█████████▋| 104/107 [02:27<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 98%|█████████▊| 105/107 [02:28<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 99%|█████████▉| 106/107 [02:29<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 77: 100%|██████████| 107/107 [02:34<00:00, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 51%|█████▏ | 55/107 [01:03<00:20, 2.51batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 78: 52%|█████▏ | 56/107 [01:27<06:28, 7.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 53%|█████▎ | 57/107 [01:31<05:20, 6.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 54%|█████▍ | 58/107 [01:32<03:58, 4.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 55%|█████▌ | 59/107 [01:35<03:19, 4.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 56%|█████▌ | 60/107 [01:36<02:37, 3.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 57%|█████▋ | 61/107 [01:37<02:03, 2.68s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 58%|█████▊ | 62/107 [01:39<01:51, 2.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 59%|█████▉ | 63/107 [01:40<01:30, 2.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 60%|█████▉ | 64/107 [01:42<01:24, 1.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 61%|██████ | 65/107 [01:43<01:09, 1.65s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 62%|██████▏ | 66/107 [01:45<01:08, 1.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 63%|██████▎ | 67/107 [01:46<01:00, 1.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 64%|██████▎ | 68/107 [01:47<00:51, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 64%|██████▍ | 69/107 [01:48<00:49, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 65%|██████▌ | 70/107 [01:49<00:46, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 66%|██████▋ | 71/107 [01:51<00:49, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 67%|██████▋ | 72/107 [01:52<00:47, 1.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 68%|██████▊ | 73/107 [01:53<00:39, 1.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 69%|██████▉ | 74/107 [01:54<00:39, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 70%|███████ | 75/107 [01:55<00:39, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 71%|███████ | 76/107 [01:56<00:34, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 72%|███████▏ | 77/107 [01:57<00:32, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 73%|███████▎ | 78/107 [01:59<00:37, 1.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 74%|███████▍ | 79/107 [02:00<00:31, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 75%|███████▍ | 80/107 [02:01<00:30, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 76%|███████▌ | 81/107 [02:01<00:25, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 77%|███████▋ | 82/107 [02:02<00:23, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 78%|███████▊ | 83/107 [02:03<00:23, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 79%|███████▊ | 84/107 [02:04<00:20, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 79%|███████▉ | 85/107 [02:05<00:19, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 80%|████████ | 86/107 [02:06<00:18, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 81%|████████▏ | 87/107 [02:07<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 82%|████████▏ | 88/107 [02:07<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 83%|████████▎ | 89/107 [02:08<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 84%|████████▍ | 90/107 [02:10<00:16, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 85%|████████▌ | 91/107 [02:10<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 86%|████████▌ | 92/107 [02:12<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 87%|████████▋ | 93/107 [02:13<00:15, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 88%|████████▊ | 94/107 [02:14<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 89%|████████▉ | 95/107 [02:19<00:25, 2.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 90%|████████▉ | 96/107 [02:20<00:20, 1.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 91%|█████████ | 97/107 [02:22<00:19, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 92%|█████████▏| 98/107 [02:23<00:14, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 93%|█████████▎| 99/107 [02:24<00:11, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 93%|█████████▎| 100/107 [02:25<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 94%|█████████▍| 101/107 [02:25<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 95%|█████████▌| 102/107 [02:27<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 96%|█████████▋| 103/107 [02:27<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 97%|█████████▋| 104/107 [02:28<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 98%|█████████▊| 105/107 [02:29<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 99%|█████████▉| 106/107 [02:30<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 78: 100%|██████████| 107/107 [02:35<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 51%|█████▏ | 55/107 [01:03<00:28, 1.84batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 79: 52%|█████▏ | 56/107 [01:27<06:34, 7.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 53%|█████▎ | 57/107 [01:31<05:21, 6.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 54%|█████▍ | 58/107 [01:32<04:00, 4.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 55%|█████▌ | 59/107 [01:34<03:19, 4.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 56%|█████▌ | 60/107 [01:36<02:40, 3.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 57%|█████▋ | 61/107 [01:37<02:05, 2.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 58%|█████▊ | 62/107 [01:39<01:51, 2.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 59%|█████▉ | 63/107 [01:40<01:31, 2.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 60%|█████▉ | 64/107 [01:42<01:21, 1.89s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 61%|██████ | 65/107 [01:43<01:11, 1.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 62%|██████▏ | 66/107 [01:45<01:10, 1.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 63%|██████▎ | 67/107 [01:46<01:00, 1.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 64%|██████▎ | 68/107 [01:47<00:51, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 64%|██████▍ | 69/107 [01:48<00:48, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 65%|██████▌ | 70/107 [01:49<00:45, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 66%|██████▋ | 71/107 [01:51<00:49, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 67%|██████▋ | 72/107 [01:52<00:51, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 68%|██████▊ | 73/107 [01:53<00:41, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 69%|██████▉ | 74/107 [01:54<00:40, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 70%|███████ | 75/107 [01:55<00:34, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 71%|███████ | 76/107 [01:56<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 72%|███████▏ | 77/107 [01:57<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 73%|███████▎ | 78/107 [01:59<00:36, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 74%|███████▍ | 79/107 [01:59<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 75%|███████▍ | 80/107 [02:00<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 76%|███████▌ | 81/107 [02:01<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 77%|███████▋ | 82/107 [02:02<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 78%|███████▊ | 83/107 [02:03<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 79%|███████▊ | 84/107 [02:04<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 79%|███████▉ | 85/107 [02:04<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 80%|████████ | 86/107 [02:05<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 81%|████████▏ | 87/107 [02:06<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 82%|████████▏ | 88/107 [02:07<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 83%|████████▎ | 89/107 [02:08<00:15, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 84%|████████▍ | 90/107 [02:09<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 85%|████████▌ | 91/107 [02:10<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 86%|████████▌ | 92/107 [02:12<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 87%|████████▋ | 93/107 [02:12<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 88%|████████▊ | 94/107 [02:14<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 89%|████████▉ | 95/107 [02:18<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 90%|████████▉ | 96/107 [02:19<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 91%|█████████ | 97/107 [02:21<00:18, 1.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 92%|█████████▏| 98/107 [02:22<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 93%|█████████▎| 99/107 [02:23<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 93%|█████████▎| 100/107 [02:24<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 94%|█████████▍| 101/107 [02:25<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 95%|█████████▌| 102/107 [02:26<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 96%|█████████▋| 103/107 [02:27<00:03, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 97%|█████████▋| 104/107 [02:27<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 98%|█████████▊| 105/107 [02:28<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 99%|█████████▉| 106/107 [02:29<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 79: 100%|██████████| 107/107 [02:35<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 51%|█████▏ | 55/107 [01:02<00:30, 1.70batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 80: 52%|█████▏ | 56/107 [01:27<06:47, 7.99s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 53%|█████▎ | 57/107 [01:31<05:31, 6.63s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 54%|█████▍ | 58/107 [01:32<04:06, 5.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 55%|█████▌ | 59/107 [01:35<03:33, 4.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 56%|█████▌ | 60/107 [01:36<02:46, 3.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 57%|█████▋ | 61/107 [01:37<02:06, 2.74s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 58%|█████▊ | 62/107 [01:42<02:29, 3.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 59%|█████▉ | 63/107 [01:43<01:55, 2.63s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 60%|█████▉ | 64/107 [01:45<01:38, 2.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 61%|██████ | 65/107 [01:46<01:20, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 62%|██████▏ | 66/107 [01:47<01:17, 1.89s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 63%|██████▎ | 67/107 [01:49<01:06, 1.65s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 64%|██████▎ | 68/107 [01:49<00:56, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 64%|██████▍ | 69/107 [01:51<00:52, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 65%|██████▌ | 70/107 [01:52<00:54, 1.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 66%|██████▋ | 71/107 [01:54<00:54, 1.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 67%|██████▋ | 72/107 [01:55<00:49, 1.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 68%|██████▊ | 73/107 [01:56<00:40, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 69%|██████▉ | 74/107 [01:57<00:38, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 70%|███████ | 75/107 [01:58<00:33, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 71%|███████ | 76/107 [01:59<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 72%|███████▏ | 77/107 [02:00<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 73%|███████▎ | 78/107 [02:01<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 74%|███████▍ | 79/107 [02:02<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 75%|███████▍ | 80/107 [02:03<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 76%|███████▌ | 81/107 [02:04<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 77%|███████▋ | 82/107 [02:05<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 78%|███████▊ | 83/107 [02:06<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 79%|███████▊ | 84/107 [02:06<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 79%|███████▉ | 85/107 [02:07<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 80%|████████ | 86/107 [02:08<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 81%|████████▏ | 87/107 [02:09<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 82%|████████▏ | 88/107 [02:10<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 83%|████████▎ | 89/107 [02:11<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 84%|████████▍ | 90/107 [02:12<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 85%|████████▌ | 91/107 [02:13<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 86%|████████▌ | 92/107 [02:14<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 87%|████████▋ | 93/107 [02:15<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 88%|████████▊ | 94/107 [02:17<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 89%|████████▉ | 95/107 [02:21<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 90%|████████▉ | 96/107 [02:22<00:19, 1.79s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 91%|█████████ | 97/107 [02:24<00:19, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 92%|█████████▏| 98/107 [02:25<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 93%|█████████▎| 99/107 [02:26<00:10, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 93%|█████████▎| 100/107 [02:27<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 94%|█████████▍| 101/107 [02:27<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 95%|█████████▌| 102/107 [02:29<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 96%|█████████▋| 103/107 [02:29<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 97%|█████████▋| 104/107 [02:30<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 98%|█████████▊| 105/107 [02:31<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 99%|█████████▉| 106/107 [02:32<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 80: 100%|██████████| 107/107 [02:38<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 51%|█████▏ | 55/107 [01:05<00:25, 2.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 81: 52%|█████▏ | 56/107 [01:30<06:48, 8.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 53%|█████▎ | 57/107 [01:33<05:22, 6.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 54%|█████▍ | 58/107 [01:35<04:06, 5.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 55%|█████▌ | 59/107 [01:37<03:18, 4.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 56%|█████▌ | 60/107 [01:38<02:38, 3.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 57%|█████▋ | 61/107 [01:40<02:05, 2.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 58%|█████▊ | 62/107 [01:42<01:54, 2.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 59%|█████▉ | 63/107 [01:43<01:35, 2.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 60%|█████▉ | 64/107 [01:44<01:20, 1.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 61%|██████ | 65/107 [01:46<01:12, 1.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 62%|██████▏ | 66/107 [01:47<01:07, 1.65s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 63%|██████▎ | 67/107 [01:48<00:58, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 64%|██████▎ | 68/107 [01:49<00:51, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 64%|██████▍ | 69/107 [01:50<00:49, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 65%|██████▌ | 70/107 [01:51<00:44, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 66%|██████▋ | 71/107 [01:53<00:47, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 67%|██████▋ | 72/107 [01:54<00:46, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 68%|██████▊ | 73/107 [01:55<00:38, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 69%|██████▉ | 74/107 [01:57<00:44, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 70%|███████ | 75/107 [01:58<00:38, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 71%|███████ | 76/107 [01:59<00:33, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 72%|███████▏ | 77/107 [02:00<00:32, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 73%|███████▎ | 78/107 [02:01<00:37, 1.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 74%|███████▍ | 79/107 [02:02<00:31, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 75%|███████▍ | 80/107 [02:03<00:30, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 76%|███████▌ | 81/107 [02:04<00:25, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 77%|███████▋ | 82/107 [02:05<00:23, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 78%|███████▊ | 83/107 [02:06<00:23, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 79%|███████▊ | 84/107 [02:07<00:21, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 79%|███████▉ | 85/107 [02:07<00:19, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 80%|████████ | 86/107 [02:08<00:18, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 81%|████████▏ | 87/107 [02:09<00:17, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 82%|████████▏ | 88/107 [02:10<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 83%|████████▎ | 89/107 [02:11<00:15, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 84%|████████▍ | 90/107 [02:12<00:17, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 85%|████████▌ | 91/107 [02:13<00:15, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 86%|████████▌ | 92/107 [02:15<00:18, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 87%|████████▋ | 93/107 [02:16<00:15, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 88%|████████▊ | 94/107 [02:17<00:15, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 89%|████████▉ | 95/107 [02:22<00:26, 2.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 90%|████████▉ | 96/107 [02:23<00:20, 1.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 91%|█████████ | 97/107 [02:25<00:19, 1.94s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 92%|█████████▏| 98/107 [02:26<00:14, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 93%|█████████▎| 99/107 [02:26<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 93%|█████████▎| 100/107 [02:27<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 94%|█████████▍| 101/107 [02:28<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 95%|█████████▌| 102/107 [02:29<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 96%|█████████▋| 103/107 [02:30<00:03, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 97%|█████████▋| 104/107 [02:31<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 98%|█████████▊| 105/107 [02:32<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 99%|█████████▉| 106/107 [02:33<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 81: 100%|██████████| 107/107 [02:38<00:00, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 51%|█████▏ | 55/107 [01:04<00:19, 2.68batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 82: 52%|█████▏ | 56/107 [01:29<06:37, 7.79s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 53%|█████▎ | 57/107 [01:34<05:42, 6.86s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 54%|█████▍ | 58/107 [01:35<04:14, 5.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 55%|█████▌ | 59/107 [01:38<03:37, 4.53s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 56%|█████▌ | 60/107 [01:40<02:45, 3.53s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 57%|█████▋ | 61/107 [01:41<02:08, 2.79s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 58%|█████▊ | 62/107 [01:44<02:14, 2.98s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 59%|█████▉ | 63/107 [01:45<01:45, 2.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 60%|█████▉ | 64/107 [01:47<01:31, 2.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 61%|██████ | 65/107 [01:48<01:18, 1.87s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 62%|██████▏ | 66/107 [01:50<01:18, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 63%|██████▎ | 67/107 [01:51<01:05, 1.64s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 64%|██████▎ | 68/107 [01:52<00:55, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 64%|██████▍ | 69/107 [01:53<00:51, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 65%|██████▌ | 70/107 [01:55<00:53, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 66%|██████▋ | 71/107 [01:56<00:53, 1.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 67%|██████▋ | 72/107 [01:57<00:49, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 68%|██████▊ | 73/107 [01:58<00:39, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 69%|██████▉ | 74/107 [01:59<00:38, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 70%|███████ | 75/107 [02:00<00:34, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 71%|███████ | 76/107 [02:01<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 72%|███████▏ | 77/107 [02:02<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 73%|███████▎ | 78/107 [02:04<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 74%|███████▍ | 79/107 [02:04<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 75%|███████▍ | 80/107 [02:05<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 76%|███████▌ | 81/107 [02:06<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 77%|███████▋ | 82/107 [02:07<00:23, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 78%|███████▊ | 83/107 [02:08<00:23, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 79%|███████▊ | 84/107 [02:09<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 79%|███████▉ | 85/107 [02:10<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 80%|████████ | 86/107 [02:10<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 81%|████████▏ | 87/107 [02:11<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 82%|████████▏ | 88/107 [02:12<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 83%|████████▎ | 89/107 [02:13<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 84%|████████▍ | 90/107 [02:14<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 85%|████████▌ | 91/107 [02:15<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 86%|████████▌ | 92/107 [02:17<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 87%|████████▋ | 93/107 [02:17<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 88%|████████▊ | 94/107 [02:19<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 89%|████████▉ | 95/107 [02:23<00:25, 2.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 90%|████████▉ | 96/107 [02:24<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 91%|█████████ | 97/107 [02:26<00:19, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 92%|█████████▏| 98/107 [02:27<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 93%|█████████▎| 99/107 [02:28<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 93%|█████████▎| 100/107 [02:29<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 94%|█████████▍| 101/107 [02:30<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 95%|█████████▌| 102/107 [02:31<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 96%|█████████▋| 103/107 [02:32<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 97%|█████████▋| 104/107 [02:32<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 98%|█████████▊| 105/107 [02:33<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 99%|█████████▉| 106/107 [02:34<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 82: 100%|██████████| 107/107 [02:40<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 51%|█████▏ | 55/107 [01:03<00:24, 2.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 83: 52%|█████▏ | 56/107 [01:29<06:53, 8.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 53%|█████▎ | 57/107 [01:32<05:32, 6.64s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 54%|█████▍ | 58/107 [01:33<04:08, 5.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 55%|█████▌ | 59/107 [01:36<03:25, 4.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 56%|█████▌ | 60/107 [01:37<02:41, 3.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 57%|█████▋ | 61/107 [01:38<02:07, 2.77s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 58%|█████▊ | 62/107 [01:40<01:54, 2.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 59%|█████▉ | 63/107 [01:42<01:33, 2.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 60%|█████▉ | 64/107 [01:43<01:23, 1.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 61%|██████ | 65/107 [01:45<01:15, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 62%|██████▏ | 66/107 [01:46<01:10, 1.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 63%|██████▎ | 67/107 [01:47<01:00, 1.51s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 64%|██████▎ | 68/107 [01:48<00:51, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 64%|██████▍ | 69/107 [01:49<00:50, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 65%|██████▌ | 70/107 [01:50<00:46, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 66%|██████▋ | 71/107 [01:52<00:50, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 67%|██████▋ | 72/107 [01:53<00:47, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 68%|██████▊ | 73/107 [01:54<00:39, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 69%|██████▉ | 74/107 [01:56<00:44, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 70%|███████ | 75/107 [01:57<00:37, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 71%|███████ | 76/107 [01:57<00:33, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 72%|███████▏ | 77/107 [01:58<00:31, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 73%|███████▎ | 78/107 [02:00<00:36, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 74%|███████▍ | 79/107 [02:01<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 75%|███████▍ | 80/107 [02:02<00:30, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 76%|███████▌ | 81/107 [02:03<00:25, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 77%|███████▋ | 82/107 [02:04<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 78%|███████▊ | 83/107 [02:05<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 79%|███████▊ | 84/107 [02:05<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 79%|███████▉ | 85/107 [02:06<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 80%|████████ | 86/107 [02:07<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 81%|████████▏ | 87/107 [02:08<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 82%|████████▏ | 88/107 [02:09<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 83%|████████▎ | 89/107 [02:10<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 84%|████████▍ | 90/107 [02:11<00:16, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 85%|████████▌ | 91/107 [02:12<00:14, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 86%|████████▌ | 92/107 [02:13<00:17, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 87%|████████▋ | 93/107 [02:14<00:15, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 88%|████████▊ | 94/107 [02:16<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 89%|████████▉ | 95/107 [02:20<00:25, 2.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 90%|████████▉ | 96/107 [02:21<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 91%|█████████ | 97/107 [02:23<00:19, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 92%|█████████▏| 98/107 [02:24<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 93%|█████████▎| 99/107 [02:25<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 93%|█████████▎| 100/107 [02:26<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 94%|█████████▍| 101/107 [02:26<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 95%|█████████▌| 102/107 [02:28<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 96%|█████████▋| 103/107 [02:28<00:04, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 97%|█████████▋| 104/107 [02:29<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 98%|█████████▊| 105/107 [02:30<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 99%|█████████▉| 106/107 [02:31<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 83: 100%|██████████| 107/107 [02:37<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 51%|█████▏ | 55/107 [01:05<00:17, 3.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 84: 52%|█████▏ | 56/107 [01:30<06:38, 7.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 53%|█████▎ | 57/107 [01:35<05:48, 6.97s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 54%|█████▍ | 58/107 [01:36<04:17, 5.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 55%|█████▌ | 59/107 [01:38<03:29, 4.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 56%|█████▌ | 60/107 [01:40<02:42, 3.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 57%|█████▋ | 61/107 [01:41<02:07, 2.77s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 58%|█████▊ | 62/107 [01:45<02:21, 3.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 59%|█████▉ | 63/107 [01:46<01:51, 2.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 60%|█████▉ | 64/107 [01:48<01:36, 2.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 61%|██████ | 65/107 [01:49<01:18, 1.86s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 62%|██████▏ | 66/107 [01:51<01:16, 1.87s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 63%|██████▎ | 67/107 [01:52<01:04, 1.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 64%|██████▎ | 68/107 [01:52<00:54, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 64%|██████▍ | 69/107 [01:54<00:51, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 65%|██████▌ | 70/107 [01:55<00:47, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 66%|██████▋ | 71/107 [01:57<00:55, 1.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 67%|██████▋ | 72/107 [01:58<00:50, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 68%|██████▊ | 73/107 [01:59<00:40, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 69%|██████▉ | 74/107 [02:00<00:39, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 70%|███████ | 75/107 [02:01<00:34, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 71%|███████ | 76/107 [02:02<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 72%|███████▏ | 77/107 [02:03<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 73%|███████▎ | 78/107 [02:04<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 74%|███████▍ | 79/107 [02:05<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 75%|███████▍ | 80/107 [02:06<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 76%|███████▌ | 81/107 [02:07<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 77%|███████▋ | 82/107 [02:08<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 78%|███████▊ | 83/107 [02:09<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 79%|███████▊ | 84/107 [02:09<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 79%|███████▉ | 85/107 [02:10<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 80%|████████ | 86/107 [02:11<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 81%|████████▏ | 87/107 [02:12<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 82%|████████▏ | 88/107 [02:13<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 83%|████████▎ | 89/107 [02:14<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 84%|████████▍ | 90/107 [02:15<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 85%|████████▌ | 91/107 [02:16<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 86%|████████▌ | 92/107 [02:17<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 87%|████████▋ | 93/107 [02:18<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 88%|████████▊ | 94/107 [02:20<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 89%|████████▉ | 95/107 [02:24<00:25, 2.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 90%|████████▉ | 96/107 [02:25<00:19, 1.79s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 91%|█████████ | 97/107 [02:27<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 92%|█████████▏| 98/107 [02:28<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 93%|█████████▎| 99/107 [02:29<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 93%|█████████▎| 100/107 [02:30<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 94%|█████████▍| 101/107 [02:30<00:06, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 95%|█████████▌| 102/107 [02:32<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 96%|█████████▋| 103/107 [02:32<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 97%|█████████▋| 104/107 [02:33<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 98%|█████████▊| 105/107 [02:34<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 99%|█████████▉| 106/107 [02:35<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 84: 100%|██████████| 107/107 [02:41<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 51%|█████▏ | 55/107 [01:04<00:29, 1.76batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 85: 52%|█████▏ | 56/107 [01:30<06:59, 8.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 53%|█████▎ | 57/107 [01:34<05:37, 6.75s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 54%|█████▍ | 58/107 [01:35<04:12, 5.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 55%|█████▌ | 59/107 [01:37<03:19, 4.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 56%|█████▌ | 60/107 [01:38<02:37, 3.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 57%|█████▋ | 61/107 [01:40<02:06, 2.74s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 58%|█████▊ | 62/107 [01:42<01:53, 2.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 59%|█████▉ | 63/107 [01:43<01:33, 2.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 60%|█████▉ | 64/107 [01:45<01:24, 1.98s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 61%|██████ | 65/107 [01:45<01:09, 1.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 62%|██████▏ | 66/107 [01:47<01:07, 1.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 63%|██████▎ | 67/107 [01:48<00:59, 1.49s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 64%|██████▎ | 68/107 [01:49<00:50, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 64%|██████▍ | 69/107 [01:50<00:49, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 65%|██████▌ | 70/107 [01:52<00:45, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 66%|██████▋ | 71/107 [01:53<00:50, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 67%|██████▋ | 72/107 [01:55<00:53, 1.53s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 68%|██████▊ | 73/107 [01:56<00:43, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 69%|██████▉ | 74/107 [01:57<00:41, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 70%|███████ | 75/107 [01:58<00:35, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 71%|███████ | 76/107 [01:59<00:32, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 72%|███████▏ | 77/107 [02:00<00:30, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 73%|███████▎ | 78/107 [02:01<00:36, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 74%|███████▍ | 79/107 [02:02<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 75%|███████▍ | 80/107 [02:03<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 76%|███████▌ | 81/107 [02:04<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 77%|███████▋ | 82/107 [02:05<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 78%|███████▊ | 83/107 [02:06<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 79%|███████▊ | 84/107 [02:06<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 79%|███████▉ | 85/107 [02:07<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 80%|████████ | 86/107 [02:08<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 81%|████████▏ | 87/107 [02:09<00:17, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 82%|████████▏ | 88/107 [02:10<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 83%|████████▎ | 89/107 [02:11<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 84%|████████▍ | 90/107 [02:12<00:16, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 85%|████████▌ | 91/107 [02:13<00:15, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 86%|████████▌ | 92/107 [02:14<00:17, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 87%|████████▋ | 93/107 [02:15<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 88%|████████▊ | 94/107 [02:17<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 89%|████████▉ | 95/107 [02:21<00:25, 2.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 90%|████████▉ | 96/107 [02:22<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 91%|█████████ | 97/107 [02:24<00:19, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 92%|█████████▏| 98/107 [02:25<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 93%|█████████▎| 99/107 [02:26<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 93%|█████████▎| 100/107 [02:27<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 94%|█████████▍| 101/107 [02:27<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 95%|█████████▌| 102/107 [02:29<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 96%|█████████▋| 103/107 [02:29<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 97%|█████████▋| 104/107 [02:30<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 98%|█████████▊| 105/107 [02:31<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 99%|█████████▉| 106/107 [02:32<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 85: 100%|██████████| 107/107 [02:38<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 51%|█████▏ | 55/107 [01:05<00:18, 2.75batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 86: 52%|█████▏ | 56/107 [01:32<06:59, 8.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 53%|█████▎ | 57/107 [01:35<05:34, 6.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 54%|█████▍ | 58/107 [01:36<04:07, 5.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 55%|█████▌ | 59/107 [01:40<03:41, 4.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 56%|█████▌ | 60/107 [01:41<02:51, 3.64s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 57%|█████▋ | 61/107 [01:42<02:10, 2.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 58%|█████▊ | 62/107 [01:47<02:32, 3.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 59%|█████▉ | 63/107 [01:48<01:59, 2.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 60%|█████▉ | 64/107 [01:49<01:41, 2.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 61%|██████ | 65/107 [01:50<01:22, 1.97s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 62%|██████▏ | 66/107 [01:52<01:21, 1.98s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 63%|██████▎ | 67/107 [01:54<01:09, 1.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 64%|██████▎ | 68/107 [01:54<00:57, 1.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 64%|██████▍ | 69/107 [01:56<00:53, 1.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 65%|██████▌ | 70/107 [01:57<00:48, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 66%|██████▋ | 71/107 [01:59<00:57, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 67%|██████▋ | 72/107 [02:00<00:52, 1.49s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 68%|██████▊ | 73/107 [02:01<00:42, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 69%|██████▉ | 74/107 [02:02<00:40, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 70%|███████ | 75/107 [02:03<00:35, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 71%|███████ | 76/107 [02:04<00:31, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 72%|███████▏ | 77/107 [02:05<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 73%|███████▎ | 78/107 [02:07<00:35, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 74%|███████▍ | 79/107 [02:07<00:30, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 75%|███████▍ | 80/107 [02:09<00:30, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 76%|███████▌ | 81/107 [02:09<00:25, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 77%|███████▋ | 82/107 [02:10<00:23, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 78%|███████▊ | 83/107 [02:11<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 79%|███████▊ | 84/107 [02:12<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 79%|███████▉ | 85/107 [02:13<00:19, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 80%|████████ | 86/107 [02:13<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 81%|████████▏ | 87/107 [02:14<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 82%|████████▏ | 88/107 [02:15<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 83%|████████▎ | 89/107 [02:16<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 84%|████████▍ | 90/107 [02:17<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 85%|████████▌ | 91/107 [02:18<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 86%|████████▌ | 92/107 [02:20<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 87%|████████▋ | 93/107 [02:20<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 88%|████████▊ | 94/107 [02:22<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 89%|████████▉ | 95/107 [02:26<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 90%|████████▉ | 96/107 [02:27<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 91%|█████████ | 97/107 [02:29<00:19, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 92%|█████████▏| 98/107 [02:30<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 93%|█████████▎| 99/107 [02:31<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 93%|█████████▎| 100/107 [02:32<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 94%|█████████▍| 101/107 [02:33<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 95%|█████████▌| 102/107 [02:34<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 96%|█████████▋| 103/107 [02:35<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 97%|█████████▋| 104/107 [02:35<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 98%|█████████▊| 105/107 [02:36<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 99%|█████████▉| 106/107 [02:37<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 86: 100%|██████████| 107/107 [02:43<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 51%|█████▏ | 55/107 [01:08<00:24, 2.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 87: 52%|█████▏ | 56/107 [01:35<07:06, 8.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 53%|█████▎ | 57/107 [01:38<05:41, 6.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 54%|█████▍ | 58/107 [01:39<04:18, 5.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 55%|█████▌ | 59/107 [01:41<03:24, 4.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 56%|█████▌ | 60/107 [01:43<02:41, 3.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 57%|█████▋ | 61/107 [01:44<02:08, 2.79s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 58%|█████▊ | 62/107 [01:46<01:58, 2.63s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 59%|█████▉ | 63/107 [01:48<01:38, 2.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 60%|█████▉ | 64/107 [01:49<01:22, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 61%|██████ | 65/107 [01:50<01:06, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 62%|██████▏ | 66/107 [01:51<01:06, 1.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 63%|██████▎ | 67/107 [01:53<00:59, 1.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 64%|██████▎ | 68/107 [01:53<00:51, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 64%|██████▍ | 69/107 [01:55<00:49, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 65%|██████▌ | 70/107 [01:56<00:46, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 66%|██████▋ | 71/107 [01:58<00:49, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 67%|██████▋ | 72/107 [01:59<00:47, 1.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 68%|██████▊ | 73/107 [02:00<00:39, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 69%|██████▉ | 74/107 [02:01<00:39, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 70%|███████ | 75/107 [02:02<00:40, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 71%|███████ | 76/107 [02:03<00:35, 1.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 72%|███████▏ | 77/107 [02:04<00:33, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 73%|███████▎ | 78/107 [02:06<00:38, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 74%|███████▍ | 79/107 [02:07<00:32, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 75%|███████▍ | 80/107 [02:08<00:30, 1.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 76%|███████▌ | 81/107 [02:09<00:25, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 77%|███████▋ | 82/107 [02:09<00:23, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 78%|███████▊ | 83/107 [02:10<00:23, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 79%|███████▊ | 84/107 [02:11<00:20, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 79%|███████▉ | 85/107 [02:12<00:19, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 80%|████████ | 86/107 [02:13<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 81%|████████▏ | 87/107 [02:14<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 82%|████████▏ | 88/107 [02:14<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 83%|████████▎ | 89/107 [02:15<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 84%|████████▍ | 90/107 [02:17<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 85%|████████▌ | 91/107 [02:17<00:14, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 86%|████████▌ | 92/107 [02:19<00:17, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 87%|████████▋ | 93/107 [02:20<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 88%|████████▊ | 94/107 [02:21<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 89%|████████▉ | 95/107 [02:26<00:25, 2.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 90%|████████▉ | 96/107 [02:27<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 91%|█████████ | 97/107 [02:29<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 92%|█████████▏| 98/107 [02:30<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 93%|█████████▎| 99/107 [02:31<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 93%|█████████▎| 100/107 [02:31<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 94%|█████████▍| 101/107 [02:32<00:06, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 95%|█████████▌| 102/107 [02:33<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 96%|█████████▋| 103/107 [02:34<00:03, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 97%|█████████▋| 104/107 [02:35<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 98%|█████████▊| 105/107 [02:36<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 99%|█████████▉| 106/107 [02:37<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 87: 100%|██████████| 107/107 [02:42<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 51%|█████▏ | 55/107 [01:09<00:30, 1.70batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 88: 52%|█████▏ | 56/107 [01:36<07:10, 8.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 53%|█████▎ | 57/107 [01:40<05:47, 6.96s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 54%|█████▍ | 58/107 [01:41<04:19, 5.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 55%|█████▌ | 59/107 [01:43<03:24, 4.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 56%|█████▌ | 60/107 [01:44<02:40, 3.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 57%|█████▋ | 61/107 [01:46<02:06, 2.74s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 58%|█████▊ | 62/107 [01:48<01:54, 2.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 59%|█████▉ | 63/107 [01:49<01:31, 2.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 60%|█████▉ | 64/107 [01:50<01:23, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 61%|██████ | 65/107 [01:51<01:12, 1.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 62%|██████▏ | 66/107 [01:53<01:10, 1.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 63%|██████▎ | 67/107 [01:54<01:00, 1.51s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 64%|██████▎ | 68/107 [01:55<00:51, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 64%|██████▍ | 69/107 [01:56<00:49, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 65%|██████▌ | 70/107 [01:57<00:45, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 66%|██████▋ | 71/107 [01:59<00:47, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 67%|██████▋ | 72/107 [02:00<00:45, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 68%|██████▊ | 73/107 [02:01<00:38, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 69%|██████▉ | 74/107 [02:03<00:43, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 70%|███████ | 75/107 [02:03<00:37, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 71%|███████ | 76/107 [02:04<00:32, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 72%|███████▏ | 77/107 [02:05<00:30, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 73%|███████▎ | 78/107 [02:07<00:36, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 74%|███████▍ | 79/107 [02:08<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 75%|███████▍ | 80/107 [02:09<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 76%|███████▌ | 81/107 [02:09<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 77%|███████▋ | 82/107 [02:10<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 78%|███████▊ | 83/107 [02:11<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 79%|███████▊ | 84/107 [02:12<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 79%|███████▉ | 85/107 [02:13<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 80%|████████ | 86/107 [02:14<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 81%|████████▏ | 87/107 [02:15<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 82%|████████▏ | 88/107 [02:15<00:15, 1.21batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 83%|████████▎ | 89/107 [02:16<00:15, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 84%|████████▍ | 90/107 [02:18<00:17, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 85%|████████▌ | 91/107 [02:18<00:15, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 86%|████████▌ | 92/107 [02:20<00:18, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 87%|████████▋ | 93/107 [02:21<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 88%|████████▊ | 94/107 [02:23<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 89%|████████▉ | 95/107 [02:27<00:25, 2.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 90%|████████▉ | 96/107 [02:28<00:20, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 91%|█████████ | 97/107 [02:30<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 92%|█████████▏| 98/107 [02:31<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 93%|█████████▎| 99/107 [02:32<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 93%|█████████▎| 100/107 [02:33<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 94%|█████████▍| 101/107 [02:33<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 95%|█████████▌| 102/107 [02:35<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 96%|█████████▋| 103/107 [02:35<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 97%|█████████▋| 104/107 [02:36<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 98%|█████████▊| 105/107 [02:37<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 99%|█████████▉| 106/107 [02:38<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 88: 100%|██████████| 107/107 [02:44<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 51%|█████▏ | 55/107 [01:06<00:30, 1.70batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 89: 52%|█████▏ | 56/107 [01:33<07:17, 8.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 53%|█████▎ | 57/107 [01:36<05:45, 6.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 54%|█████▍ | 58/107 [01:38<04:17, 5.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 55%|█████▌ | 59/107 [01:41<03:43, 4.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 56%|█████▌ | 60/107 [01:43<02:57, 3.77s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 57%|█████▋ | 61/107 [01:43<02:13, 2.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 58%|█████▊ | 62/107 [01:48<02:32, 3.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 59%|█████▉ | 63/107 [01:49<02:00, 2.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 60%|█████▉ | 64/107 [01:51<01:41, 2.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 61%|██████ | 65/107 [01:52<01:23, 2.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 62%|██████▏ | 66/107 [01:54<01:21, 1.99s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 63%|██████▎ | 67/107 [01:55<01:08, 1.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 64%|██████▎ | 68/107 [01:56<00:57, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 64%|██████▍ | 69/107 [01:57<00:54, 1.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 65%|██████▌ | 70/107 [01:58<00:48, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 66%|██████▋ | 71/107 [02:01<00:59, 1.64s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 67%|██████▋ | 72/107 [02:02<00:52, 1.51s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 68%|██████▊ | 73/107 [02:02<00:42, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 69%|██████▉ | 74/107 [02:04<00:40, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 70%|███████ | 75/107 [02:04<00:35, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 71%|███████ | 76/107 [02:05<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 72%|███████▏ | 77/107 [02:06<00:30, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 73%|███████▎ | 78/107 [02:08<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 74%|███████▍ | 79/107 [02:09<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 75%|███████▍ | 80/107 [02:10<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 76%|███████▌ | 81/107 [02:10<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 77%|███████▋ | 82/107 [02:11<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 78%|███████▊ | 83/107 [02:12<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 79%|███████▊ | 84/107 [02:13<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 79%|███████▉ | 85/107 [02:14<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 80%|████████ | 86/107 [02:15<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 81%|████████▏ | 87/107 [02:15<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 82%|████████▏ | 88/107 [02:16<00:15, 1.21batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 83%|████████▎ | 89/107 [02:17<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 84%|████████▍ | 90/107 [02:18<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 85%|████████▌ | 91/107 [02:19<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 86%|████████▌ | 92/107 [02:21<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 87%|████████▋ | 93/107 [02:22<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 88%|████████▊ | 94/107 [02:23<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 89%|████████▉ | 95/107 [02:28<00:25, 2.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 90%|████████▉ | 96/107 [02:29<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 91%|█████████ | 97/107 [02:31<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 92%|█████████▏| 98/107 [02:32<00:14, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 93%|█████████▎| 99/107 [02:33<00:11, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 93%|█████████▎| 100/107 [02:33<00:08, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 94%|█████████▍| 101/107 [02:34<00:06, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 95%|█████████▌| 102/107 [02:35<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 96%|█████████▋| 103/107 [02:36<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 97%|█████████▋| 104/107 [02:37<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 98%|█████████▊| 105/107 [02:38<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 99%|█████████▉| 106/107 [02:39<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 89: 100%|██████████| 107/107 [02:44<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 51%|█████▏ | 55/107 [01:06<00:28, 1.80batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 90: 52%|█████▏ | 56/107 [01:34<07:19, 8.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 53%|█████▎ | 57/107 [01:36<05:40, 6.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 54%|█████▍ | 58/107 [01:38<04:12, 5.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 55%|█████▌ | 59/107 [01:42<03:52, 4.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 56%|█████▌ | 60/107 [01:43<03:00, 3.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 57%|█████▋ | 61/107 [01:44<02:17, 2.99s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 58%|█████▊ | 62/107 [01:48<02:23, 3.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 59%|█████▉ | 63/107 [01:49<01:51, 2.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 60%|█████▉ | 64/107 [01:51<01:35, 2.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 61%|██████ | 65/107 [01:52<01:19, 1.89s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 62%|██████▏ | 66/107 [01:53<01:14, 1.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 63%|██████▎ | 67/107 [01:55<01:07, 1.68s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 64%|██████▎ | 68/107 [01:56<00:56, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 64%|██████▍ | 69/107 [01:57<00:52, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 65%|██████▌ | 70/107 [01:58<00:54, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 66%|██████▋ | 71/107 [02:00<00:54, 1.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 67%|██████▋ | 72/107 [02:01<00:49, 1.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 68%|██████▊ | 73/107 [02:02<00:40, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 69%|██████▉ | 74/107 [02:03<00:39, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 70%|███████ | 75/107 [02:04<00:34, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 71%|███████ | 76/107 [02:05<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 72%|███████▏ | 77/107 [02:06<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 73%|███████▎ | 78/107 [02:08<00:35, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 74%|███████▍ | 79/107 [02:08<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 75%|███████▍ | 80/107 [02:09<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 76%|███████▌ | 81/107 [02:10<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 77%|███████▋ | 82/107 [02:11<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 78%|███████▊ | 83/107 [02:12<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 79%|███████▊ | 84/107 [02:13<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 79%|███████▉ | 85/107 [02:13<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 80%|████████ | 86/107 [02:14<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 81%|████████▏ | 87/107 [02:15<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 82%|████████▏ | 88/107 [02:16<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 83%|████████▎ | 89/107 [02:17<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 84%|████████▍ | 90/107 [02:18<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 85%|████████▌ | 91/107 [02:19<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 86%|████████▌ | 92/107 [02:21<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 87%|████████▋ | 93/107 [02:21<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 88%|████████▊ | 94/107 [02:23<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 89%|████████▉ | 95/107 [02:27<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 90%|████████▉ | 96/107 [02:28<00:20, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 91%|█████████ | 97/107 [02:30<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 92%|█████████▏| 98/107 [02:31<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 93%|█████████▎| 99/107 [02:32<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 93%|█████████▎| 100/107 [02:33<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 94%|█████████▍| 101/107 [02:34<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 95%|█████████▌| 102/107 [02:35<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 96%|█████████▋| 103/107 [02:36<00:03, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 97%|█████████▋| 104/107 [02:36<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 98%|█████████▊| 105/107 [02:37<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 99%|█████████▉| 106/107 [02:38<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 90: 100%|██████████| 107/107 [02:44<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 51%|█████▏ | 55/107 [01:07<00:22, 2.33batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 91: 52%|█████▏ | 56/107 [01:35<07:16, 8.56s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 53%|█████▎ | 57/107 [01:38<05:42, 6.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 54%|█████▍ | 58/107 [01:41<04:34, 5.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 55%|█████▌ | 59/107 [01:44<03:56, 4.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 56%|█████▌ | 60/107 [01:45<02:57, 3.78s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 57%|█████▋ | 61/107 [01:46<02:16, 2.97s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 58%|█████▊ | 62/107 [03:07<19:42, 26.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 59%|█████▉ | 63/107 [03:08<13:45, 18.76s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 60%|█████▉ | 64/107 [03:10<09:45, 13.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 61%|██████ | 65/107 [03:11<06:54, 9.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 62%|██████▏ | 66/107 [03:13<05:06, 7.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 63%|██████▎ | 67/107 [03:14<03:47, 5.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 64%|██████▎ | 68/107 [03:15<02:49, 4.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 64%|██████▍ | 69/107 [03:17<02:13, 3.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 65%|██████▌ | 70/107 [03:18<01:45, 2.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 66%|██████▋ | 71/107 [03:20<01:32, 2.56s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 67%|██████▋ | 72/107 [03:21<01:17, 2.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 68%|██████▊ | 73/107 [03:22<01:00, 1.78s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 69%|██████▉ | 74/107 [03:24<00:53, 1.63s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 70%|███████ | 75/107 [03:24<00:44, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 71%|███████ | 76/107 [03:25<00:38, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 72%|███████▏ | 77/107 [03:26<00:35, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 73%|███████▎ | 78/107 [03:28<00:40, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 74%|███████▍ | 79/107 [03:29<00:34, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 75%|███████▍ | 80/107 [03:30<00:32, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 76%|███████▌ | 81/107 [03:31<00:27, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 77%|███████▋ | 82/107 [03:32<00:24, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 78%|███████▊ | 83/107 [03:33<00:24, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 79%|███████▊ | 84/107 [03:34<00:21, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 79%|███████▉ | 85/107 [03:34<00:20, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 80%|████████ | 86/107 [03:35<00:19, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 81%|████████▏ | 87/107 [03:36<00:18, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 82%|████████▏ | 88/107 [03:37<00:16, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 83%|████████▎ | 89/107 [03:38<00:16, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 84%|████████▍ | 90/107 [03:39<00:17, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 85%|████████▌ | 91/107 [03:40<00:15, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 86%|████████▌ | 92/107 [03:42<00:18, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 87%|████████▋ | 93/107 [03:43<00:15, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 88%|████████▊ | 94/107 [03:44<00:15, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 89%|████████▉ | 95/107 [03:49<00:27, 2.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 90%|████████▉ | 96/107 [03:50<00:20, 1.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 91%|█████████ | 97/107 [03:52<00:20, 2.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 92%|█████████▏| 98/107 [03:53<00:14, 1.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 93%|█████████▎| 99/107 [03:54<00:11, 1.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 93%|█████████▎| 100/107 [03:55<00:08, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 94%|█████████▍| 101/107 [03:56<00:06, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 95%|█████████▌| 102/107 [03:57<00:05, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 96%|█████████▋| 103/107 [03:58<00:04, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 97%|█████████▋| 104/107 [03:58<00:02, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 98%|█████████▊| 105/107 [03:59<00:01, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 99%|█████████▉| 106/107 [04:01<00:01, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 91: 100%|██████████| 107/107 [04:06<00:00, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 51%|█████▏ | 55/107 [01:11<00:20, 2.55batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 92: 52%|█████▏ | 56/107 [01:41<07:43, 9.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 53%|█████▎ | 57/107 [01:46<06:36, 7.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 54%|█████▍ | 58/107 [01:48<04:53, 6.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 55%|█████▌ | 59/107 [01:49<03:47, 4.74s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 56%|█████▌ | 60/107 [01:51<02:54, 3.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 57%|█████▋ | 61/107 [01:52<02:18, 3.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 58%|█████▊ | 62/107 [01:56<02:31, 3.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 59%|█████▉ | 63/107 [01:57<01:58, 2.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 60%|█████▉ | 64/107 [01:59<01:40, 2.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 61%|██████ | 65/107 [02:00<01:24, 2.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 62%|██████▏ | 66/107 [02:02<01:22, 2.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 63%|██████▎ | 67/107 [02:04<01:13, 1.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 64%|██████▎ | 68/107 [02:05<01:02, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 64%|██████▍ | 69/107 [02:06<00:56, 1.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 65%|██████▌ | 70/107 [02:07<00:50, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 66%|██████▋ | 71/107 [02:09<00:59, 1.64s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 67%|██████▋ | 72/107 [02:10<00:53, 1.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 68%|██████▊ | 73/107 [02:11<00:43, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 69%|██████▉ | 74/107 [02:12<00:41, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 70%|███████ | 75/107 [02:13<00:36, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 71%|███████ | 76/107 [02:14<00:32, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 72%|███████▏ | 77/107 [02:15<00:30, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 73%|███████▎ | 78/107 [02:17<00:40, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 74%|███████▍ | 79/107 [02:18<00:35, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 75%|███████▍ | 80/107 [02:20<00:35, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 76%|███████▌ | 81/107 [02:20<00:29, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 77%|███████▋ | 82/107 [02:21<00:27, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 78%|███████▊ | 83/107 [02:23<00:26, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 79%|███████▊ | 84/107 [02:23<00:23, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 79%|███████▉ | 85/107 [02:24<00:21, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 80%|████████ | 86/107 [02:25<00:21, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 81%|████████▏ | 87/107 [02:26<00:19, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 82%|████████▏ | 88/107 [02:27<00:17, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 83%|████████▎ | 89/107 [02:28<00:17, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 84%|████████▍ | 90/107 [02:29<00:18, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 85%|████████▌ | 91/107 [02:30<00:16, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 86%|████████▌ | 92/107 [02:32<00:19, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 87%|████████▋ | 93/107 [02:33<00:15, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 88%|████████▊ | 94/107 [02:35<00:16, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 89%|████████▉ | 95/107 [02:39<00:26, 2.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 90%|████████▉ | 96/107 [02:40<00:20, 1.87s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 91%|█████████ | 97/107 [02:42<00:20, 2.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 92%|█████████▏| 98/107 [02:43<00:14, 1.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 93%|█████████▎| 99/107 [02:44<00:11, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 93%|█████████▎| 100/107 [02:45<00:08, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 94%|█████████▍| 101/107 [02:46<00:06, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 95%|█████████▌| 102/107 [02:47<00:05, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 96%|█████████▋| 103/107 [02:48<00:04, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 97%|█████████▋| 104/107 [02:49<00:02, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 98%|█████████▊| 105/107 [02:49<00:01, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 99%|█████████▉| 106/107 [02:51<00:01, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 92: 100%|██████████| 107/107 [02:56<00:00, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 51%|█████▏ | 55/107 [01:09<00:22, 2.26batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 93: 52%|█████▏ | 56/107 [01:39<07:47, 9.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 53%|█████▎ | 57/107 [01:43<06:28, 7.77s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 54%|█████▍ | 58/107 [01:45<04:44, 5.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 55%|█████▌ | 59/107 [01:50<04:30, 5.64s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 56%|█████▌ | 60/107 [01:52<03:33, 4.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 57%|█████▋ | 61/107 [01:54<02:49, 3.69s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 58%|█████▊ | 62/107 [01:56<02:28, 3.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 59%|█████▉ | 63/107 [01:57<01:55, 2.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 60%|█████▉ | 64/107 [01:58<01:32, 2.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 61%|██████ | 65/107 [01:59<01:18, 1.87s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 62%|██████▏ | 66/107 [02:01<01:15, 1.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 63%|██████▎ | 67/107 [02:02<01:06, 1.67s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 64%|██████▎ | 68/107 [02:03<00:55, 1.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 64%|██████▍ | 69/107 [02:04<00:52, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 65%|██████▌ | 70/107 [02:06<00:48, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 66%|██████▋ | 71/107 [02:08<00:57, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 67%|██████▋ | 72/107 [02:09<00:52, 1.51s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 68%|██████▊ | 73/107 [02:10<00:43, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 69%|██████▉ | 74/107 [02:11<00:41, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 70%|███████ | 75/107 [02:12<00:36, 1.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 71%|███████ | 76/107 [02:13<00:33, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 72%|███████▏ | 77/107 [02:14<00:32, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 73%|███████▎ | 78/107 [02:16<00:41, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 74%|███████▍ | 79/107 [02:17<00:36, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 75%|███████▍ | 80/107 [02:19<00:35, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 76%|███████▌ | 81/107 [02:19<00:29, 1.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 77%|███████▋ | 82/107 [02:20<00:27, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 78%|███████▊ | 83/107 [02:22<00:27, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 79%|███████▊ | 84/107 [02:22<00:24, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 79%|███████▉ | 85/107 [02:23<00:21, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 80%|████████ | 86/107 [02:24<00:21, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 81%|████████▏ | 87/107 [02:25<00:19, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 82%|████████▏ | 88/107 [02:26<00:17, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 83%|████████▎ | 89/107 [02:27<00:17, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 84%|████████▍ | 90/107 [02:29<00:18, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 85%|████████▌ | 91/107 [02:29<00:16, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 86%|████████▌ | 92/107 [02:31<00:19, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 87%|████████▋ | 93/107 [02:32<00:15, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 88%|████████▊ | 94/107 [02:34<00:16, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 89%|████████▉ | 95/107 [02:38<00:27, 2.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 90%|████████▉ | 96/107 [02:39<00:21, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 91%|█████████ | 97/107 [02:42<00:20, 2.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 92%|█████████▏| 98/107 [02:43<00:15, 1.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 93%|█████████▎| 99/107 [02:44<00:12, 1.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 93%|█████████▎| 100/107 [02:45<00:09, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 94%|█████████▍| 101/107 [02:46<00:07, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 95%|█████████▌| 102/107 [02:47<00:06, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 96%|█████████▋| 103/107 [02:48<00:04, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 97%|█████████▋| 104/107 [02:49<00:03, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 98%|█████████▊| 105/107 [02:50<00:02, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 99%|█████████▉| 106/107 [02:51<00:01, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 93: 100%|██████████| 107/107 [02:57<00:00, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 51%|█████▏ | 55/107 [01:09<00:21, 2.40batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 94: 52%|█████▏ | 56/107 [01:38<07:40, 9.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 53%|█████▎ | 57/107 [01:40<05:55, 7.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 54%|█████▍ | 58/107 [01:42<04:23, 5.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 55%|█████▌ | 59/107 [01:44<03:36, 4.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 56%|█████▌ | 60/107 [01:47<03:06, 3.96s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 57%|█████▋ | 61/107 [01:48<02:25, 3.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 58%|█████▊ | 62/107 [01:50<02:09, 2.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 59%|█████▉ | 63/107 [01:52<01:45, 2.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 60%|█████▉ | 64/107 [01:53<01:33, 2.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 61%|██████ | 65/107 [01:55<01:20, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 62%|██████▏ | 66/107 [01:56<01:13, 1.79s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 63%|██████▎ | 67/107 [01:57<01:04, 1.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 64%|██████▎ | 68/107 [01:58<00:54, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 64%|██████▍ | 69/107 [02:00<00:52, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 65%|██████▌ | 70/107 [02:01<00:47, 1.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 66%|██████▋ | 71/107 [02:02<00:50, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 67%|██████▋ | 72/107 [02:04<00:47, 1.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 68%|██████▊ | 73/107 [02:04<00:38, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 69%|██████▉ | 74/107 [02:05<00:38, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 70%|███████ | 75/107 [02:06<00:33, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 71%|███████ | 76/107 [02:08<00:35, 1.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 72%|███████▏ | 77/107 [02:09<00:33, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 73%|███████▎ | 78/107 [02:10<00:37, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 74%|███████▍ | 79/107 [02:11<00:32, 1.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 75%|███████▍ | 80/107 [02:12<00:30, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 76%|███████▌ | 81/107 [02:13<00:25, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 77%|███████▋ | 82/107 [02:14<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 78%|███████▊ | 83/107 [02:15<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 79%|███████▊ | 84/107 [02:15<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 79%|███████▉ | 85/107 [02:16<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 80%|████████ | 86/107 [02:17<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 81%|████████▏ | 87/107 [02:18<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 82%|████████▏ | 88/107 [02:19<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 83%|████████▎ | 89/107 [02:20<00:15, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 84%|████████▍ | 90/107 [02:21<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 85%|████████▌ | 91/107 [02:22<00:14, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 86%|████████▌ | 92/107 [02:23<00:18, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 87%|████████▋ | 93/107 [02:24<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 88%|████████▊ | 94/107 [02:26<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 89%|████████▉ | 95/107 [02:30<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 90%|████████▉ | 96/107 [02:31<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 91%|█████████ | 97/107 [02:33<00:18, 1.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 92%|█████████▏| 98/107 [02:34<00:14, 1.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 93%|█████████▎| 99/107 [02:35<00:10, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 93%|█████████▎| 100/107 [02:36<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 94%|█████████▍| 101/107 [02:36<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 95%|█████████▌| 102/107 [02:37<00:05, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 96%|█████████▋| 103/107 [02:38<00:03, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 97%|█████████▋| 104/107 [02:39<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 98%|█████████▊| 105/107 [02:40<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 99%|█████████▉| 106/107 [02:41<00:00, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 94: 100%|██████████| 107/107 [02:47<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 51%|█████▏ | 55/107 [01:08<00:22, 2.33batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 95: 52%|█████▏ | 56/107 [01:37<07:40, 9.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 53%|█████▎ | 57/107 [01:40<05:57, 7.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 54%|█████▍ | 58/107 [01:42<04:26, 5.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 55%|█████▌ | 59/107 [01:45<03:55, 4.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 56%|█████▌ | 60/107 [01:48<03:16, 4.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 57%|█████▋ | 61/107 [01:49<02:35, 3.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 58%|█████▊ | 62/107 [01:53<02:39, 3.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 59%|█████▉ | 63/107 [01:55<02:06, 2.87s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 60%|█████▉ | 64/107 [01:56<01:44, 2.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 61%|██████ | 65/107 [01:57<01:26, 2.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 62%|██████▏ | 66/107 [01:59<01:22, 2.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 63%|██████▎ | 67/107 [02:00<01:12, 1.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 64%|██████▎ | 68/107 [02:01<01:00, 1.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 64%|██████▍ | 69/107 [02:03<00:55, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 65%|██████▌ | 70/107 [02:04<00:49, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 66%|██████▋ | 71/107 [02:05<00:51, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 67%|██████▋ | 72/107 [02:06<00:47, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 68%|██████▊ | 73/107 [02:08<00:46, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 69%|██████▉ | 74/107 [02:09<00:43, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 70%|███████ | 75/107 [02:10<00:37, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 71%|███████ | 76/107 [02:11<00:33, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 72%|███████▏ | 77/107 [02:12<00:31, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 73%|███████▎ | 78/107 [02:13<00:36, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 74%|███████▍ | 79/107 [02:14<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 75%|███████▍ | 80/107 [02:15<00:29, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 76%|███████▌ | 81/107 [02:16<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 77%|███████▋ | 82/107 [02:17<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 78%|███████▊ | 83/107 [02:18<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 79%|███████▊ | 84/107 [02:19<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 79%|███████▉ | 85/107 [02:19<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 80%|████████ | 86/107 [02:20<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 81%|████████▏ | 87/107 [02:21<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 82%|████████▏ | 88/107 [02:22<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 83%|████████▎ | 89/107 [02:23<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 84%|████████▍ | 90/107 [02:24<00:16, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 85%|████████▌ | 91/107 [02:25<00:14, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 86%|████████▌ | 92/107 [02:26<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 87%|████████▋ | 93/107 [02:27<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 88%|████████▊ | 94/107 [02:29<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 89%|████████▉ | 95/107 [02:33<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 90%|████████▉ | 96/107 [02:34<00:20, 1.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 91%|█████████ | 97/107 [02:36<00:19, 1.96s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 92%|█████████▏| 98/107 [02:37<00:14, 1.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 93%|█████████▎| 99/107 [02:38<00:11, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 93%|█████████▎| 100/107 [02:39<00:08, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 94%|█████████▍| 101/107 [02:40<00:06, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 95%|█████████▌| 102/107 [02:41<00:06, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 96%|█████████▋| 103/107 [02:42<00:04, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 97%|█████████▋| 104/107 [02:43<00:03, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 98%|█████████▊| 105/107 [02:44<00:02, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 99%|█████████▉| 106/107 [02:45<00:01, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 95: 100%|██████████| 107/107 [02:52<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 96: 51%|█████▏ | 55/107 [01:09<00:19, 2.63batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 96: 52%|█████▏ | 56/107 [01:39<07:56, 9.35s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 53%|█████▎ | 57/107 [01:42<06:12, 7.44s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 54%|█████▍ | 58/107 [01:45<04:50, 5.94s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 55%|█████▌ | 59/107 [01:48<04:13, 5.29s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 56%|█████▌ | 60/107 [01:50<03:16, 4.19s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 57%|█████▋ | 61/107 [01:52<02:34, 3.36s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 58%|█████▊ | 62/107 [01:54<02:13, 2.98s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 59%|█████▉ | 63/107 [01:55<01:47, 2.44s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 60%|█████▉ | 64/107 [01:56<01:33, 2.17s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 61%|██████ | 65/107 [01:58<01:30, 2.15s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 62%|██████▏ | 66/107 [02:00<01:21, 1.98s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 63%|██████▎ | 67/107 [02:01<01:09, 1.74s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 64%|██████▎ | 68/107 [02:02<00:58, 1.50s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 64%|██████▍ | 69/107 [02:03<00:53, 1.40s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 65%|██████▌ | 70/107 [02:04<00:48, 1.32s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 66%|██████▋ | 71/107 [02:06<00:51, 1.43s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 67%|██████▋ | 72/107 [02:08<00:53, 1.54s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 68%|██████▊ | 73/107 [02:09<00:43, 1.27s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 69%|██████▉ | 74/107 [02:10<00:41, 1.25s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 70%|███████ | 75/107 [02:11<00:35, 1.10s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 71%|███████ | 76/107 [02:11<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 72%|███████▏ | 77/107 [02:12<00:29, 1.00batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 73%|███████▎ | 78/107 [02:14<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 74%|███████▍ | 79/107 [02:15<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 75%|███████▍ | 80/107 [02:16<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 76%|███████▌ | 81/107 [02:17<00:25, 1.04batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 77%|███████▋ | 82/107 [02:17<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 78%|███████▊ | 83/107 [02:18<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 79%|███████▊ | 84/107 [02:19<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 79%|███████▉ | 85/107 [02:20<00:19, 1.11batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 80%|████████ | 86/107 [02:21<00:20, 1.05batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 81%|████████▏ | 87/107 [02:22<00:19, 1.05batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 82%|████████▏ | 88/107 [02:23<00:17, 1.07batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 83%|████████▎ | 89/107 [02:24<00:17, 1.02batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 84%|████████▍ | 90/107 [02:25<00:18, 1.09s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 85%|████████▌ | 91/107 [02:26<00:16, 1.02s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 86%|████████▌ | 92/107 [02:28<00:19, 1.30s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 87%|████████▋ | 93/107 [02:29<00:16, 1.15s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 88%|████████▊ | 94/107 [02:31<00:17, 1.31s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 89%|████████▉ | 95/107 [02:35<00:27, 2.29s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 90%|████████▉ | 96/107 [02:36<00:21, 1.95s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 91%|█████████ | 97/107 [02:39<00:20, 2.06s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 92%|█████████▏| 98/107 [02:40<00:15, 1.71s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 93%|█████████▎| 99/107 [02:41<00:11, 1.48s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 93%|█████████▎| 100/107 [02:41<00:08, 1.28s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 94%|█████████▍| 101/107 [02:42<00:06, 1.14s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 95%|█████████▌| 102/107 [02:44<00:06, 1.21s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 96%|█████████▋| 103/107 [02:44<00:04, 1.10s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 97%|█████████▋| 104/107 [02:45<00:03, 1.01s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 98%|█████████▊| 105/107 [02:46<00:01, 1.01batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 99%|█████████▉| 106/107 [02:47<00:01, 1.07s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 96: 100%|██████████| 107/107 [02:53<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 97: 51%|█████▏ | 55/107 [01:09<00:28, 1.82batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 97: 52%|█████▏ | 56/107 [01:37<07:33, 8.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 53%|█████▎ | 57/107 [01:42<06:13, 7.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 54%|█████▍ | 58/107 [01:43<04:39, 5.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 55%|█████▌ | 59/107 [01:45<03:40, 4.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 56%|█████▌ | 60/107 [01:47<02:52, 3.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 57%|█████▋ | 61/107 [01:48<02:15, 2.94s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 58%|█████▊ | 62/107 [01:50<02:03, 2.74s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 59%|█████▉ | 63/107 [01:51<01:39, 2.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 60%|█████▉ | 64/107 [01:52<01:22, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 61%|██████ | 65/107 [01:54<01:11, 1.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 62%|██████▏ | 66/107 [01:55<01:07, 1.65s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 63%|██████▎ | 67/107 [01:56<01:00, 1.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 64%|██████▎ | 68/107 [01:57<00:51, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 64%|██████▍ | 69/107 [01:58<00:50, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 65%|██████▌ | 70/107 [02:00<00:46, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 66%|██████▋ | 71/107 [02:02<00:56, 1.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 67%|██████▋ | 72/107 [02:03<00:53, 1.53s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 68%|██████▊ | 73/107 [02:04<00:44, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 69%|██████▉ | 74/107 [02:06<00:44, 1.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 70%|███████ | 75/107 [02:06<00:38, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 71%|███████ | 76/107 [02:07<00:33, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 72%|███████▏ | 77/107 [02:08<00:31, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 73%|███████▎ | 78/107 [02:10<00:36, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 74%|███████▍ | 79/107 [02:11<00:31, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 75%|███████▍ | 80/107 [02:12<00:29, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 76%|███████▌ | 81/107 [02:12<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 77%|███████▋ | 82/107 [02:13<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 78%|███████▊ | 83/107 [02:14<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 79%|███████▊ | 84/107 [02:15<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 79%|███████▉ | 85/107 [02:16<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 80%|████████ | 86/107 [02:17<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 81%|████████▏ | 87/107 [02:17<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 82%|████████▏ | 88/107 [02:18<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 83%|████████▎ | 89/107 [02:19<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 84%|████████▍ | 90/107 [02:20<00:16, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 85%|████████▌ | 91/107 [02:21<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 86%|████████▌ | 92/107 [02:23<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 87%|████████▋ | 93/107 [02:24<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 88%|████████▊ | 94/107 [02:25<00:15, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 89%|████████▉ | 95/107 [02:30<00:28, 2.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 90%|████████▉ | 96/107 [02:31<00:21, 1.98s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 91%|█████████ | 97/107 [02:34<00:20, 2.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 92%|█████████▏| 98/107 [02:35<00:15, 1.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 93%|█████████▎| 99/107 [02:35<00:11, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 93%|█████████▎| 100/107 [02:36<00:08, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 94%|█████████▍| 101/107 [02:37<00:06, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 95%|█████████▌| 102/107 [02:38<00:05, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 96%|█████████▋| 103/107 [02:39<00:04, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 97%|█████████▋| 104/107 [02:40<00:02, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 98%|█████████▊| 105/107 [02:41<00:01, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 99%|█████████▉| 106/107 [02:42<00:00, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 97: 100%|██████████| 107/107 [02:48<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 51%|█████▏ | 55/107 [01:10<00:28, 1.84batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 98: 52%|█████▏ | 56/107 [01:39<07:46, 9.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 53%|█████▎ | 57/107 [01:42<06:10, 7.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 54%|█████▍ | 58/107 [01:44<04:39, 5.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 55%|█████▌ | 59/107 [01:46<03:40, 4.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 56%|█████▌ | 60/107 [01:48<02:53, 3.68s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 57%|█████▋ | 61/107 [01:49<02:10, 2.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 58%|█████▊ | 62/107 [01:53<02:35, 3.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 59%|█████▉ | 63/107 [01:55<02:00, 2.75s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 60%|█████▉ | 64/107 [01:56<01:40, 2.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 61%|██████ | 65/107 [01:57<01:26, 2.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 62%|██████▏ | 66/107 [01:59<01:22, 2.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 63%|██████▎ | 67/107 [02:00<01:11, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 64%|██████▎ | 68/107 [02:01<00:59, 1.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 64%|██████▍ | 69/107 [02:03<00:58, 1.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 65%|██████▌ | 70/107 [02:05<00:57, 1.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 66%|██████▋ | 71/107 [02:06<00:58, 1.63s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 67%|██████▋ | 72/107 [02:08<00:58, 1.68s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 68%|██████▊ | 73/107 [02:09<00:46, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 69%|██████▉ | 74/107 [02:10<00:43, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 70%|███████ | 75/107 [02:11<00:37, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 71%|███████ | 76/107 [02:12<00:33, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 72%|███████▏ | 77/107 [02:13<00:31, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 73%|███████▎ | 78/107 [02:14<00:36, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 74%|███████▍ | 79/107 [02:15<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 75%|███████▍ | 80/107 [02:16<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 76%|███████▌ | 81/107 [02:17<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 77%|███████▋ | 82/107 [02:18<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 78%|███████▊ | 83/107 [02:19<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 79%|███████▊ | 84/107 [02:19<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 79%|███████▉ | 85/107 [02:20<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 80%|████████ | 86/107 [02:21<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 81%|████████▏ | 87/107 [02:22<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 82%|████████▏ | 88/107 [02:23<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 83%|████████▎ | 89/107 [02:24<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 84%|████████▍ | 90/107 [02:25<00:17, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 85%|████████▌ | 91/107 [02:26<00:16, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 86%|████████▌ | 92/107 [02:28<00:20, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 87%|████████▋ | 93/107 [02:29<00:17, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 88%|████████▊ | 94/107 [02:31<00:17, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 89%|████████▉ | 95/107 [02:35<00:27, 2.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 90%|████████▉ | 96/107 [02:36<00:21, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 91%|█████████ | 97/107 [02:39<00:20, 2.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 92%|█████████▏| 98/107 [02:39<00:14, 1.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 93%|█████████▎| 99/107 [02:40<00:11, 1.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 93%|█████████▎| 100/107 [02:41<00:08, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 94%|█████████▍| 101/107 [02:42<00:06, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 95%|█████████▌| 102/107 [02:43<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 96%|█████████▋| 103/107 [02:44<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 97%|█████████▋| 104/107 [02:44<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 98%|█████████▊| 105/107 [02:45<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 99%|█████████▉| 106/107 [02:46<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 98: 100%|██████████| 107/107 [02:52<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 51%|█████▏ | 55/107 [01:10<00:19, 2.62batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 99: 52%|█████▏ | 56/107 [01:39<07:31, 8.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 53%|█████▎ | 57/107 [01:43<06:17, 7.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 54%|█████▍ | 58/107 [01:45<04:40, 5.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 55%|█████▌ | 59/107 [01:46<03:35, 4.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 56%|█████▌ | 60/107 [01:48<02:46, 3.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 57%|█████▋ | 61/107 [01:49<02:09, 2.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 58%|█████▊ | 62/107 [01:54<02:41, 3.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 59%|█████▉ | 63/107 [01:56<02:19, 3.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 60%|█████▉ | 64/107 [01:58<01:52, 2.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 61%|██████ | 65/107 [01:59<01:33, 2.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 62%|██████▏ | 66/107 [02:01<01:26, 2.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 63%|██████▎ | 67/107 [02:02<01:14, 1.86s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 64%|██████▎ | 68/107 [02:03<01:00, 1.56s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 64%|██████▍ | 69/107 [02:04<00:56, 1.49s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 65%|██████▌ | 70/107 [02:05<00:50, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 66%|██████▋ | 71/107 [02:07<00:52, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 67%|██████▋ | 72/107 [02:09<00:53, 1.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 68%|██████▊ | 73/107 [02:10<00:43, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 69%|██████▉ | 74/107 [02:11<00:41, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 70%|███████ | 75/107 [02:12<00:36, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 71%|███████ | 76/107 [02:12<00:31, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 72%|███████▏ | 77/107 [02:13<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 73%|███████▎ | 78/107 [02:15<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 74%|███████▍ | 79/107 [02:16<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 75%|███████▍ | 80/107 [02:17<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 76%|███████▌ | 81/107 [02:18<00:26, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 77%|███████▋ | 82/107 [02:19<00:25, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 78%|███████▊ | 83/107 [02:20<00:25, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 79%|███████▊ | 84/107 [02:21<00:23, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 79%|███████▉ | 85/107 [02:22<00:21, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 80%|████████ | 86/107 [02:23<00:20, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 81%|████████▏ | 87/107 [02:24<00:19, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 82%|████████▏ | 88/107 [02:25<00:18, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 83%|████████▎ | 89/107 [02:25<00:16, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 84%|████████▍ | 90/107 [02:27<00:17, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 85%|████████▌ | 91/107 [02:27<00:15, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 86%|████████▌ | 92/107 [02:29<00:18, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 87%|████████▋ | 93/107 [02:30<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 88%|████████▊ | 94/107 [02:32<00:16, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 89%|████████▉ | 95/107 [02:37<00:28, 2.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 90%|████████▉ | 96/107 [02:38<00:21, 1.97s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 91%|█████████ | 97/107 [02:40<00:20, 2.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 92%|█████████▏| 98/107 [02:41<00:15, 1.68s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 93%|█████████▎| 99/107 [02:42<00:11, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 93%|█████████▎| 100/107 [02:42<00:08, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 94%|█████████▍| 101/107 [02:43<00:06, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 95%|█████████▌| 102/107 [02:45<00:05, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 96%|█████████▋| 103/107 [02:45<00:04, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 97%|█████████▋| 104/107 [02:46<00:02, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 98%|█████████▊| 105/107 [02:47<00:01, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 99%|█████████▉| 106/107 [02:48<00:01, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 99: 100%|██████████| 107/107 [02:54<00:00, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 51%|█████▏ | 55/107 [01:11<00:30, 1.72batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 100: 52%|█████▏ | 56/107 [01:41<07:50, 9.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 53%|█████▎ | 57/107 [01:45<06:29, 7.78s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 54%|█████▍ | 58/107 [01:46<04:45, 5.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 55%|█████▌ | 59/107 [01:49<03:53, 4.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 56%|█████▌ | 60/107 [01:51<03:02, 3.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 57%|█████▋ | 61/107 [01:51<02:17, 3.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 58%|█████▊ | 62/107 [01:55<02:28, 3.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 59%|█████▉ | 63/107 [01:57<01:56, 2.65s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 60%|█████▉ | 64/107 [01:58<01:38, 2.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 61%|██████ | 65/107 [01:59<01:18, 1.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 62%|██████▏ | 66/107 [02:01<01:15, 1.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 63%|██████▎ | 67/107 [02:02<01:04, 1.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 64%|██████▎ | 68/107 [02:03<00:54, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 64%|██████▍ | 69/107 [02:04<00:50, 1.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 65%|██████▌ | 70/107 [02:05<00:45, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 66%|██████▋ | 71/107 [02:07<00:48, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 67%|██████▋ | 72/107 [02:08<00:52, 1.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 68%|██████▊ | 73/107 [02:09<00:42, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 69%|██████▉ | 74/107 [02:10<00:40, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 70%|███████ | 75/107 [02:11<00:34, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 71%|███████ | 76/107 [02:12<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 72%|███████▏ | 77/107 [02:13<00:30, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 73%|███████▎ | 78/107 [02:15<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 74%|███████▍ | 79/107 [02:15<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 75%|███████▍ | 80/107 [02:16<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 76%|███████▌ | 81/107 [02:17<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 77%|███████▋ | 82/107 [02:18<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 78%|███████▊ | 83/107 [02:19<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 79%|███████▊ | 84/107 [02:20<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 79%|███████▉ | 85/107 [02:20<00:18, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 80%|████████ | 86/107 [02:21<00:17, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 81%|████████▏ | 87/107 [02:22<00:16, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 82%|████████▏ | 88/107 [02:23<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 83%|████████▎ | 89/107 [02:24<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 84%|████████▍ | 90/107 [02:25<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 85%|████████▌ | 91/107 [02:26<00:14, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 86%|████████▌ | 92/107 [02:27<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 87%|████████▋ | 93/107 [02:28<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 88%|████████▊ | 94/107 [02:30<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 89%|████████▉ | 95/107 [02:34<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 90%|████████▉ | 96/107 [02:35<00:19, 1.79s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 91%|█████████ | 97/107 [02:37<00:18, 1.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 92%|█████████▏| 98/107 [02:38<00:14, 1.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 93%|█████████▎| 99/107 [02:39<00:10, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 93%|█████████▎| 100/107 [02:40<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 94%|█████████▍| 101/107 [02:40<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 95%|█████████▌| 102/107 [02:42<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 96%|█████████▋| 103/107 [02:42<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 97%|█████████▋| 104/107 [02:43<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 98%|█████████▊| 105/107 [02:44<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 99%|█████████▉| 106/107 [02:45<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 100: 100%|██████████| 107/107 [02:51<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 51%|█████▏ | 55/107 [01:13<00:26, 1.97batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 101: 52%|█████▏ | 56/107 [01:43<07:54, 9.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 53%|█████▎ | 57/107 [01:46<06:16, 7.53s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 54%|█████▍ | 58/107 [01:47<04:40, 5.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 55%|█████▌ | 59/107 [01:49<03:40, 4.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 56%|█████▌ | 60/107 [01:51<02:51, 3.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 57%|█████▋ | 61/107 [01:52<02:18, 3.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 58%|█████▊ | 62/107 [01:55<02:04, 2.77s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 59%|█████▉ | 63/107 [01:56<01:39, 2.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 60%|█████▉ | 64/107 [01:58<01:31, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 61%|██████ | 65/107 [01:59<01:17, 1.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 62%|██████▏ | 66/107 [02:00<01:10, 1.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 63%|██████▎ | 67/107 [02:01<01:01, 1.53s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 64%|██████▎ | 68/107 [02:02<00:52, 1.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 64%|██████▍ | 69/107 [02:03<00:49, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 65%|██████▌ | 70/107 [02:04<00:45, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 66%|██████▋ | 71/107 [02:06<00:48, 1.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 67%|██████▋ | 72/107 [02:08<00:52, 1.51s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 68%|██████▊ | 73/107 [02:09<00:42, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 69%|██████▉ | 74/107 [02:10<00:40, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 70%|███████ | 75/107 [02:10<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 71%|███████ | 76/107 [02:11<00:30, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 72%|███████▏ | 77/107 [02:12<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 73%|███████▎ | 78/107 [02:14<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 74%|███████▍ | 79/107 [02:15<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 75%|███████▍ | 80/107 [02:16<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 76%|███████▌ | 81/107 [02:16<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 77%|███████▋ | 82/107 [02:17<00:22, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 78%|███████▊ | 83/107 [02:18<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 79%|███████▊ | 84/107 [02:19<00:20, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 79%|███████▉ | 85/107 [02:20<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 80%|████████ | 86/107 [02:21<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 81%|████████▏ | 87/107 [02:21<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 82%|████████▏ | 88/107 [02:22<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 83%|████████▎ | 89/107 [02:23<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 84%|████████▍ | 90/107 [02:24<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 85%|████████▌ | 91/107 [02:25<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 86%|████████▌ | 92/107 [02:27<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 87%|████████▋ | 93/107 [02:28<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 88%|████████▊ | 94/107 [02:29<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 89%|████████▉ | 95/107 [02:34<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 90%|████████▉ | 96/107 [02:35<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 91%|█████████ | 97/107 [02:37<00:19, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 92%|█████████▏| 98/107 [02:38<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 93%|█████████▎| 99/107 [02:38<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 93%|█████████▎| 100/107 [02:39<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 94%|█████████▍| 101/107 [02:40<00:06, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 95%|█████████▌| 102/107 [02:41<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 96%|█████████▋| 103/107 [02:42<00:04, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 97%|█████████▋| 104/107 [02:43<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 98%|█████████▊| 105/107 [02:43<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 99%|█████████▉| 106/107 [02:45<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 101: 100%|██████████| 107/107 [02:51<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 51%|█████▏ | 55/107 [01:10<00:30, 1.72batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 102: 52%|█████▏ | 56/107 [01:41<08:00, 9.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 53%|█████▎ | 57/107 [01:44<06:20, 7.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 54%|█████▍ | 58/107 [01:45<04:41, 5.74s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 55%|█████▌ | 59/107 [01:48<03:48, 4.76s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 56%|█████▌ | 60/107 [01:49<02:53, 3.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 57%|█████▋ | 61/107 [01:50<02:10, 2.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 58%|█████▊ | 62/107 [01:55<02:34, 3.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 59%|█████▉ | 63/107 [01:56<02:00, 2.74s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 60%|█████▉ | 64/107 [01:57<01:38, 2.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 61%|██████ | 65/107 [01:58<01:20, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 62%|██████▏ | 66/107 [02:00<01:17, 1.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 63%|██████▎ | 67/107 [02:01<01:07, 1.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 64%|██████▎ | 68/107 [02:02<00:56, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 64%|██████▍ | 69/107 [02:03<00:52, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 65%|██████▌ | 70/107 [02:04<00:47, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 66%|██████▋ | 71/107 [02:06<00:56, 1.56s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 67%|██████▋ | 72/107 [02:08<00:50, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 68%|██████▊ | 73/107 [02:08<00:41, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 69%|██████▉ | 74/107 [02:10<00:40, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 70%|███████ | 75/107 [02:10<00:35, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 71%|███████ | 76/107 [02:11<00:31, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 72%|███████▏ | 77/107 [02:12<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 73%|███████▎ | 78/107 [02:14<00:35, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 74%|███████▍ | 79/107 [02:15<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 75%|███████▍ | 80/107 [02:16<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 76%|███████▌ | 81/107 [02:16<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 77%|███████▋ | 82/107 [02:17<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 78%|███████▊ | 83/107 [02:18<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 79%|███████▊ | 84/107 [02:19<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 79%|███████▉ | 85/107 [02:20<00:18, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 80%|████████ | 86/107 [02:21<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 81%|████████▏ | 87/107 [02:21<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 82%|████████▏ | 88/107 [02:22<00:15, 1.25batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 83%|████████▎ | 89/107 [02:23<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 84%|████████▍ | 90/107 [02:24<00:16, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 85%|████████▌ | 91/107 [02:25<00:14, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 86%|████████▌ | 92/107 [02:27<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 87%|████████▋ | 93/107 [02:28<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 88%|████████▊ | 94/107 [02:29<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 89%|████████▉ | 95/107 [02:33<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 90%|████████▉ | 96/107 [02:34<00:19, 1.79s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 91%|█████████ | 97/107 [02:37<00:19, 1.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 92%|█████████▏| 98/107 [02:37<00:14, 1.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 93%|█████████▎| 99/107 [02:38<00:10, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 93%|█████████▎| 100/107 [02:39<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 94%|█████████▍| 101/107 [02:40<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 95%|█████████▌| 102/107 [02:41<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 96%|█████████▋| 103/107 [02:42<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 97%|█████████▋| 104/107 [02:43<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 98%|█████████▊| 105/107 [02:43<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 99%|█████████▉| 106/107 [02:45<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 102: 100%|██████████| 107/107 [02:51<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 51%|█████▏ | 55/107 [01:10<00:17, 3.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 103: 52%|█████▏ | 56/107 [01:40<07:53, 9.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 53%|█████▎ | 57/107 [01:45<06:34, 7.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 54%|█████▍ | 58/107 [01:46<04:50, 5.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 55%|█████▌ | 59/107 [01:49<03:56, 4.94s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 56%|█████▌ | 60/107 [01:50<03:01, 3.86s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 57%|█████▋ | 61/107 [01:51<02:18, 3.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 58%|█████▊ | 62/107 [01:56<02:31, 3.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 59%|█████▉ | 63/107 [01:57<01:57, 2.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 60%|█████▉ | 64/107 [01:58<01:39, 2.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 61%|██████ | 65/107 [01:59<01:20, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 62%|██████▏ | 66/107 [02:01<01:15, 1.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 63%|██████▎ | 67/107 [02:02<01:07, 1.69s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 64%|██████▎ | 68/107 [02:03<00:55, 1.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 64%|██████▍ | 69/107 [02:04<00:52, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 65%|██████▌ | 70/107 [02:05<00:48, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 66%|██████▋ | 71/107 [02:07<00:56, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 67%|██████▋ | 72/107 [02:09<00:51, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 68%|██████▊ | 73/107 [02:09<00:41, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 69%|██████▉ | 74/107 [02:11<00:39, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 70%|███████ | 75/107 [02:11<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 71%|███████ | 76/107 [02:12<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 72%|███████▏ | 77/107 [02:13<00:29, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 73%|███████▎ | 78/107 [02:15<00:35, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 74%|███████▍ | 79/107 [02:16<00:30, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 75%|███████▍ | 80/107 [02:17<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 76%|███████▌ | 81/107 [02:17<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 77%|███████▋ | 82/107 [02:18<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 78%|███████▊ | 83/107 [02:19<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 79%|███████▊ | 84/107 [02:20<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 79%|███████▉ | 85/107 [02:21<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 80%|████████ | 86/107 [02:22<00:18, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 81%|████████▏ | 87/107 [02:22<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 82%|████████▏ | 88/107 [02:23<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 83%|████████▎ | 89/107 [02:24<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 84%|████████▍ | 90/107 [02:25<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 85%|████████▌ | 91/107 [02:26<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 86%|████████▌ | 92/107 [02:28<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 87%|████████▋ | 93/107 [02:29<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 88%|████████▊ | 94/107 [02:30<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 89%|████████▉ | 95/107 [02:34<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 90%|████████▉ | 96/107 [02:35<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 91%|█████████ | 97/107 [02:38<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 92%|█████████▏| 98/107 [02:38<00:14, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 93%|█████████▎| 99/107 [02:39<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 93%|█████████▎| 100/107 [02:40<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 94%|█████████▍| 101/107 [02:41<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 95%|█████████▌| 102/107 [02:42<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 96%|█████████▋| 103/107 [02:43<00:03, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 97%|█████████▋| 104/107 [02:43<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 98%|█████████▊| 105/107 [02:44<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 99%|█████████▉| 106/107 [02:45<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 103: 100%|██████████| 107/107 [02:51<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 51%|█████▏ | 55/107 [01:12<00:26, 1.97batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 104: 52%|█████▏ | 56/107 [01:42<08:08, 9.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 53%|█████▎ | 57/107 [01:45<06:20, 7.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 54%|█████▍ | 58/107 [01:46<04:37, 5.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 55%|█████▌ | 59/107 [01:50<04:08, 5.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 56%|█████▌ | 60/107 [01:52<03:08, 4.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 57%|█████▋ | 61/107 [01:53<02:22, 3.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 58%|█████▊ | 62/107 [01:57<02:32, 3.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 59%|█████▉ | 63/107 [01:58<01:58, 2.69s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 60%|█████▉ | 64/107 [01:59<01:37, 2.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 61%|██████ | 65/107 [02:00<01:17, 1.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 62%|██████▏ | 66/107 [02:02<01:17, 1.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 63%|██████▎ | 67/107 [02:03<01:07, 1.69s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 64%|██████▎ | 68/107 [02:04<00:56, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 64%|██████▍ | 69/107 [02:05<00:52, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 65%|██████▌ | 70/107 [02:06<00:47, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 66%|██████▋ | 71/107 [02:09<00:55, 1.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 67%|██████▋ | 72/107 [02:10<00:50, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 68%|██████▊ | 73/107 [02:10<00:40, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 69%|██████▉ | 74/107 [02:12<00:39, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 70%|███████ | 75/107 [02:12<00:34, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 71%|███████ | 76/107 [02:13<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 72%|███████▏ | 77/107 [02:14<00:30, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 73%|███████▎ | 78/107 [02:16<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 74%|███████▍ | 79/107 [02:17<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 75%|███████▍ | 80/107 [02:18<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 76%|███████▌ | 81/107 [02:18<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 77%|███████▋ | 82/107 [02:19<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 78%|███████▊ | 83/107 [02:20<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 79%|███████▊ | 84/107 [02:21<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 79%|███████▉ | 85/107 [02:22<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 80%|████████ | 86/107 [02:23<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 81%|████████▏ | 87/107 [02:24<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 82%|████████▏ | 88/107 [02:24<00:15, 1.21batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 83%|████████▎ | 89/107 [02:25<00:15, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 84%|████████▍ | 90/107 [02:26<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 85%|████████▌ | 91/107 [02:27<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 86%|████████▌ | 92/107 [02:29<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 87%|████████▋ | 93/107 [02:30<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 88%|████████▊ | 94/107 [02:31<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 89%|████████▉ | 95/107 [02:36<00:25, 2.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 90%|████████▉ | 96/107 [02:37<00:20, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 91%|█████████ | 97/107 [02:39<00:19, 1.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 92%|█████████▏| 98/107 [02:40<00:14, 1.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 93%|█████████▎| 99/107 [02:41<00:11, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 93%|█████████▎| 100/107 [02:42<00:08, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 94%|█████████▍| 101/107 [02:42<00:06, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 95%|█████████▌| 102/107 [02:44<00:05, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 96%|█████████▋| 103/107 [02:44<00:04, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 97%|█████████▋| 104/107 [02:45<00:02, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 98%|█████████▊| 105/107 [02:46<00:01, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 99%|█████████▉| 106/107 [02:47<00:01, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 104: 100%|██████████| 107/107 [02:53<00:00, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 51%|█████▏ | 55/107 [01:12<00:28, 1.79batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 105: 52%|█████▏ | 56/107 [01:43<08:12, 9.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 53%|█████▎ | 57/107 [01:48<06:43, 8.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 54%|█████▍ | 58/107 [01:49<04:55, 6.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 55%|█████▌ | 59/107 [01:52<04:09, 5.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 56%|█████▌ | 60/107 [01:53<03:06, 3.97s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 57%|█████▋ | 61/107 [01:54<02:23, 3.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 58%|█████▊ | 62/107 [01:58<02:25, 3.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 59%|█████▉ | 63/107 [01:59<01:54, 2.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 60%|█████▉ | 64/107 [02:00<01:35, 2.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 61%|██████ | 65/107 [02:01<01:16, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 62%|██████▏ | 66/107 [02:03<01:11, 1.75s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 63%|██████▎ | 67/107 [02:04<01:09, 1.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 64%|██████▎ | 68/107 [02:05<00:57, 1.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 64%|██████▍ | 69/107 [02:07<00:53, 1.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 65%|██████▌ | 70/107 [02:08<00:48, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 66%|██████▋ | 71/107 [02:09<00:51, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 67%|██████▋ | 72/107 [02:11<00:54, 1.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 68%|██████▊ | 73/107 [02:12<00:44, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 69%|██████▉ | 74/107 [02:13<00:41, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 70%|███████ | 75/107 [02:14<00:36, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 71%|███████ | 76/107 [02:15<00:32, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 72%|███████▏ | 77/107 [02:16<00:30, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 73%|███████▎ | 78/107 [02:18<00:36, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 74%|███████▍ | 79/107 [02:18<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 75%|███████▍ | 80/107 [02:19<00:29, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 76%|███████▌ | 81/107 [02:20<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 77%|███████▋ | 82/107 [02:21<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 78%|███████▊ | 83/107 [02:22<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 79%|███████▊ | 84/107 [02:23<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 79%|███████▉ | 85/107 [02:23<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 80%|████████ | 86/107 [02:24<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 81%|████████▏ | 87/107 [02:25<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 82%|████████▏ | 88/107 [02:26<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 83%|████████▎ | 89/107 [02:27<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 84%|████████▍ | 90/107 [02:28<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 85%|████████▌ | 91/107 [02:29<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 86%|████████▌ | 92/107 [02:31<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 87%|████████▋ | 93/107 [02:31<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 88%|████████▊ | 94/107 [02:33<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 89%|████████▉ | 95/107 [02:37<00:25, 2.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 90%|████████▉ | 96/107 [02:38<00:20, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 91%|█████████ | 97/107 [02:40<00:19, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 92%|█████████▏| 98/107 [02:41<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 93%|█████████▎| 99/107 [02:42<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 93%|█████████▎| 100/107 [02:43<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 94%|█████████▍| 101/107 [02:44<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 95%|█████████▌| 102/107 [02:45<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 96%|█████████▋| 103/107 [02:46<00:04, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 97%|█████████▋| 104/107 [02:46<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 98%|█████████▊| 105/107 [02:47<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 99%|█████████▉| 106/107 [02:48<00:00, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 105: 100%|██████████| 107/107 [02:54<00:00, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 51%|█████▏ | 55/107 [01:12<00:23, 2.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 106: 52%|█████▏ | 56/107 [01:43<08:11, 9.64s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 53%|█████▎ | 57/107 [01:47<06:28, 7.78s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 54%|█████▍ | 58/107 [01:48<04:46, 5.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 55%|█████▌ | 59/107 [01:51<03:56, 4.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 56%|█████▌ | 60/107 [01:52<03:00, 3.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 57%|█████▋ | 61/107 [01:53<02:17, 3.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 58%|█████▊ | 62/107 [01:58<02:37, 3.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 59%|█████▉ | 63/107 [01:59<02:01, 2.76s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 60%|█████▉ | 64/107 [02:00<01:42, 2.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 61%|██████ | 65/107 [02:02<01:24, 2.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 62%|██████▏ | 66/107 [02:03<01:19, 1.94s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 63%|██████▎ | 67/107 [02:04<01:08, 1.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 64%|██████▎ | 68/107 [02:05<00:57, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 64%|██████▍ | 69/107 [02:07<00:52, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 65%|██████▌ | 70/107 [02:08<00:47, 1.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 66%|██████▋ | 71/107 [02:10<00:56, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 67%|██████▋ | 72/107 [02:11<00:51, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 68%|██████▊ | 73/107 [02:12<00:41, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 69%|██████▉ | 74/107 [02:13<00:40, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 70%|███████ | 75/107 [02:14<00:35, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 71%|███████ | 76/107 [02:15<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 72%|███████▏ | 77/107 [02:16<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 73%|███████▎ | 78/107 [02:17<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 74%|███████▍ | 79/107 [02:18<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 75%|███████▍ | 80/107 [02:19<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 76%|███████▌ | 81/107 [02:20<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 77%|███████▋ | 82/107 [02:21<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 78%|███████▊ | 83/107 [02:22<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 79%|███████▊ | 84/107 [02:22<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 79%|███████▉ | 85/107 [02:23<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 80%|████████ | 86/107 [02:24<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 81%|████████▏ | 87/107 [02:25<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 82%|████████▏ | 88/107 [02:26<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 83%|████████▎ | 89/107 [02:26<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 84%|████████▍ | 90/107 [02:28<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 85%|████████▌ | 91/107 [02:28<00:14, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 86%|████████▌ | 92/107 [02:30<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 87%|████████▋ | 93/107 [02:31<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 88%|████████▊ | 94/107 [02:33<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 89%|████████▉ | 95/107 [02:37<00:25, 2.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 90%|████████▉ | 96/107 [02:38<00:19, 1.79s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 91%|█████████ | 97/107 [02:40<00:19, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 92%|█████████▏| 98/107 [02:41<00:14, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 93%|█████████▎| 99/107 [02:42<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 93%|█████████▎| 100/107 [02:43<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 94%|█████████▍| 101/107 [02:43<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 95%|█████████▌| 102/107 [02:45<00:05, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 96%|█████████▋| 103/107 [02:45<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 97%|█████████▋| 104/107 [02:46<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 98%|█████████▊| 105/107 [02:47<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 99%|█████████▉| 106/107 [02:48<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 106: 100%|██████████| 107/107 [02:54<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 51%|█████▏ | 55/107 [01:13<00:33, 1.53batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 107: 52%|█████▏ | 56/107 [01:44<08:22, 9.86s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 53%|█████▎ | 57/107 [01:47<06:33, 7.87s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 54%|█████▍ | 58/107 [01:49<04:48, 5.89s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 55%|█████▌ | 59/107 [01:52<04:10, 5.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 56%|█████▌ | 60/107 [01:54<03:14, 4.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 57%|█████▋ | 61/107 [01:55<02:27, 3.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 58%|█████▊ | 62/107 [01:59<02:29, 3.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 59%|█████▉ | 63/107 [02:00<01:56, 2.65s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 60%|█████▉ | 64/107 [02:01<01:39, 2.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 61%|██████ | 65/107 [02:02<01:21, 1.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 62%|██████▏ | 66/107 [02:04<01:18, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 63%|██████▎ | 67/107 [02:05<01:05, 1.65s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 64%|██████▎ | 68/107 [02:06<00:55, 1.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 64%|██████▍ | 69/107 [02:07<00:52, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 65%|██████▌ | 70/107 [02:09<00:54, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 66%|██████▋ | 71/107 [02:11<00:54, 1.51s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 67%|██████▋ | 72/107 [02:12<00:49, 1.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 68%|██████▊ | 73/107 [02:13<00:40, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 69%|██████▉ | 74/107 [02:14<00:39, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 70%|███████ | 75/107 [02:14<00:34, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 71%|███████ | 76/107 [02:15<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 72%|███████▏ | 77/107 [02:16<00:30, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 73%|███████▎ | 78/107 [02:18<00:35, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 74%|███████▍ | 79/107 [02:19<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 75%|███████▍ | 80/107 [02:20<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 76%|███████▌ | 81/107 [02:21<00:24, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 77%|███████▋ | 82/107 [02:21<00:22, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 78%|███████▊ | 83/107 [02:22<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 79%|███████▊ | 84/107 [02:23<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 79%|███████▉ | 85/107 [02:24<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 80%|████████ | 86/107 [02:25<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 81%|████████▏ | 87/107 [02:26<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 82%|████████▏ | 88/107 [02:26<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 83%|████████▎ | 89/107 [02:27<00:15, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 84%|████████▍ | 90/107 [02:29<00:17, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 85%|████████▌ | 91/107 [02:29<00:15, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 86%|████████▌ | 92/107 [02:31<00:18, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 87%|████████▋ | 93/107 [02:32<00:15, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 88%|████████▊ | 94/107 [02:34<00:15, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 89%|████████▉ | 95/107 [02:38<00:26, 2.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 90%|████████▉ | 96/107 [02:39<00:20, 1.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 91%|█████████ | 97/107 [02:41<00:19, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 92%|█████████▏| 98/107 [02:42<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 93%|█████████▎| 99/107 [02:43<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 93%|█████████▎| 100/107 [02:44<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 94%|█████████▍| 101/107 [02:44<00:06, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 95%|█████████▌| 102/107 [02:46<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 96%|█████████▋| 103/107 [02:47<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 97%|█████████▋| 104/107 [02:47<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 98%|█████████▊| 105/107 [02:48<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 99%|█████████▉| 106/107 [02:49<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 107: 100%|██████████| 107/107 [02:55<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 51%|█████▏ | 55/107 [01:12<00:17, 2.93batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 108: 52%|█████▏ | 56/107 [01:44<08:13, 9.68s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 53%|█████▎ | 57/107 [01:47<06:28, 7.77s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 54%|█████▍ | 58/107 [01:48<04:46, 5.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 55%|█████▌ | 59/107 [01:51<04:00, 5.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 56%|█████▌ | 60/107 [01:53<03:04, 3.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 57%|█████▋ | 61/107 [01:54<02:19, 3.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 58%|█████▊ | 62/107 [01:58<02:31, 3.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 59%|█████▉ | 63/107 [01:59<01:58, 2.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 60%|█████▉ | 64/107 [02:00<01:37, 2.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 61%|██████ | 65/107 [02:02<01:21, 1.94s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 62%|██████▏ | 66/107 [02:03<01:17, 1.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 63%|██████▎ | 67/107 [02:05<01:09, 1.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 64%|██████▎ | 68/107 [02:06<00:57, 1.49s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 64%|██████▍ | 69/107 [02:07<00:54, 1.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 65%|██████▌ | 70/107 [02:08<00:49, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 66%|██████▋ | 71/107 [02:10<00:58, 1.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 67%|██████▋ | 72/107 [02:12<00:52, 1.51s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 68%|██████▊ | 73/107 [02:12<00:43, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 69%|██████▉ | 74/107 [02:13<00:40, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 70%|███████ | 75/107 [02:14<00:35, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 71%|███████ | 76/107 [02:15<00:31, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 72%|███████▏ | 77/107 [02:16<00:31, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 73%|███████▎ | 78/107 [02:18<00:36, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 74%|███████▍ | 79/107 [02:19<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 75%|███████▍ | 80/107 [02:20<00:30, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 76%|███████▌ | 81/107 [02:20<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 77%|███████▋ | 82/107 [02:21<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 78%|███████▊ | 83/107 [02:22<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 79%|███████▊ | 84/107 [02:23<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 79%|███████▉ | 85/107 [02:24<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 80%|████████ | 86/107 [02:25<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 81%|████████▏ | 87/107 [02:25<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 82%|████████▏ | 88/107 [02:26<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 83%|████████▎ | 89/107 [02:27<00:15, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 84%|████████▍ | 90/107 [03:57<07:45, 27.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 85%|████████▌ | 91/107 [03:57<05:11, 19.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 86%|████████▌ | 92/107 [03:59<03:33, 14.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 87%|████████▋ | 93/107 [04:00<02:23, 10.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 88%|████████▊ | 94/107 [04:02<01:39, 7.64s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 89%|████████▉ | 95/107 [04:07<01:20, 6.74s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 90%|████████▉ | 96/107 [04:08<00:55, 5.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 91%|█████████ | 97/107 [04:10<00:42, 4.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 92%|█████████▏| 98/107 [04:11<00:29, 3.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 93%|█████████▎| 99/107 [04:12<00:20, 2.56s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 93%|█████████▎| 100/107 [04:13<00:14, 2.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 94%|█████████▍| 101/107 [04:14<00:10, 1.67s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 95%|█████████▌| 102/107 [04:15<00:07, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 96%|█████████▋| 103/107 [04:16<00:05, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 97%|█████████▋| 104/107 [04:17<00:03, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 98%|█████████▊| 105/107 [04:18<00:02, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 99%|█████████▉| 106/107 [04:19<00:01, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 108: 100%|██████████| 107/107 [04:25<00:00, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 109: 51%|█████▏ | 55/107 [01:12<00:22, 2.34batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 109: 52%|█████▏ | 56/107 [01:44<08:21, 9.84s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 53%|█████▎ | 57/107 [01:47<06:36, 7.94s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 54%|█████▍ | 58/107 [01:49<04:54, 6.01s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 55%|█████▌ | 59/107 [01:51<03:50, 4.81s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 56%|█████▌ | 60/107 [01:52<02:58, 3.80s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 57%|█████▋ | 61/107 [01:53<02:18, 3.00s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 58%|█████▊ | 62/107 [01:56<02:04, 2.76s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 59%|█████▉ | 63/107 [01:57<01:40, 2.28s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 60%|█████▉ | 64/107 [01:58<01:30, 2.12s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 61%|██████ | 65/107 [01:59<01:13, 1.75s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 62%|██████▏ | 66/107 [02:01<01:11, 1.76s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 63%|██████▎ | 67/107 [02:02<01:02, 1.56s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 64%|██████▎ | 68/107 [02:03<00:53, 1.38s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 64%|██████▍ | 69/107 [02:04<00:51, 1.35s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 65%|██████▌ | 70/107 [02:06<00:46, 1.26s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 66%|██████▋ | 71/107 [02:07<00:49, 1.39s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 67%|██████▋ | 72/107 [02:08<00:47, 1.35s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 68%|██████▊ | 73/107 [02:10<00:45, 1.35s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 69%|██████▉ | 74/107 [02:11<00:42, 1.30s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 70%|███████ | 75/107 [02:12<00:36, 1.15s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 71%|███████ | 76/107 [02:13<00:32, 1.06s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 72%|███████▏ | 77/107 [02:14<00:31, 1.04s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 73%|███████▎ | 78/107 [02:15<00:36, 1.26s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 74%|███████▍ | 79/107 [02:16<00:31, 1.11s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 75%|███████▍ | 80/107 [02:17<00:30, 1.12s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 76%|███████▌ | 81/107 [02:18<00:25, 1.02batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 77%|███████▋ | 82/107 [02:19<00:23, 1.06batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 78%|███████▊ | 83/107 [02:20<00:23, 1.02batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 79%|███████▊ | 84/107 [02:21<00:21, 1.09batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 79%|███████▉ | 85/107 [02:21<00:19, 1.13batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 80%|████████ | 86/107 [02:22<00:18, 1.13batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 81%|████████▏ | 87/107 [02:23<00:17, 1.17batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 82%|████████▏ | 88/107 [02:24<00:15, 1.21batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 83%|████████▎ | 89/107 [02:25<00:15, 1.16batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 84%|████████▍ | 90/107 [02:26<00:16, 1.02batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 85%|████████▌ | 91/107 [02:27<00:14, 1.07batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 86%|████████▌ | 92/107 [02:29<00:18, 1.20s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 87%|████████▋ | 93/107 [02:30<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 88%|████████▊ | 94/107 [02:31<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 89%|████████▉ | 95/107 [02:35<00:25, 2.14s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 90%|████████▉ | 96/107 [02:36<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 91%|█████████ | 97/107 [02:39<00:19, 1.93s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 92%|█████████▏| 98/107 [02:39<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 93%|█████████▎| 99/107 [02:40<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 93%|█████████▎| 100/107 [02:41<00:08, 1.20s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 94%|█████████▍| 101/107 [02:42<00:06, 1.06s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 95%|█████████▌| 102/107 [02:43<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 96%|█████████▋| 103/107 [02:44<00:04, 1.03s/batch, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 97%|█████████▋| 104/107 [02:45<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 98%|█████████▊| 105/107 [02:46<00:01, 1.07batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 99%|█████████▉| 106/107 [02:47<00:00, 1.00batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 109: 100%|██████████| 107/107 [02:53<00:00, 1.10batch/s, batch_idx=54, gpu=0, loss=0.020, v_num=zf852wxn]\nEpoch 110: 51%|█████▏ | 55/107 [01:13<00:31, 1.66batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 110: 52%|█████▏ | 56/107 [01:45<08:27, 9.96s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 53%|█████▎ | 57/107 [01:48<06:34, 7.89s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 54%|█████▍ | 58/107 [01:49<04:48, 5.89s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 55%|█████▌ | 59/107 [01:53<04:11, 5.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 56%|█████▌ | 60/107 [01:54<03:13, 4.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 57%|█████▋ | 61/107 [01:55<02:27, 3.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 58%|█████▊ | 62/107 [01:59<02:29, 3.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 59%|█████▉ | 63/107 [02:00<01:56, 2.65s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 60%|█████▉ | 64/107 [02:02<01:40, 2.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 61%|██████ | 65/107 [02:03<01:23, 1.99s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 62%|██████▏ | 66/107 [02:04<01:17, 1.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 63%|██████▎ | 67/107 [02:06<01:08, 1.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 64%|██████▎ | 68/107 [02:07<00:57, 1.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 64%|██████▍ | 69/107 [02:08<00:53, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 65%|██████▌ | 70/107 [02:10<00:54, 1.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 66%|██████▋ | 71/107 [02:11<00:54, 1.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 67%|██████▋ | 72/107 [02:12<00:50, 1.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 68%|██████▊ | 73/107 [02:13<00:40, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 69%|██████▉ | 74/107 [02:14<00:39, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 70%|███████ | 75/107 [02:15<00:34, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 71%|███████ | 76/107 [02:16<00:31, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 72%|███████▏ | 77/107 [02:17<00:29, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 73%|███████▎ | 78/107 [02:19<00:35, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 74%|███████▍ | 79/107 [02:19<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 75%|███████▍ | 80/107 [02:21<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 76%|███████▌ | 81/107 [02:21<00:24, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 77%|███████▋ | 82/107 [02:22<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 78%|███████▊ | 83/107 [02:23<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 79%|███████▊ | 84/107 [02:24<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 79%|███████▉ | 85/107 [02:25<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 80%|████████ | 86/107 [02:25<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 81%|████████▏ | 87/107 [02:26<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 82%|████████▏ | 88/107 [02:27<00:15, 1.23batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 83%|████████▎ | 89/107 [02:28<00:15, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 84%|████████▍ | 90/107 [02:29<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 85%|████████▌ | 91/107 [02:30<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 86%|████████▌ | 92/107 [02:32<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 87%|████████▋ | 93/107 [02:33<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 88%|████████▊ | 94/107 [02:34<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 89%|████████▉ | 95/107 [02:38<00:25, 2.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 90%|████████▉ | 96/107 [02:39<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 91%|█████████ | 97/107 [02:42<00:19, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 92%|█████████▏| 98/107 [02:42<00:14, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 93%|█████████▎| 99/107 [02:43<00:11, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 93%|█████████▎| 100/107 [02:44<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 94%|█████████▍| 101/107 [02:45<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 95%|█████████▌| 102/107 [02:46<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 96%|█████████▋| 103/107 [02:47<00:04, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 97%|█████████▋| 104/107 [02:48<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 98%|█████████▊| 105/107 [02:48<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 99%|█████████▉| 106/107 [02:50<00:01, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 110: 100%|██████████| 107/107 [02:56<00:00, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 51%|█████▏ | 55/107 [01:18<00:37, 1.40batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 111: 52%|█████▏ | 56/107 [01:50<08:42, 10.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 53%|█████▎ | 57/107 [01:53<06:47, 8.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 54%|█████▍ | 58/107 [01:55<05:00, 6.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 55%|█████▌ | 59/107 [01:58<04:17, 5.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 56%|█████▌ | 60/107 [02:00<03:16, 4.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 57%|█████▋ | 61/107 [02:01<02:30, 3.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 58%|█████▊ | 62/107 [02:05<02:39, 3.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 59%|█████▉ | 63/107 [02:06<02:03, 2.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 60%|█████▉ | 64/107 [02:08<01:45, 2.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 61%|██████ | 65/107 [02:09<01:25, 2.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 62%|██████▏ | 66/107 [02:11<01:20, 1.97s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 63%|██████▎ | 67/107 [02:12<01:09, 1.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 64%|██████▎ | 68/107 [02:13<00:57, 1.49s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 64%|██████▍ | 69/107 [02:14<00:53, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 65%|██████▌ | 70/107 [02:16<00:57, 1.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 66%|██████▋ | 71/107 [02:17<00:56, 1.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 67%|██████▋ | 72/107 [02:19<00:51, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 68%|██████▊ | 73/107 [02:19<00:41, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 69%|██████▉ | 74/107 [02:20<00:40, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 70%|███████ | 75/107 [02:21<00:35, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 71%|███████ | 76/107 [02:22<00:32, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 72%|███████▏ | 77/107 [02:23<00:30, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 73%|███████▎ | 78/107 [02:25<00:36, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 74%|███████▍ | 79/107 [02:26<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 75%|███████▍ | 80/107 [02:27<00:30, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 76%|███████▌ | 81/107 [02:28<00:25, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 77%|███████▋ | 82/107 [02:28<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 78%|███████▊ | 83/107 [02:29<00:22, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 79%|███████▊ | 84/107 [02:30<00:20, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 79%|███████▉ | 85/107 [02:31<00:19, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 80%|████████ | 86/107 [02:32<00:18, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 81%|████████▏ | 87/107 [02:33<00:17, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 82%|████████▏ | 88/107 [02:33<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 83%|████████▎ | 89/107 [02:34<00:16, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 84%|████████▍ | 90/107 [02:36<00:17, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 85%|████████▌ | 91/107 [02:36<00:14, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 86%|████████▌ | 92/107 [02:38<00:17, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 87%|████████▋ | 93/107 [02:39<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 88%|████████▊ | 94/107 [02:41<00:15, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 89%|████████▉ | 95/107 [02:45<00:25, 2.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 90%|████████▉ | 96/107 [02:46<00:20, 1.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 91%|█████████ | 97/107 [02:48<00:19, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 92%|█████████▏| 98/107 [02:49<00:14, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 93%|█████████▎| 99/107 [02:50<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 93%|█████████▎| 100/107 [02:51<00:08, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 94%|█████████▍| 101/107 [02:51<00:06, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 95%|█████████▌| 102/107 [02:53<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 96%|█████████▋| 103/107 [02:53<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 97%|█████████▋| 104/107 [02:54<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 98%|█████████▊| 105/107 [02:55<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 99%|█████████▉| 106/107 [02:56<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 111: 100%|██████████| 107/107 [03:02<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 51%|█████▏ | 55/107 [01:13<00:24, 2.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 112: 52%|█████▏ | 56/107 [01:45<08:29, 9.98s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 53%|█████▎ | 57/107 [01:49<06:38, 7.96s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 54%|█████▍ | 58/107 [01:50<04:50, 5.94s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 55%|█████▌ | 59/107 [01:53<04:07, 5.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 56%|█████▌ | 60/107 [01:55<03:13, 4.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 57%|█████▋ | 61/107 [01:56<02:25, 3.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 58%|█████▊ | 62/107 [02:00<02:34, 3.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 59%|█████▉ | 63/107 [02:01<01:59, 2.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 60%|█████▉ | 64/107 [02:03<01:41, 2.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 61%|██████ | 65/107 [02:04<01:24, 2.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 62%|██████▏ | 66/107 [02:05<01:19, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 63%|██████▎ | 67/107 [02:07<01:10, 1.76s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 64%|██████▎ | 68/107 [02:08<00:58, 1.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 64%|██████▍ | 69/107 [02:09<00:54, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 65%|██████▌ | 70/107 [02:10<00:49, 1.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 66%|██████▋ | 71/107 [02:12<00:52, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 67%|██████▋ | 72/107 [02:13<00:48, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 68%|██████▊ | 73/107 [02:14<00:46, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 69%|██████▉ | 74/107 [02:16<00:43, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 70%|███████ | 75/107 [02:17<00:37, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 71%|███████ | 76/107 [02:17<00:33, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 72%|███████▏ | 77/107 [02:18<00:31, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 73%|███████▎ | 78/107 [02:20<00:36, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 74%|███████▍ | 79/107 [02:21<00:30, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 75%|███████▍ | 80/107 [02:22<00:29, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 76%|███████▌ | 81/107 [02:23<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 77%|███████▋ | 82/107 [02:23<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 78%|███████▊ | 83/107 [02:24<00:22, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 79%|███████▊ | 84/107 [02:25<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 79%|███████▉ | 85/107 [02:26<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 80%|████████ | 86/107 [02:27<00:18, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 81%|████████▏ | 87/107 [02:28<00:17, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 82%|████████▏ | 88/107 [02:28<00:15, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 83%|████████▎ | 89/107 [02:29<00:15, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 84%|████████▍ | 90/107 [02:31<00:16, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 85%|████████▌ | 91/107 [02:31<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 86%|████████▌ | 92/107 [02:33<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 87%|████████▋ | 93/107 [02:34<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 88%|████████▊ | 94/107 [02:35<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 89%|████████▉ | 95/107 [02:40<00:26, 2.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 90%|████████▉ | 96/107 [02:41<00:20, 1.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 91%|█████████ | 97/107 [02:43<00:19, 1.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 92%|█████████▏| 98/107 [02:44<00:14, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 93%|█████████▎| 99/107 [02:45<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 93%|█████████▎| 100/107 [02:46<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 94%|█████████▍| 101/107 [02:46<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 95%|█████████▌| 102/107 [02:48<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 96%|█████████▋| 103/107 [02:48<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 97%|█████████▋| 104/107 [02:49<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 98%|█████████▊| 105/107 [02:50<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 99%|█████████▉| 106/107 [02:51<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 112: 100%|██████████| 107/107 [02:57<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 51%|█████▏ | 55/107 [01:15<00:22, 2.30batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 113: 52%|█████▏ | 56/107 [01:48<08:40, 10.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 53%|█████▎ | 57/107 [01:51<06:50, 8.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 54%|█████▍ | 58/107 [01:53<05:03, 6.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 55%|█████▌ | 59/107 [01:55<03:57, 4.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 56%|█████▌ | 60/107 [01:56<03:04, 3.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 57%|█████▋ | 61/107 [01:57<02:22, 3.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 58%|█████▊ | 62/107 [01:59<02:06, 2.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 59%|█████▉ | 63/107 [02:01<01:41, 2.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 60%|█████▉ | 64/107 [02:02<01:30, 2.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 61%|██████ | 65/107 [02:03<01:13, 1.74s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 62%|██████▏ | 66/107 [02:05<01:10, 1.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 63%|██████▎ | 67/107 [02:06<01:02, 1.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 64%|██████▎ | 68/107 [02:07<00:53, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 64%|██████▍ | 69/107 [02:08<00:50, 1.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 65%|██████▌ | 70/107 [02:09<00:46, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 66%|██████▋ | 71/107 [02:11<00:50, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 67%|██████▋ | 72/107 [02:12<00:47, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 68%|██████▊ | 73/107 [02:13<00:39, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 69%|██████▉ | 74/107 [02:14<00:39, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 70%|███████ | 75/107 [02:16<00:40, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 71%|███████ | 76/107 [02:16<00:35, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 72%|███████▏ | 77/107 [02:17<00:32, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 73%|███████▎ | 78/107 [02:19<00:37, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 74%|███████▍ | 79/107 [02:20<00:32, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 75%|███████▍ | 80/107 [02:21<00:30, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 76%|███████▌ | 81/107 [02:22<00:25, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 77%|███████▋ | 82/107 [02:23<00:23, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 78%|███████▊ | 83/107 [02:24<00:23, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 79%|███████▊ | 84/107 [02:24<00:21, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 79%|███████▉ | 85/107 [02:25<00:19, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 80%|████████ | 86/107 [02:26<00:18, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 81%|████████▏ | 87/107 [02:27<00:17, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 82%|████████▏ | 88/107 [02:28<00:16, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 83%|████████▎ | 89/107 [02:29<00:16, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 84%|████████▍ | 90/107 [02:30<00:17, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 85%|████████▌ | 91/107 [02:31<00:15, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 86%|████████▌ | 92/107 [02:33<00:18, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 87%|████████▋ | 93/107 [02:34<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 88%|████████▊ | 94/107 [02:35<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 89%|████████▉ | 95/107 [02:39<00:25, 2.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 90%|████████▉ | 96/107 [02:40<00:20, 1.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 91%|█████████ | 97/107 [02:43<00:19, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 92%|█████████▏| 98/107 [02:43<00:14, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 93%|█████████▎| 99/107 [02:44<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 93%|█████████▎| 100/107 [02:45<00:08, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 94%|█████████▍| 101/107 [02:46<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 95%|█████████▌| 102/107 [02:47<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 96%|█████████▋| 103/107 [02:48<00:04, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 97%|█████████▋| 104/107 [02:48<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 98%|█████████▊| 105/107 [02:49<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 99%|█████████▉| 106/107 [02:51<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 113: 100%|██████████| 107/107 [02:57<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 51%|█████▏ | 55/107 [01:15<00:26, 1.93batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 114: 52%|█████▏ | 56/107 [01:49<08:47, 10.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 53%|█████▎ | 57/107 [01:52<06:47, 8.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 54%|█████▍ | 58/107 [01:53<04:56, 6.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 55%|█████▌ | 59/107 [01:57<04:20, 5.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 56%|█████▌ | 60/107 [01:58<03:20, 4.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 57%|█████▋ | 61/107 [01:59<02:31, 3.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 58%|█████▊ | 62/107 [02:03<02:33, 3.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 59%|█████▉ | 63/107 [02:04<01:59, 2.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 60%|█████▉ | 64/107 [02:06<01:41, 2.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 61%|██████ | 65/107 [02:07<01:23, 1.98s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 62%|██████▏ | 66/107 [02:09<01:19, 1.94s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 63%|██████▎ | 67/107 [02:10<01:06, 1.65s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 64%|██████▎ | 68/107 [02:11<00:55, 1.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 64%|██████▍ | 69/107 [02:12<00:59, 1.56s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 65%|██████▌ | 70/107 [02:14<00:52, 1.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 66%|██████▋ | 71/107 [02:15<00:53, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 67%|██████▋ | 72/107 [02:16<00:48, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 68%|██████▊ | 73/107 [02:17<00:39, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 69%|██████▉ | 74/107 [02:18<00:38, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 70%|███████ | 75/107 [02:19<00:34, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 71%|███████ | 76/107 [02:20<00:31, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 72%|███████▏ | 77/107 [02:21<00:30, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 73%|███████▎ | 78/107 [02:23<00:35, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 74%|███████▍ | 79/107 [02:23<00:30, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 75%|███████▍ | 80/107 [02:24<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 76%|███████▌ | 81/107 [02:25<00:24, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 77%|███████▋ | 82/107 [02:26<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 78%|███████▊ | 83/107 [02:27<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 79%|███████▊ | 84/107 [02:28<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 79%|███████▉ | 85/107 [02:28<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 80%|████████ | 86/107 [02:29<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 81%|████████▏ | 87/107 [02:30<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 82%|████████▏ | 88/107 [02:31<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 83%|████████▎ | 89/107 [02:32<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 84%|████████▍ | 90/107 [02:33<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 85%|████████▌ | 91/107 [02:34<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 86%|████████▌ | 92/107 [02:36<00:17, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 87%|████████▋ | 93/107 [02:36<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 88%|████████▊ | 94/107 [02:38<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 89%|████████▉ | 95/107 [02:42<00:25, 2.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 90%|████████▉ | 96/107 [02:43<00:20, 1.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 91%|█████████ | 97/107 [02:46<00:19, 1.96s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 92%|█████████▏| 98/107 [02:47<00:14, 1.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 93%|█████████▎| 99/107 [02:47<00:11, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 93%|█████████▎| 100/107 [02:48<00:08, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 94%|█████████▍| 101/107 [02:49<00:06, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 95%|█████████▌| 102/107 [02:50<00:05, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 96%|█████████▋| 103/107 [02:51<00:04, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 97%|█████████▋| 104/107 [02:52<00:02, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 98%|█████████▊| 105/107 [02:53<00:01, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 99%|█████████▉| 106/107 [02:54<00:01, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 114: 100%|██████████| 107/107 [03:00<00:00, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 51%|█████▏ | 55/107 [01:15<00:24, 2.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 115: 52%|█████▏ | 56/107 [01:49<08:44, 10.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 53%|█████▎ | 57/107 [01:52<06:47, 8.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 54%|█████▍ | 58/107 [01:53<05:00, 6.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 55%|█████▌ | 59/107 [01:55<03:53, 4.87s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 56%|█████▌ | 60/107 [01:57<03:03, 3.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 57%|█████▋ | 61/107 [01:58<02:21, 3.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 58%|█████▊ | 62/107 [02:00<02:04, 2.76s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 59%|█████▉ | 63/107 [02:01<01:41, 2.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 60%|█████▉ | 64/107 [02:03<01:27, 2.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 61%|██████ | 65/107 [02:04<01:15, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 62%|██████▏ | 66/107 [02:05<01:08, 1.67s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 63%|██████▎ | 67/107 [02:06<01:00, 1.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 64%|██████▎ | 68/107 [02:07<00:52, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 64%|██████▍ | 69/107 [02:09<00:50, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 65%|██████▌ | 70/107 [02:10<00:46, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 66%|██████▋ | 71/107 [02:11<00:49, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 67%|██████▋ | 72/107 [02:13<00:47, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 68%|██████▊ | 73/107 [02:13<00:39, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 69%|██████▉ | 74/107 [02:15<00:39, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 70%|███████ | 75/107 [02:16<00:40, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 71%|███████ | 76/107 [02:17<00:35, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 72%|███████▏ | 77/107 [02:18<00:32, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 73%|███████▎ | 78/107 [02:20<00:37, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 74%|███████▍ | 79/107 [02:21<00:32, 1.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 75%|███████▍ | 80/107 [02:22<00:30, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 76%|███████▌ | 81/107 [02:22<00:25, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 77%|███████▋ | 82/107 [02:23<00:23, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 78%|███████▊ | 83/107 [02:24<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 79%|███████▊ | 84/107 [02:25<00:20, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 79%|███████▉ | 85/107 [02:26<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 80%|████████ | 86/107 [02:26<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 81%|████████▏ | 87/107 [02:27<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 82%|████████▏ | 88/107 [02:28<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 83%|████████▎ | 89/107 [02:29<00:15, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 84%|████████▍ | 90/107 [02:30<00:16, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 85%|████████▌ | 91/107 [02:31<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 86%|████████▌ | 92/107 [02:33<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 87%|████████▋ | 93/107 [02:34<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 88%|████████▊ | 94/107 [02:35<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 89%|████████▉ | 95/107 [02:39<00:25, 2.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 90%|████████▉ | 96/107 [02:41<00:19, 1.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 91%|█████████ | 97/107 [02:43<00:19, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 92%|█████████▏| 98/107 [02:44<00:14, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 93%|█████████▎| 99/107 [02:44<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 93%|█████████▎| 100/107 [02:45<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 94%|█████████▍| 101/107 [02:46<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 95%|█████████▌| 102/107 [02:47<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 96%|█████████▋| 103/107 [02:48<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 97%|█████████▋| 104/107 [02:49<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 98%|█████████▊| 105/107 [02:50<00:01, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 99%|█████████▉| 106/107 [02:51<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 115: 100%|██████████| 107/107 [02:57<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 51%|█████▏ | 55/107 [01:18<00:32, 1.62batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 116: 52%|█████▏ | 56/107 [01:51<08:59, 10.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 53%|█████▎ | 57/107 [01:56<07:16, 8.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 54%|█████▍ | 58/107 [01:57<05:15, 6.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 55%|█████▌ | 59/107 [01:59<04:09, 5.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 56%|█████▌ | 60/107 [02:01<03:08, 4.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 57%|█████▋ | 61/107 [02:01<02:20, 3.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 58%|█████▊ | 62/107 [02:06<02:39, 3.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 59%|█████▉ | 63/107 [02:07<02:04, 2.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 60%|█████▉ | 64/107 [02:09<01:44, 2.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 61%|██████ | 65/107 [02:10<01:25, 2.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 62%|██████▏ | 66/107 [02:11<01:18, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 63%|██████▎ | 67/107 [02:13<01:09, 1.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 64%|██████▎ | 68/107 [02:14<00:57, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 64%|██████▍ | 69/107 [02:15<00:52, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 65%|██████▌ | 70/107 [02:16<00:47, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 66%|██████▋ | 71/107 [02:18<00:57, 1.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 67%|██████▋ | 72/107 [02:19<00:51, 1.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 68%|██████▊ | 73/107 [02:20<00:41, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 69%|██████▉ | 74/107 [02:21<00:42, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 70%|███████ | 75/107 [02:23<00:41, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 71%|███████ | 76/107 [02:24<00:37, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 72%|███████▏ | 77/107 [02:25<00:35, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 73%|███████▎ | 78/107 [02:27<00:39, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 74%|███████▍ | 79/107 [02:27<00:32, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 75%|███████▍ | 80/107 [02:28<00:30, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 76%|███████▌ | 81/107 [02:29<00:25, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 77%|███████▋ | 82/107 [02:30<00:23, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 78%|███████▊ | 83/107 [02:31<00:22, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 79%|███████▊ | 84/107 [02:32<00:20, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 79%|███████▉ | 85/107 [02:32<00:18, 1.17batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 80%|████████ | 86/107 [02:33<00:18, 1.16batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 81%|████████▏ | 87/107 [02:34<00:16, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 82%|████████▏ | 88/107 [02:35<00:15, 1.22batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 83%|████████▎ | 89/107 [02:36<00:15, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 84%|████████▍ | 90/107 [02:37<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 85%|████████▌ | 91/107 [02:38<00:14, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 86%|████████▌ | 92/107 [02:39<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 87%|████████▋ | 93/107 [02:40<00:14, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 88%|████████▊ | 94/107 [02:42<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 89%|████████▉ | 95/107 [02:46<00:25, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 90%|████████▉ | 96/107 [02:47<00:19, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 91%|█████████ | 97/107 [02:49<00:19, 1.90s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 92%|█████████▏| 98/107 [02:50<00:14, 1.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 93%|█████████▎| 99/107 [02:51<00:10, 1.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 93%|█████████▎| 100/107 [02:52<00:08, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 94%|█████████▍| 101/107 [02:52<00:06, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 95%|█████████▌| 102/107 [02:54<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 96%|█████████▋| 103/107 [02:54<00:03, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 97%|█████████▋| 104/107 [02:55<00:02, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 98%|█████████▊| 105/107 [02:56<00:01, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 99%|█████████▉| 106/107 [02:57<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 116: 100%|██████████| 107/107 [03:04<00:00, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 51%|█████▏ | 55/107 [01:19<00:21, 2.47batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 117: 52%|█████▏ | 56/107 [01:54<09:12, 10.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 53%|█████▎ | 57/107 [01:57<07:10, 8.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 54%|█████▍ | 58/107 [01:59<05:23, 6.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 55%|█████▌ | 59/107 [02:02<04:18, 5.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 56%|█████▌ | 60/107 [02:03<03:20, 4.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 57%|█████▋ | 61/107 [02:05<02:35, 3.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 58%|█████▊ | 62/107 [02:07<02:20, 3.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 59%|█████▉ | 63/107 [02:08<01:50, 2.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 60%|█████▉ | 64/107 [02:10<01:32, 2.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 61%|██████ | 65/107 [02:10<01:14, 1.77s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 62%|██████▏ | 66/107 [02:12<01:13, 1.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 63%|██████▎ | 67/107 [02:13<01:04, 1.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 64%|██████▎ | 68/107 [02:14<00:55, 1.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 64%|██████▍ | 69/107 [02:16<00:53, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 65%|██████▌ | 70/107 [02:17<00:49, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 66%|██████▋ | 71/107 [02:19<00:52, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 67%|██████▋ | 72/107 [02:21<00:54, 1.56s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 68%|██████▊ | 73/107 [02:21<00:43, 1.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 69%|██████▉ | 74/107 [02:22<00:41, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 70%|███████ | 75/107 [02:23<00:37, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 71%|███████ | 76/107 [02:24<00:33, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 72%|███████▏ | 77/107 [02:25<00:31, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 73%|███████▎ | 78/107 [02:27<00:39, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 74%|███████▍ | 79/107 [02:28<00:33, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 75%|███████▍ | 80/107 [02:29<00:31, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 76%|███████▌ | 81/107 [02:30<00:26, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 77%|███████▋ | 82/107 [02:31<00:24, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 78%|███████▊ | 83/107 [02:32<00:24, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 79%|███████▊ | 84/107 [02:33<00:22, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 79%|███████▉ | 85/107 [02:34<00:20, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 80%|████████ | 86/107 [02:34<00:19, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 81%|████████▏ | 87/107 [02:35<00:17, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 82%|████████▏ | 88/107 [02:36<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 83%|████████▎ | 89/107 [02:37<00:16, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 84%|████████▍ | 90/107 [02:38<00:17, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 85%|████████▌ | 91/107 [02:39<00:15, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 86%|████████▌ | 92/107 [02:41<00:18, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 87%|████████▋ | 93/107 [02:42<00:15, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 88%|████████▊ | 94/107 [02:43<00:16, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 89%|████████▉ | 95/107 [02:48<00:27, 2.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 90%|████████▉ | 96/107 [02:49<00:21, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 91%|█████████ | 97/107 [02:52<00:20, 2.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 92%|█████████▏| 98/107 [02:52<00:15, 1.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 93%|█████████▎| 99/107 [02:53<00:11, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 93%|█████████▎| 100/107 [02:54<00:08, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 94%|█████████▍| 101/107 [02:55<00:06, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 95%|█████████▌| 102/107 [02:56<00:05, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 96%|█████████▋| 103/107 [02:57<00:04, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 97%|█████████▋| 104/107 [02:58<00:02, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 98%|█████████▊| 105/107 [02:59<00:01, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 99%|█████████▉| 106/107 [03:00<00:01, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 117: 100%|██████████| 107/107 [03:06<00:00, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 51%|█████▏ | 55/107 [01:21<00:22, 2.30batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 118: 52%|█████▏ | 56/107 [01:59<10:02, 11.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 53%|█████▎ | 57/107 [02:02<07:46, 9.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 54%|█████▍ | 58/107 [02:05<05:51, 7.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 55%|█████▌ | 59/107 [02:07<04:41, 5.86s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 56%|█████▌ | 60/107 [02:09<03:32, 4.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 57%|█████▋ | 61/107 [02:10<02:37, 3.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 58%|█████▊ | 62/107 [02:14<02:46, 3.69s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 59%|█████▉ | 63/107 [02:15<02:09, 2.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 60%|█████▉ | 64/107 [02:17<01:47, 2.49s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 61%|██████ | 65/107 [02:19<01:37, 2.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 62%|██████▏ | 66/107 [02:21<01:33, 2.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 63%|██████▎ | 67/107 [02:22<01:20, 2.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 64%|██████▎ | 68/107 [02:23<01:06, 1.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 64%|██████▍ | 69/107 [02:24<01:00, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 65%|██████▌ | 70/107 [02:26<00:53, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 66%|██████▋ | 71/107 [02:27<00:54, 1.51s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 67%|██████▋ | 72/107 [02:28<00:50, 1.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 68%|██████▊ | 73/107 [02:29<00:41, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 69%|██████▉ | 74/107 [02:31<00:48, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 70%|███████ | 75/107 [02:32<00:41, 1.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 71%|███████ | 76/107 [02:33<00:36, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 72%|███████▏ | 77/107 [02:34<00:34, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 73%|███████▎ | 78/107 [02:36<00:40, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 74%|███████▍ | 79/107 [02:37<00:34, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 75%|███████▍ | 80/107 [02:38<00:33, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 76%|███████▌ | 81/107 [02:39<00:28, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 77%|███████▋ | 82/107 [02:40<00:25, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 78%|███████▊ | 83/107 [02:41<00:25, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 79%|███████▊ | 84/107 [02:42<00:22, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 79%|███████▉ | 85/107 [02:42<00:20, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 80%|████████ | 86/107 [02:43<00:19, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 81%|████████▏ | 87/107 [02:44<00:18, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 82%|████████▏ | 88/107 [02:45<00:16, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 83%|████████▎ | 89/107 [02:46<00:16, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 84%|████████▍ | 90/107 [02:47<00:17, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 85%|████████▌ | 91/107 [02:48<00:15, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 86%|████████▌ | 92/107 [02:50<00:19, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 87%|████████▋ | 93/107 [02:51<00:17, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 88%|████████▊ | 94/107 [02:54<00:19, 1.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 89%|████████▉ | 95/107 [03:00<00:35, 2.96s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 90%|████████▉ | 96/107 [03:01<00:27, 2.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 91%|█████████ | 97/107 [03:04<00:25, 2.53s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 92%|█████████▏| 98/107 [03:05<00:18, 2.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 93%|█████████▎| 99/107 [03:06<00:13, 1.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 93%|█████████▎| 100/107 [03:06<00:10, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 94%|█████████▍| 101/107 [03:07<00:07, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 95%|█████████▌| 102/107 [03:09<00:06, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 96%|█████████▋| 103/107 [03:09<00:04, 1.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 97%|█████████▋| 104/107 [03:10<00:03, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 98%|█████████▊| 105/107 [03:11<00:02, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 99%|█████████▉| 106/107 [03:13<00:01, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 118: 100%|██████████| 107/107 [03:19<00:00, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 51%|█████▏ | 55/107 [01:25<00:29, 1.77batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 119: 52%|█████▏ | 56/107 [02:03<10:08, 11.94s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 53%|█████▎ | 57/107 [02:07<07:54, 9.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 54%|█████▍ | 58/107 [02:08<05:46, 7.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 55%|█████▌ | 59/107 [02:13<05:00, 6.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 56%|█████▌ | 60/107 [02:14<03:48, 4.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 57%|█████▋ | 61/107 [02:16<02:55, 3.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 58%|█████▊ | 62/107 [02:18<02:29, 3.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 59%|█████▉ | 63/107 [02:19<01:58, 2.69s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 60%|█████▉ | 64/107 [02:21<01:41, 2.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 61%|██████ | 65/107 [02:21<01:20, 1.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 62%|██████▏ | 66/107 [02:23<01:17, 1.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 63%|██████▎ | 67/107 [02:24<01:06, 1.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 64%|██████▎ | 68/107 [02:25<00:56, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 64%|██████▍ | 69/107 [02:27<00:52, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 65%|██████▌ | 70/107 [02:28<00:48, 1.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 66%|██████▋ | 71/107 [02:29<00:51, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 67%|██████▋ | 72/107 [02:31<00:54, 1.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 68%|██████▊ | 73/107 [02:32<00:44, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 69%|██████▉ | 74/107 [02:33<00:42, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 70%|███████ | 75/107 [02:34<00:37, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 71%|███████ | 76/107 [02:35<00:33, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 72%|███████▏ | 77/107 [02:36<00:31, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 73%|███████▎ | 78/107 [02:38<00:38, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 74%|███████▍ | 79/107 [02:39<00:32, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 75%|███████▍ | 80/107 [02:40<00:31, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 76%|███████▌ | 81/107 [02:41<00:26, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 77%|███████▋ | 82/107 [02:42<00:24, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 78%|███████▊ | 83/107 [02:43<00:24, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 79%|███████▊ | 84/107 [02:43<00:22, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 79%|███████▉ | 85/107 [02:44<00:20, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 80%|████████ | 86/107 [02:45<00:19, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 81%|████████▏ | 87/107 [02:46<00:17, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 82%|████████▏ | 88/107 [02:47<00:16, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 83%|████████▎ | 89/107 [02:48<00:16, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 84%|████████▍ | 90/107 [02:49<00:17, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 85%|████████▌ | 91/107 [02:50<00:15, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 86%|████████▌ | 92/107 [02:52<00:18, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 87%|████████▋ | 93/107 [02:53<00:15, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 88%|████████▊ | 94/107 [02:54<00:16, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 89%|████████▉ | 95/107 [02:59<00:27, 2.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 90%|████████▉ | 96/107 [03:00<00:21, 1.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 91%|█████████ | 97/107 [03:02<00:20, 2.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 92%|█████████▏| 98/107 [03:03<00:15, 1.68s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 93%|█████████▎| 99/107 [03:04<00:11, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 93%|█████████▎| 100/107 [03:05<00:08, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 94%|█████████▍| 101/107 [03:06<00:06, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 95%|█████████▌| 102/107 [03:07<00:06, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 96%|█████████▋| 103/107 [03:08<00:04, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 97%|█████████▋| 104/107 [03:09<00:03, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 98%|█████████▊| 105/107 [03:10<00:02, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 99%|█████████▉| 106/107 [03:11<00:01, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 119: 100%|██████████| 107/107 [03:18<00:00, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 51%|█████▏ | 55/107 [01:22<00:34, 1.50batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 120: 52%|█████▏ | 56/107 [01:58<09:24, 11.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 53%|█████▎ | 57/107 [02:02<07:34, 9.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 54%|█████▍ | 58/107 [02:03<05:28, 6.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 55%|█████▌ | 59/107 [02:06<04:28, 5.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 56%|█████▌ | 60/107 [02:08<03:23, 4.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 57%|█████▋ | 61/107 [02:09<02:32, 3.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 58%|█████▊ | 62/107 [02:11<02:23, 3.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 59%|█████▉ | 63/107 [02:13<01:53, 2.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 60%|█████▉ | 64/107 [02:14<01:34, 2.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 61%|██████ | 65/107 [02:15<01:21, 1.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 62%|██████▏ | 66/107 [02:17<01:19, 1.94s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 63%|██████▎ | 67/107 [02:19<01:18, 1.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 64%|██████▎ | 68/107 [02:20<01:07, 1.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 64%|██████▍ | 69/107 [02:22<01:07, 1.78s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 65%|██████▌ | 70/107 [02:24<01:01, 1.67s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 66%|██████▋ | 71/107 [02:26<01:03, 1.77s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 67%|██████▋ | 72/107 [02:27<01:00, 1.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 68%|██████▊ | 73/107 [02:29<00:56, 1.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 69%|██████▉ | 74/107 [02:30<00:51, 1.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 70%|███████ | 75/107 [02:31<00:42, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 71%|███████ | 76/107 [02:32<00:36, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 72%|███████▏ | 77/107 [02:33<00:34, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 73%|███████▎ | 78/107 [02:35<00:39, 1.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 74%|███████▍ | 79/107 [02:36<00:33, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 75%|███████▍ | 80/107 [02:37<00:32, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 76%|███████▌ | 81/107 [02:37<00:26, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 77%|███████▋ | 82/107 [02:38<00:24, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 78%|███████▊ | 83/107 [02:39<00:23, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 79%|███████▊ | 84/107 [02:40<00:21, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 79%|███████▉ | 85/107 [02:41<00:19, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 80%|████████ | 86/107 [02:42<00:18, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 81%|████████▏ | 87/107 [02:43<00:17, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 82%|████████▏ | 88/107 [02:43<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 83%|████████▎ | 89/107 [02:44<00:15, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 84%|████████▍ | 90/107 [02:46<00:16, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 85%|████████▌ | 91/107 [02:46<00:14, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 86%|████████▌ | 92/107 [02:48<00:17, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 87%|████████▋ | 93/107 [02:49<00:14, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 88%|████████▊ | 94/107 [02:50<00:15, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 89%|████████▉ | 95/107 [02:55<00:26, 2.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 90%|████████▉ | 96/107 [02:56<00:20, 1.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 91%|█████████ | 97/107 [02:58<00:19, 1.94s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 92%|█████████▏| 98/107 [02:59<00:14, 1.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 93%|█████████▎| 99/107 [03:00<00:11, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 93%|█████████▎| 100/107 [03:01<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 94%|█████████▍| 101/107 [03:01<00:06, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 95%|█████████▌| 102/107 [03:03<00:05, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 96%|█████████▋| 103/107 [03:03<00:04, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 97%|█████████▋| 104/107 [03:04<00:02, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 98%|█████████▊| 105/107 [03:05<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 99%|█████████▉| 106/107 [03:06<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 120: 100%|██████████| 107/107 [03:13<00:00, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 51%|█████▏ | 55/107 [01:20<00:27, 1.86batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 121: 52%|█████▏ | 56/107 [01:56<09:28, 11.15s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 53%|█████▎ | 57/107 [01:59<07:21, 8.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 54%|█████▍ | 58/107 [02:00<05:22, 6.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 55%|█████▌ | 59/107 [02:06<05:02, 6.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 56%|█████▌ | 60/107 [02:07<03:47, 4.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 57%|█████▋ | 61/107 [02:08<02:50, 3.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 58%|█████▊ | 62/107 [02:12<02:42, 3.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 59%|█████▉ | 63/107 [02:13<02:04, 2.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 60%|█████▉ | 64/107 [02:14<01:43, 2.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 61%|██████ | 65/107 [02:16<01:26, 2.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 62%|██████▏ | 66/107 [02:18<01:34, 2.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 63%|██████▎ | 67/107 [02:20<01:22, 2.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 64%|██████▎ | 68/107 [02:21<01:06, 1.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 64%|██████▍ | 69/107 [02:23<01:07, 1.77s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 65%|██████▌ | 70/107 [02:24<00:58, 1.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 66%|██████▋ | 71/107 [02:25<00:56, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 67%|██████▋ | 72/107 [02:27<00:51, 1.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 68%|██████▊ | 73/107 [02:27<00:41, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 69%|██████▉ | 74/107 [02:28<00:39, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 70%|███████ | 75/107 [02:29<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 71%|███████ | 76/107 [02:30<00:30, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 72%|███████▏ | 77/107 [02:31<00:29, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 73%|███████▎ | 78/107 [02:33<00:35, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 74%|███████▍ | 79/107 [02:33<00:30, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 75%|███████▍ | 80/107 [02:35<00:29, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 76%|███████▌ | 81/107 [02:35<00:24, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 77%|███████▋ | 82/107 [02:36<00:22, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 78%|███████▊ | 83/107 [02:37<00:22, 1.07batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 79%|███████▊ | 84/107 [02:38<00:20, 1.14batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 79%|███████▉ | 85/107 [02:38<00:18, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 80%|████████ | 86/107 [02:39<00:17, 1.18batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 81%|████████▏ | 87/107 [02:40<00:16, 1.20batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 82%|████████▏ | 88/107 [02:41<00:15, 1.24batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 83%|████████▎ | 89/107 [02:42<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 84%|████████▍ | 90/107 [02:43<00:16, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 85%|████████▌ | 91/107 [02:44<00:14, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 86%|████████▌ | 92/107 [02:46<00:17, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 87%|████████▋ | 93/107 [02:46<00:14, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 88%|████████▊ | 94/107 [02:48<00:15, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 89%|████████▉ | 95/107 [02:53<00:28, 2.37s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 90%|████████▉ | 96/107 [02:54<00:22, 2.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 91%|█████████ | 97/107 [02:57<00:21, 2.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 92%|█████████▏| 98/107 [02:58<00:16, 1.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 93%|█████████▎| 99/107 [02:59<00:12, 1.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 93%|█████████▎| 100/107 [03:00<00:09, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 94%|█████████▍| 101/107 [03:01<00:07, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 95%|█████████▌| 102/107 [03:02<00:06, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 96%|█████████▋| 103/107 [03:03<00:04, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 97%|█████████▋| 104/107 [03:04<00:03, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 98%|█████████▊| 105/107 [03:05<00:01, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 99%|█████████▉| 106/107 [03:06<00:01, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 121: 100%|██████████| 107/107 [03:13<00:00, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 51%|█████▏ | 55/107 [01:19<00:21, 2.44batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 122: 52%|█████▏ | 56/107 [01:57<09:52, 11.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 53%|█████▎ | 57/107 [02:00<07:38, 9.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 54%|█████▍ | 58/107 [02:02<05:35, 6.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 55%|█████▌ | 59/107 [02:05<04:32, 5.67s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 56%|█████▌ | 60/107 [02:07<03:31, 4.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 57%|█████▋ | 61/107 [02:08<02:41, 3.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 58%|█████▊ | 62/107 [02:10<02:24, 3.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 59%|█████▉ | 63/107 [02:12<01:55, 2.63s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 60%|█████▉ | 64/107 [02:14<01:44, 2.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 61%|██████ | 65/107 [02:15<01:26, 2.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 62%|██████▏ | 66/107 [02:17<01:23, 2.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 63%|██████▎ | 67/107 [02:18<01:10, 1.77s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 64%|██████▎ | 68/107 [02:19<00:58, 1.51s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 64%|██████▍ | 69/107 [02:20<00:54, 1.44s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 65%|██████▌ | 70/107 [02:21<00:49, 1.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 66%|██████▋ | 71/107 [02:23<00:52, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 67%|██████▋ | 72/107 [02:24<00:48, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 68%|██████▊ | 73/107 [02:25<00:39, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 69%|██████▉ | 74/107 [02:26<00:39, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 70%|███████ | 75/107 [02:27<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 71%|███████ | 76/107 [02:28<00:32, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 72%|███████▏ | 77/107 [02:29<00:36, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 73%|███████▎ | 78/107 [02:31<00:40, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 74%|███████▍ | 79/107 [02:32<00:33, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 75%|███████▍ | 80/107 [02:33<00:31, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 76%|███████▌ | 81/107 [02:34<00:26, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 77%|███████▋ | 82/107 [02:35<00:25, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 78%|███████▊ | 83/107 [02:36<00:24, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 79%|███████▊ | 84/107 [02:37<00:21, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 79%|███████▉ | 85/107 [02:37<00:19, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 80%|████████ | 86/107 [02:38<00:18, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 81%|████████▏ | 87/107 [02:39<00:17, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 82%|████████▏ | 88/107 [02:40<00:15, 1.19batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 83%|████████▎ | 89/107 [02:41<00:15, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 84%|████████▍ | 90/107 [02:42<00:16, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 85%|████████▌ | 91/107 [02:43<00:14, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 86%|████████▌ | 92/107 [02:45<00:18, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 87%|████████▋ | 93/107 [02:45<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 88%|████████▊ | 94/107 [02:47<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 89%|████████▉ | 95/107 [02:52<00:27, 2.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 90%|████████▉ | 96/107 [02:53<00:21, 2.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 91%|█████████ | 97/107 [02:55<00:20, 2.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 92%|█████████▏| 98/107 [02:56<00:15, 1.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 93%|█████████▎| 99/107 [02:57<00:11, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 93%|█████████▎| 100/107 [02:58<00:08, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 94%|█████████▍| 101/107 [02:59<00:07, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 95%|█████████▌| 102/107 [03:01<00:07, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 96%|█████████▋| 103/107 [03:02<00:05, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 97%|█████████▋| 104/107 [03:03<00:03, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 98%|█████████▊| 105/107 [03:03<00:02, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 99%|█████████▉| 106/107 [03:05<00:01, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 122: 100%|██████████| 107/107 [03:11<00:00, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 51%|█████▏ | 55/107 [01:21<00:25, 2.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 123: 52%|█████▏ | 56/107 [01:57<09:32, 11.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 53%|█████▎ | 57/107 [02:01<07:41, 9.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 54%|█████▍ | 58/107 [02:03<05:36, 6.86s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 55%|█████▌ | 59/107 [02:06<04:36, 5.76s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 56%|█████▌ | 60/107 [02:07<03:25, 4.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 57%|█████▋ | 61/107 [02:08<02:37, 3.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 58%|█████▊ | 62/107 [02:12<02:35, 3.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 59%|█████▉ | 63/107 [02:14<02:09, 2.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 60%|█████▉ | 64/107 [02:16<02:00, 2.81s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 61%|██████ | 65/107 [02:18<01:45, 2.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 62%|██████▏ | 66/107 [02:20<01:37, 2.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 63%|██████▎ | 67/107 [02:22<01:27, 2.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 64%|██████▎ | 68/107 [02:23<01:11, 1.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 64%|██████▍ | 69/107 [02:24<01:05, 1.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 65%|██████▌ | 70/107 [02:25<00:58, 1.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 66%|██████▋ | 71/107 [02:28<01:08, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 67%|██████▋ | 72/107 [02:29<01:00, 1.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 68%|██████▊ | 73/107 [02:30<00:48, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 69%|██████▉ | 74/107 [02:31<00:45, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 70%|███████ | 75/107 [02:32<00:38, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 71%|███████ | 76/107 [02:33<00:33, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 72%|███████▏ | 77/107 [02:34<00:31, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 73%|███████▎ | 78/107 [02:36<00:36, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 74%|███████▍ | 79/107 [02:37<00:31, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 75%|███████▍ | 80/107 [02:38<00:29, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 76%|███████▌ | 81/107 [02:38<00:25, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 77%|███████▋ | 82/107 [02:39<00:22, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 78%|███████▊ | 83/107 [02:40<00:23, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 79%|███████▊ | 84/107 [02:41<00:20, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 79%|███████▉ | 85/107 [02:42<00:19, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 80%|████████ | 86/107 [02:43<00:18, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 81%|████████▏ | 87/107 [02:43<00:17, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 82%|████████▏ | 88/107 [02:44<00:15, 1.21batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 83%|████████▎ | 89/107 [02:45<00:15, 1.15batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 84%|████████▍ | 90/107 [02:46<00:16, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 85%|████████▌ | 91/107 [02:47<00:14, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 86%|████████▌ | 92/107 [02:49<00:18, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 87%|████████▋ | 93/107 [02:50<00:15, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 88%|████████▊ | 94/107 [02:51<00:15, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 89%|████████▉ | 95/107 [02:56<00:26, 2.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 90%|████████▉ | 96/107 [02:57<00:20, 1.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 91%|█████████ | 97/107 [02:59<00:19, 1.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 92%|█████████▏| 98/107 [03:00<00:14, 1.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 93%|█████████▎| 99/107 [03:01<00:11, 1.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 93%|█████████▎| 100/107 [03:02<00:08, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 94%|█████████▍| 101/107 [03:02<00:06, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 95%|█████████▌| 102/107 [03:04<00:05, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 96%|█████████▋| 103/107 [03:04<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 97%|█████████▋| 104/107 [03:05<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 98%|█████████▊| 105/107 [03:06<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 99%|█████████▉| 106/107 [03:07<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 123: 100%|██████████| 107/107 [03:14<00:00, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 51%|█████▏ | 55/107 [01:20<00:30, 1.69batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 124: 52%|█████▏ | 56/107 [01:56<09:32, 11.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 53%|█████▎ | 57/107 [01:58<07:07, 8.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 54%|█████▍ | 58/107 [02:01<05:32, 6.78s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 55%|█████▌ | 59/107 [02:05<04:39, 5.82s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 56%|█████▌ | 60/107 [02:06<03:33, 4.55s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 57%|█████▋ | 61/107 [02:07<02:42, 3.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 58%|█████▊ | 62/107 [02:10<02:32, 3.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 59%|█████▉ | 63/107 [02:12<02:04, 2.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 60%|█████▉ | 64/107 [02:14<01:49, 2.54s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 61%|██████ | 65/107 [02:15<01:32, 2.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 62%|██████▏ | 66/107 [02:17<01:22, 2.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 63%|██████▎ | 67/107 [02:18<01:08, 1.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 64%|██████▎ | 68/107 [02:19<00:57, 1.49s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 64%|██████▍ | 69/107 [02:20<00:53, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 65%|██████▌ | 70/107 [02:22<01:00, 1.63s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 66%|██████▋ | 71/107 [02:24<01:02, 1.74s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 67%|██████▋ | 72/107 [02:25<00:55, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 68%|██████▊ | 73/107 [02:26<00:44, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 69%|██████▉ | 74/107 [02:27<00:42, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 70%|███████ | 75/107 [02:28<00:36, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 71%|███████ | 76/107 [02:29<00:31, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 72%|███████▏ | 77/107 [02:30<00:30, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 73%|███████▎ | 78/107 [02:32<00:36, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 74%|███████▍ | 79/107 [02:32<00:31, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 75%|███████▍ | 80/107 [02:34<00:31, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 76%|███████▌ | 81/107 [02:34<00:27, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 77%|███████▋ | 82/107 [02:35<00:25, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 78%|███████▊ | 83/107 [02:37<00:25, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 79%|███████▊ | 84/107 [02:38<00:23, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 79%|███████▉ | 85/107 [02:38<00:21, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 80%|████████ | 86/107 [02:39<00:20, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 81%|████████▏ | 87/107 [02:40<00:19, 1.03batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 82%|████████▏ | 88/107 [02:41<00:17, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 83%|████████▎ | 89/107 [02:42<00:16, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 84%|████████▍ | 90/107 [02:43<00:17, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 85%|████████▌ | 91/107 [02:44<00:15, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 86%|████████▌ | 92/107 [02:46<00:17, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 87%|████████▋ | 93/107 [02:47<00:15, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 88%|████████▊ | 94/107 [02:48<00:15, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 89%|████████▉ | 95/107 [02:53<00:25, 2.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 90%|████████▉ | 96/107 [02:54<00:20, 1.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 91%|█████████ | 97/107 [02:56<00:19, 1.94s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 92%|█████████▏| 98/107 [02:57<00:14, 1.60s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 93%|█████████▎| 99/107 [02:58<00:11, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 93%|█████████▎| 100/107 [02:58<00:08, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 94%|█████████▍| 101/107 [02:59<00:06, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 95%|█████████▌| 102/107 [03:01<00:06, 1.28s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 96%|█████████▋| 103/107 [03:02<00:04, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 97%|█████████▋| 104/107 [03:03<00:03, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 98%|█████████▊| 105/107 [03:04<00:02, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 99%|█████████▉| 106/107 [03:05<00:01, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 124: 100%|██████████| 107/107 [03:13<00:00, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 51%|█████▏ | 55/107 [01:21<00:34, 1.53batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 125: 52%|█████▏ | 56/107 [02:00<10:21, 12.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 53%|█████▎ | 57/107 [02:06<08:35, 10.31s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 54%|█████▍ | 58/107 [02:08<06:18, 7.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 55%|█████▌ | 59/107 [02:10<04:45, 5.96s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 56%|█████▌ | 60/107 [02:11<03:34, 4.57s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 57%|█████▋ | 61/107 [02:12<02:44, 3.58s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 58%|█████▊ | 62/107 [02:16<02:43, 3.63s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 59%|█████▉ | 63/107 [02:17<02:03, 2.80s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 60%|█████▉ | 64/107 [02:18<01:45, 2.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 61%|██████ | 65/107 [02:20<01:27, 2.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 62%|██████▏ | 66/107 [02:22<01:22, 2.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 63%|██████▎ | 67/107 [02:23<01:15, 1.89s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 64%|██████▎ | 68/107 [02:25<01:07, 1.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 64%|██████▍ | 69/107 [02:26<01:04, 1.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 65%|██████▌ | 70/107 [02:27<00:56, 1.53s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 66%|██████▋ | 71/107 [02:29<00:57, 1.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 67%|██████▋ | 72/107 [02:30<00:52, 1.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 68%|██████▊ | 73/107 [02:32<00:49, 1.47s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 69%|██████▉ | 74/107 [02:33<00:45, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 70%|███████ | 75/107 [02:34<00:38, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 71%|███████ | 76/107 [02:35<00:34, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 72%|███████▏ | 77/107 [02:36<00:32, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 73%|███████▎ | 78/107 [02:37<00:37, 1.29s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 74%|███████▍ | 79/107 [02:38<00:31, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 75%|███████▍ | 80/107 [02:39<00:30, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 76%|███████▌ | 81/107 [02:40<00:25, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 77%|███████▋ | 82/107 [02:41<00:23, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 78%|███████▊ | 83/107 [02:42<00:23, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 79%|███████▊ | 84/107 [02:43<00:21, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 79%|███████▉ | 85/107 [02:44<00:19, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 80%|████████ | 86/107 [02:44<00:18, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 81%|████████▏ | 87/107 [02:45<00:17, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 82%|████████▏ | 88/107 [02:46<00:16, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 83%|████████▎ | 89/107 [02:47<00:16, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 84%|████████▍ | 90/107 [02:48<00:17, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 85%|████████▌ | 91/107 [02:49<00:15, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 86%|████████▌ | 92/107 [02:51<00:18, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 87%|████████▋ | 93/107 [02:52<00:15, 1.14s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 88%|████████▊ | 94/107 [02:54<00:16, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 89%|████████▉ | 95/107 [02:58<00:26, 2.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 90%|████████▉ | 96/107 [02:59<00:20, 1.87s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 91%|█████████ | 97/107 [03:01<00:19, 1.99s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 92%|█████████▏| 98/107 [03:02<00:14, 1.66s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 93%|█████████▎| 99/107 [03:03<00:11, 1.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 93%|█████████▎| 100/107 [03:04<00:08, 1.24s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 94%|█████████▍| 101/107 [03:05<00:06, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 95%|█████████▌| 102/107 [03:06<00:05, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 96%|█████████▋| 103/107 [03:07<00:04, 1.04s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 97%|█████████▋| 104/107 [03:07<00:02, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 98%|█████████▊| 105/107 [03:08<00:01, 1.09batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 99%|█████████▉| 106/107 [03:09<00:00, 1.01batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 125: 100%|██████████| 107/107 [03:17<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 51%|█████▏ | 55/107 [01:21<00:25, 2.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 126: 52%|█████▏ | 56/107 [01:59<09:53, 11.63s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 53%|█████▎ | 57/107 [02:03<07:59, 9.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 54%|█████▍ | 58/107 [02:05<05:52, 7.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 55%|█████▌ | 59/107 [02:07<04:29, 5.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 56%|█████▌ | 60/107 [02:09<03:28, 4.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 57%|█████▋ | 61/107 [02:10<02:36, 3.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 58%|█████▊ | 62/107 [02:12<02:21, 3.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 59%|█████▉ | 63/107 [02:13<01:49, 2.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 60%|█████▉ | 64/107 [02:15<01:35, 2.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 61%|██████ | 65/107 [02:16<01:22, 1.96s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 62%|██████▏ | 66/107 [02:18<01:20, 1.97s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 63%|██████▎ | 67/107 [02:20<01:13, 1.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 64%|██████▎ | 68/107 [02:21<01:01, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 64%|██████▍ | 69/107 [02:22<00:55, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 65%|██████▌ | 70/107 [02:23<00:49, 1.34s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 66%|██████▋ | 71/107 [02:24<00:52, 1.45s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 67%|██████▋ | 72/107 [02:26<00:48, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 68%|██████▊ | 73/107 [02:26<00:39, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 69%|██████▉ | 74/107 [02:28<00:39, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 70%|███████ | 75/107 [02:28<00:34, 1.08s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 71%|███████ | 76/107 [02:29<00:32, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 72%|███████▏ | 77/107 [02:31<00:36, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 73%|███████▎ | 78/107 [02:33<00:40, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 74%|███████▍ | 79/107 [02:34<00:33, 1.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 75%|███████▍ | 80/107 [02:35<00:33, 1.23s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 76%|███████▌ | 81/107 [02:36<00:28, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 77%|███████▋ | 82/107 [02:37<00:31, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 78%|███████▊ | 83/107 [02:39<00:32, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 79%|███████▊ | 84/107 [02:40<00:28, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 79%|███████▉ | 85/107 [02:41<00:24, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 80%|████████ | 86/107 [02:42<00:23, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 81%|████████▏ | 87/107 [02:43<00:21, 1.06s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 82%|████████▏ | 88/107 [02:44<00:19, 1.00s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 83%|████████▎ | 89/107 [02:45<00:18, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 84%|████████▍ | 90/107 [02:46<00:20, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 85%|████████▌ | 91/107 [02:47<00:17, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 86%|████████▌ | 92/107 [02:49<00:20, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 87%|████████▋ | 93/107 [02:50<00:17, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 88%|████████▊ | 94/107 [02:51<00:16, 1.30s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 89%|████████▉ | 95/107 [02:56<00:28, 2.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 90%|████████▉ | 96/107 [02:58<00:22, 2.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 91%|█████████ | 97/107 [03:00<00:21, 2.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 92%|█████████▏| 98/107 [03:01<00:15, 1.73s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 93%|█████████▎| 99/107 [03:02<00:11, 1.48s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 93%|█████████▎| 100/107 [03:02<00:08, 1.27s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 94%|█████████▍| 101/107 [03:03<00:06, 1.13s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 95%|█████████▌| 102/107 [03:05<00:05, 1.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 96%|█████████▋| 103/107 [03:05<00:04, 1.07s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 97%|█████████▋| 104/107 [03:06<00:02, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 98%|█████████▊| 105/107 [03:07<00:01, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 99%|█████████▉| 106/107 [03:08<00:01, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 126: 100%|██████████| 107/107 [03:15<00:00, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 51%|█████▏ | 55/107 [01:20<00:21, 2.39batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 127: 52%|█████▏ | 56/107 [01:57<09:46, 11.50s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 53%|█████▎ | 57/107 [02:00<07:26, 8.92s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 54%|█████▍ | 58/107 [02:02<05:32, 6.79s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 55%|█████▌ | 59/107 [02:07<05:03, 6.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 56%|█████▌ | 60/107 [02:09<03:53, 4.98s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 57%|█████▋ | 61/107 [02:10<02:58, 3.88s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 58%|█████▊ | 62/107 [02:13<02:32, 3.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 59%|█████▉ | 63/107 [02:14<01:58, 2.70s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 60%|█████▉ | 64/107 [02:15<01:40, 2.33s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 61%|██████ | 65/107 [02:16<01:19, 1.89s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 62%|██████▏ | 66/107 [02:18<01:15, 1.84s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 63%|██████▎ | 67/107 [02:19<01:05, 1.63s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 64%|██████▎ | 68/107 [02:20<00:54, 1.41s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 64%|██████▍ | 69/107 [02:21<00:52, 1.39s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 65%|██████▌ | 70/107 [02:23<00:56, 1.53s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 66%|██████▋ | 71/107 [02:25<01:00, 1.67s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 67%|██████▋ | 72/107 [02:26<00:53, 1.52s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 68%|██████▊ | 73/107 [02:27<00:42, 1.26s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 69%|██████▉ | 74/107 [02:28<00:43, 1.32s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 70%|███████ | 75/107 [02:29<00:36, 1.16s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 71%|███████ | 76/107 [02:30<00:32, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 72%|███████▏ | 77/107 [02:31<00:30, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 73%|███████▎ | 78/107 [02:33<00:36, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 74%|███████▍ | 79/107 [02:33<00:30, 1.10s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 75%|███████▍ | 80/107 [02:35<00:29, 1.11s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 76%|███████▌ | 81/107 [02:35<00:24, 1.04batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 77%|███████▋ | 82/107 [02:36<00:22, 1.10batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 78%|███████▊ | 83/107 [02:38<00:28, 1.17s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 79%|███████▊ | 84/107 [02:39<00:25, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 79%|███████▉ | 85/107 [02:40<00:22, 1.03s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 80%|████████ | 86/107 [02:40<00:20, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 81%|████████▏ | 87/107 [02:41<00:18, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 82%|████████▏ | 88/107 [02:42<00:16, 1.13batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 83%|████████▎ | 89/107 [02:43<00:16, 1.12batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 84%|████████▍ | 90/107 [02:44<00:16, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 85%|████████▌ | 91/107 [02:45<00:15, 1.06batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 86%|████████▌ | 92/107 [02:47<00:18, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 87%|████████▋ | 93/107 [02:48<00:15, 1.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 88%|████████▊ | 94/107 [02:49<00:15, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 89%|████████▉ | 95/107 [02:54<00:26, 2.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 90%|████████▉ | 96/107 [02:55<00:20, 1.85s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 91%|█████████ | 97/107 [02:57<00:19, 1.95s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 92%|█████████▏| 98/107 [02:58<00:14, 1.61s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 93%|█████████▎| 99/107 [02:59<00:11, 1.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 93%|█████████▎| 100/107 [02:59<00:08, 1.20s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 94%|█████████▍| 101/107 [03:00<00:06, 1.05s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 95%|█████████▌| 102/107 [03:01<00:05, 1.12s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 96%|█████████▋| 103/107 [03:02<00:04, 1.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 97%|█████████▋| 104/107 [03:03<00:02, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 98%|█████████▊| 105/107 [03:04<00:01, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 99%|█████████▉| 106/107 [03:05<00:00, 1.02batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 127: 100%|██████████| 107/107 [03:12<00:00, 1.11batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 51%|█████▏ | 55/107 [01:22<00:27, 1.91batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn] \nValidating: 0%| | 0/52 [00:00<?, ?batch/s]\u001b[A\nEpoch 128: 52%|█████▏ | 56/107 [01:58<09:40, 11.38s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 53%|█████▎ | 57/107 [02:02<07:30, 9.01s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 54%|█████▍ | 58/107 [02:03<05:28, 6.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 55%|█████▌ | 59/107 [02:06<04:26, 5.56s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 56%|█████▌ | 60/107 [02:08<03:26, 4.40s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 57%|█████▋ | 61/107 [02:09<02:37, 3.43s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 58%|█████▊ | 62/107 [02:12<02:26, 3.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 59%|█████▉ | 63/107 [02:14<02:09, 2.94s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 60%|█████▉ | 64/107 [02:16<01:57, 2.72s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 61%|██████ | 65/107 [02:18<01:39, 2.36s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 62%|██████▏ | 66/107 [02:20<01:40, 2.46s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 63%|██████▎ | 67/107 [02:22<01:28, 2.21s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 64%|██████▎ | 68/107 [02:23<01:12, 1.87s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 64%|██████▍ | 69/107 [02:36<03:19, 5.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 65%|██████▌ | 70/107 [02:38<02:34, 4.18s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 66%|██████▋ | 71/107 [02:40<02:09, 3.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 67%|██████▋ | 72/107 [02:42<01:42, 2.93s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 68%|██████▊ | 73/107 [02:42<01:16, 2.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 69%|██████▉ | 74/107 [02:43<01:04, 1.94s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 70%|███████ | 75/107 [02:44<00:50, 1.59s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 71%|███████ | 76/107 [02:45<00:41, 1.35s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 72%|███████▏ | 77/107 [02:46<00:37, 1.25s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 73%|███████▎ | 78/107 [02:48<00:41, 1.42s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 74%|███████▍ | 79/107 [02:49<00:34, 1.22s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 75%|███████▍ | 80/107 [02:50<00:32, 1.19s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 76%|███████▌ | 81/107 [02:50<00:26, 1.02s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 77%|███████▋ | 82/107 [02:51<00:23, 1.05batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 78%|███████▊ | 83/107 [02:52<00:23, 1.00batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 79%|███████▊ | 84/107 [02:53<00:21, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 79%|███████▉ | 85/107 [02:54<00:20, 1.08batch/s, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 80%|████████ | 86/107 [03:01<00:56, 2.71s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 81%|████████▏ | 87/107 [03:03<00:49, 2.49s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 82%|████████▏ | 88/107 [03:06<00:49, 2.62s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 83%|████████▎ | 89/107 [03:09<00:50, 2.83s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 84%|████████▍ | 90/107 [03:12<00:47, 2.77s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 85%|████████▌ | 91/107 [03:15<00:47, 2.96s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 86%|████████▌ | 92/107 [03:17<00:39, 2.65s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 87%|████████▋ | 93/107 [03:18<00:29, 2.09s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]\nEpoch 128: 88%|████████▊ | 94/107 [03:19<00:24, 1.91s/batch, batch_idx=54, gpu=0, loss=0.021, v_num=zf852wxn]" ] ], [ [ "# Analysis", "_____no_output_____" ] ], [ [ "from torch.utils.data import DataLoader\n\ndata_base = '/home/tshimko/tesselate/'\n\ndata_load = [\n 'pdb_id',\n# 'model',\n 'atom_nodes',\n 'atom_adj',\n# 'atom_contact',\n# 'atom_mask',\n 'res_adj',\n 'res_dist',\n 'chain_mem',\n 'res_contact',\n 'conn_adj',\n 'res_mask',\n 'mem_mat',\n# 'idx2atom_dict',\n# 'idx2res_dict'\n ]\n\ntrain_data = TesselateDataset(data_base + 'test4.txt',\n graph_dir=data_base + 'data/graphs',\n contacts_dir=data_base + 'data/contacts',\n return_data=data_load, in_memory=True)\n\n\nfor idx, i in enumerate(DataLoader(train_data, shuffle=False, num_workers=30, pin_memory=True)):\n print(idx, i['pdb_id'])", "wandb: Network error resolved after 0:00:24.093383, resuming normal operation.\nwandb: Network error resolved after 0:00:25.815422, resuming normal operation.\n" ], [ "train_data.accessions", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code", "code" ] ]

ec7e28defb0a51ac962817acca3008d4d0f97db5

732,273

ipynb

Jupyter Notebook

examples/FaceShadowRemoval/FaceShadowRemoval.ipynb

MPF-Optimization-Laboratory/AtomicOpt.jl

a03f6a0ed152bad9b518548fafa936667deb8a67

[ "MIT" ]

1

2022-02-01T01:26:04.000Z

examples/FaceShadowRemoval/FaceShadowRemoval.ipynb

ZhenanFanUBC/AtomicOpt.jl

a03f6a0ed152bad9b518548fafa936667deb8a67

[ "MIT" ]

null

examples/FaceShadowRemoval/FaceShadowRemoval.ipynb

ZhenanFanUBC/AtomicOpt.jl

a03f6a0ed152bad9b518548fafa936667deb8a67

[ "MIT" ]

null

2,153.744118

368,850

0.955865

[ [ [ "empty" ] ] ]

[ "empty" ]

[ [ "empty" ] ]

ec7e384ff10167c72ac184e3fc82079ee29abc92

16,318

ipynb

Jupyter Notebook

examples/notebooks/choropleth-viz-example.ipynb

TileDB-Inc/mapboxgl-jupyter

9ff60e8d9a86d72ee1a309557ac3d3c02db9ef18

[ "MIT" ]

617

2017-08-08T16:33:44.000Z

2022-03-18T11:10:31.000Z

examples/notebooks/choropleth-viz-example.ipynb

TileDB-Inc/mapboxgl-jupyter

9ff60e8d9a86d72ee1a309557ac3d3c02db9ef18

[ "MIT" ]

148

2017-08-09T09:16:04.000Z

2022-01-11T05:02:27.000Z

examples/notebooks/choropleth-viz-example.ipynb

TileDB-Inc/mapboxgl-jupyter

9ff60e8d9a86d72ee1a309557ac3d3c02db9ef18

[ "MIT" ]

134

2017-08-08T17:21:26.000Z

2022-01-19T16:52:24.000Z

52.980519

3,265

0.474323

[ [ [ "# Mapboxgl Python Library for location data visualization\n\nhttps://github.com/mapbox/mapboxgl-jupyter", "_____no_output_____" ] ], [ [ "import os\nfrom mapboxgl.viz import *\nfrom mapboxgl.utils import *\n\n# Must be a public token, starting with `pk`\ntoken = os.getenv('MAPBOX_ACCESS_TOKEN')", "_____no_output_____" ] ], [ [ "## Choropleths with interpolated color assignment from GeoJSON source", "_____no_output_____" ] ], [ [ "# create choropleth from polygon features stored as GeoJSON\nviz = ChoroplethViz('https://raw.githubusercontent.com/mapbox/mapboxgl-jupyter/master/examples/data/us-states.geojson', \n access_token=token,\n color_property='density',\n color_stops=create_color_stops([0, 50, 100, 500, 1500], colors='YlOrRd'),\n color_function_type='interpolate',\n line_stroke='--',\n line_color='rgb(128,0,38)',\n line_width=1,\n line_opacity=0.9,\n opacity=0.8,\n center=(-96, 37.8),\n zoom=3,\n below_layer='waterway-label',\n legend_layout='horizontal',\n legend_key_shape='bar',\n legend_key_borders_on=False)\nviz.show()", "_____no_output_____" ] ], [ [ "## Add 3-D Extrusion", "_____no_output_____" ] ], [ [ "# adjust view angle\nviz.bearing = -15\nviz.pitch = 45\n\n# add extrusion to viz using interpolation keyed on density in GeoJSON features\nviz.height_property = 'density'\nviz.height_stops = create_numeric_stops([0, 50, 100, 500, 1500, 5000], 0, 500000)\nviz.height_function_type = 'interpolate'\n\n# render again\nviz.show()", "_____no_output_____" ] ], [ [ "## Choropleths with match-type color scheme from GeoJSON source", "_____no_output_____" ] ], [ [ "match_color_stops = [['Massachusetts', 'rgb(46,204,113)'],\n ['Utah', 'rgb(231,76,60)'],\n ['California', 'rgb(142,68,173)']]\n\nviz = ChoroplethViz('https://raw.githubusercontent.com/mapbox/mapboxgl-jupyter/master/examples/data/us-states.geojson', \n access_token=token,\n color_property='name', \n color_stops=match_color_stops, \n color_function_type='match', \n color_default='rgba(52,73,94,0.5)', \n opacity=0.8, \n center=(-96, 37.8), \n zoom=3, \n below_layer='waterway-label')\nviz.show()", "_____no_output_____" ] ], [ [ "## Vector polygon source with data-join technique\n\nIn this configuration, properties in JSON data are used to calculate colors to style polygons from the vector source.", "_____no_output_____" ] ], [ [ "# must be JSON object (need to extend to use referenced JSON file)\ndata = [{\"id\": \"01\", \"name\": \"Alabama\", \"density\": 94.65}, {\"id\": \"02\", \"name\": \"Alaska\", \"density\": 1.264}, {\"id\": \"04\", \"name\": \"Arizona\", \"density\": 57.05}, {\"id\": \"05\", \"name\": \"Arkansas\", \"density\": 56.43}, {\"id\": \"06\", \"name\": \"California\", \"density\": 241.7}, {\"id\": \"08\", \"name\": \"Colorado\", \"density\": 49.33}, {\"id\": \"09\", \"name\": \"Connecticut\", \"density\": 739.1}, {\"id\": \"10\", \"name\": \"Delaware\", \"density\": 464.3}, {\"id\": \"11\", \"name\": \"District of Columbia\", \"density\": 10065}, {\"id\": \"12\", \"name\": \"Florida\", \"density\": 353.4}, {\"id\": \"13\", \"name\": \"Georgia\", \"density\": 169.5}, {\"id\": \"15\", \"name\": \"Hawaii\", \"density\": 214.1}, {\"id\": \"16\", \"name\": \"Idaho\", \"density\": 19.15}, {\"id\": \"17\", \"name\": \"Illinois\", \"density\": 231.5}, {\"id\": \"18\", \"name\": \"Indiana\", \"density\": 181.7}, {\"id\": \"19\", \"name\": \"Iowa\", \"density\": 54.81}, {\"id\": \"20\", \"name\": \"Kansas\", \"density\": 35.09}, {\"id\": \"21\", \"name\": \"Kentucky\", \"density\": 110}, {\"id\": \"22\", \"name\": \"Louisiana\", \"density\": 105}, {\"id\": \"23\", \"name\": \"Maine\", \"density\": 43.04}, {\"id\": \"24\", \"name\": \"Maryland\", \"density\": 596.3}, {\"id\": \"25\", \"name\": \"Massachusetts\", \"density\": 840.2}, {\"id\": \"26\", \"name\": \"Michigan\", \"density\": 173.9}, {\"id\": \"27\", \"name\": \"Minnesota\", \"density\": 67.14}, {\"id\": \"28\", \"name\": \"Mississippi\", \"density\": 63.5}, {\"id\": \"29\", \"name\": \"Missouri\", \"density\": 87.26}, {\"id\": \"30\", \"name\": \"Montana\", \"density\": 6.858}, {\"id\": \"31\", \"name\": \"Nebraska\", \"density\": 23.97}, {\"id\": \"32\", \"name\": \"Nevada\", \"density\": 24.8}, {\"id\": \"33\", \"name\": \"New Hampshire\", \"density\": 147}, {\"id\": \"34\", \"name\": \"New Jersey\", \"density\": 1189}, {\"id\": \"35\", \"name\": \"New Mexico\", \"density\": 17.16}, {\"id\": \"36\", \"name\": \"New York\", \"density\": 412.3}, {\"id\": \"37\", \"name\": \"North Carolina\", \"density\": 198.2}, {\"id\": \"38\", \"name\": \"North Dakota\", \"density\": 9.916}, {\"id\": \"39\", \"name\": \"Ohio\", \"density\": 281.9}, {\"id\": \"40\", \"name\": \"Oklahoma\", \"density\": 55.22}, {\"id\": \"41\", \"name\": \"Oregon\", \"density\": 40.33}, {\"id\": \"42\", \"name\": \"Pennsylvania\", \"density\": 284.3}, {\"id\": \"44\", \"name\": \"Rhode Island\", \"density\": 1006}, {\"id\": \"45\", \"name\": \"South Carolina\", \"density\": 155.4}, {\"id\": \"46\", \"name\": \"South Dakota\", \"density\": 98.07}, {\"id\": \"47\", \"name\": \"Tennessee\", \"density\": 88.08}, {\"id\": \"48\", \"name\": \"Texas\", \"density\": 98.07}, {\"id\": \"49\", \"name\": \"Utah\", \"density\": 34.3}, {\"id\": \"50\", \"name\": \"Vermont\", \"density\": 67.73}, {\"id\": \"51\", \"name\": \"Virginia\", \"density\": 204.5}, {\"id\": \"53\", \"name\": \"Washington\", \"density\": 102.6}, {\"id\": \"54\", \"name\": \"West Virginia\", \"density\": 77.06}, {\"id\": \"55\", \"name\": \"Wisconsin\", \"density\": 105.2}, {\"id\": \"56\", \"name\": \"Wyoming\", \"density\": 5.851}, {\"id\": \"72\", \"name\": \"Puerto Rico\", \"density\": 1082}]\n\n# create choropleth map with vector source styling use data in JSON object\nviz = ChoroplethViz(data, \n access_token=token,\n vector_url='mapbox://mapbox.us_census_states_2015',\n vector_layer_name='states',\n vector_join_property='STATE_ID',\n data_join_property='id',\n color_property='density',\n color_stops=create_color_stops([0, 50, 100, 500, 1500], colors='YlOrRd'),\n line_stroke='dashed',\n line_color='rgb(128,0,38)',\n opacity=0.8,\n center=(-96, 37.8),\n zoom=3,\n below_layer='waterway-label',\n legend_layout='horizontal',\n legend_key_shape='bar',\n legend_key_borders_on=False)\nviz.show()", "_____no_output_____" ] ], [ [ "## Vector polygon source with data-join technique, categorical color scheme", "_____no_output_____" ] ], [ [ "# must be JSON object (need to extend to use referenced JSON file)\ndata = [{\"id\": \"01\", \"name\": \"Alabama\", \"density\": 94.65}, {\"id\": \"02\", \"name\": \"Alaska\", \"density\": 1.264}, {\"id\": \"04\", \"name\": \"Arizona\", \"density\": 57.05}, {\"id\": \"05\", \"name\": \"Arkansas\", \"density\": 56.43}, {\"id\": \"06\", \"name\": \"California\", \"density\": 241.7}, {\"id\": \"08\", \"name\": \"Colorado\", \"density\": 49.33}, {\"id\": \"09\", \"name\": \"Connecticut\", \"density\": 739.1}, {\"id\": \"10\", \"name\": \"Delaware\", \"density\": 464.3}, {\"id\": \"11\", \"name\": \"District of Columbia\", \"density\": 10065}, {\"id\": \"12\", \"name\": \"Florida\", \"density\": 353.4}, {\"id\": \"13\", \"name\": \"Georgia\", \"density\": 169.5}, {\"id\": \"15\", \"name\": \"Hawaii\", \"density\": 214.1}, {\"id\": \"16\", \"name\": \"Idaho\", \"density\": 19.15}, {\"id\": \"17\", \"name\": \"Illinois\", \"density\": 231.5}, {\"id\": \"18\", \"name\": \"Indiana\", \"density\": 181.7}, {\"id\": \"19\", \"name\": \"Iowa\", \"density\": 54.81}, {\"id\": \"20\", \"name\": \"Kansas\", \"density\": 35.09}, {\"id\": \"21\", \"name\": \"Kentucky\", \"density\": 110}, {\"id\": \"22\", \"name\": \"Louisiana\", \"density\": 105}, {\"id\": \"23\", \"name\": \"Maine\", \"density\": 43.04}, {\"id\": \"24\", \"name\": \"Maryland\", \"density\": 596.3}, {\"id\": \"25\", \"name\": \"Massachusetts\", \"density\": 840.2}, {\"id\": \"26\", \"name\": \"Michigan\", \"density\": 173.9}, {\"id\": \"27\", \"name\": \"Minnesota\", \"density\": 67.14}, {\"id\": \"28\", \"name\": \"Mississippi\", \"density\": 63.5}, {\"id\": \"29\", \"name\": \"Missouri\", \"density\": 87.26}, {\"id\": \"30\", \"name\": \"Montana\", \"density\": 6.858}, {\"id\": \"31\", \"name\": \"Nebraska\", \"density\": 23.97}, {\"id\": \"32\", \"name\": \"Nevada\", \"density\": 24.8}, {\"id\": \"33\", \"name\": \"New Hampshire\", \"density\": 147}, {\"id\": \"34\", \"name\": \"New Jersey\", \"density\": 1189}, {\"id\": \"35\", \"name\": \"New Mexico\", \"density\": 17.16}, {\"id\": \"36\", \"name\": \"New York\", \"density\": 412.3}, {\"id\": \"37\", \"name\": \"North Carolina\", \"density\": 198.2}, {\"id\": \"38\", \"name\": \"North Dakota\", \"density\": 9.916}, {\"id\": \"39\", \"name\": \"Ohio\", \"density\": 281.9}, {\"id\": \"40\", \"name\": \"Oklahoma\", \"density\": 55.22}, {\"id\": \"41\", \"name\": \"Oregon\", \"density\": 40.33}, {\"id\": \"42\", \"name\": \"Pennsylvania\", \"density\": 284.3}, {\"id\": \"44\", \"name\": \"Rhode Island\", \"density\": 1006}, {\"id\": \"45\", \"name\": \"South Carolina\", \"density\": 155.4}, {\"id\": \"46\", \"name\": \"South Dakota\", \"density\": 98.07}, {\"id\": \"47\", \"name\": \"Tennessee\", \"density\": 88.08}, {\"id\": \"48\", \"name\": \"Texas\", \"density\": 98.07}, {\"id\": \"49\", \"name\": \"Utah\", \"density\": 34.3}, {\"id\": \"50\", \"name\": \"Vermont\", \"density\": 67.73}, {\"id\": \"51\", \"name\": \"Virginia\", \"density\": 204.5}, {\"id\": \"53\", \"name\": \"Washington\", \"density\": 102.6}, {\"id\": \"54\", \"name\": \"West Virginia\", \"density\": 77.06}, {\"id\": \"55\", \"name\": \"Wisconsin\", \"density\": 105.2}, {\"id\": \"56\", \"name\": \"Wyoming\", \"density\": 5.851}, {\"id\": \"72\", \"name\": \"Puerto Rico\", \"density\": 1082}]\n\nmatch_color_stops = [['Massachusetts', 'rgb(46,204,113)'],\n ['Utah', 'rgb(231,76,60)'],\n ['California', 'rgb(142,68,173)']]\n\n# create choropleth map with vector source styling use data in JSON object\nviz = ChoroplethViz(data, \n access_token=token,\n vector_url='mapbox://mapbox.us_census_states_2015',\n vector_layer_name='states',\n vector_join_property='STATE_ID',\n data_join_property='id',\n color_property='name',\n color_stops=match_color_stops,\n color_default = 'rgba(52,73,94,0.5)', \n opacity=0.8,\n center=(-96, 37.8),\n zoom=3,\n below_layer='waterway-label')\nviz.show()", "_____no_output_____" ] ], [ [ "## Add 3-D Extrusion to Vector Choropleth Map", "_____no_output_____" ] ], [ [ "# adjust view angle\nviz.bearing = -15\nviz.pitch = 45\n\n# add extrusion to viz using interpolation keyed on density in GeoJSON features\nviz.height_property = 'density'\nviz.height_stops = create_numeric_stops([0, 50, 100, 500, 1500, 5000], 0, 500000)\nviz.height_function_type = 'interpolate'\n\n# render again\nviz.show()", "_____no_output_____" ] ], [ [ "# Vector polygon source joined to data in a local Pandas dataframe", "_____no_output_____" ] ], [ [ "# Load data from sample csv\nimport pandas as pd\ndata_url = 'https://raw.githubusercontent.com/mapbox/mapboxgl-jupyter/master/examples/data/2010_us_population_by_postcode.csv'\ndf = pd.read_csv(data_url).round(3)\ndf.head(2)", "_____no_output_____" ], [ "# Group pandas dataframe by a value\nmeasure = '2010 Census Population'\ndimension = 'Zip Code ZCTA'\n\ndata = df[[dimension, measure]].groupby(dimension, as_index=False).mean()\ncolor_breaks = [round(data[measure].quantile(q=x*0.1), 2) for x in range(2,11)]\ncolor_stops = create_color_stops(color_breaks, colors='PuRd')\ndata = json.loads(data.to_json(orient='records'))", "_____no_output_____" ], [ "# Create the viz\nviz = ChoroplethViz(data, \n access_token=token,\n vector_url='mapbox://rsbaumann.bv2k1pl2',\n vector_layer_name='2016_us_census_postcode',\n vector_join_property='postcode',\n data_join_property=dimension,\n color_property=measure,\n color_stops=color_stops,\n line_color = 'rgba(0,0,0,0.05)',\n line_width = 0.5,\n opacity=0.7,\n center=(-95, 45),\n zoom=2,\n below_layer='waterway-label',\n legend_key_shape='contiguous-bar')\nviz.show()", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code", "code", "code" ] ]

ec7e38b5565e2fec9b20f2cb0f5d61c93a53bdb7

18,485

ipynb

Jupyter Notebook

Lecture05_BayesianRegression/Lecture 05 - Conjugate Priors and Bayesian Linear Regression.ipynb

sawahashi3/LectureNotes

16a6bd2913cc3fd241032566ad46ff6084e4b235

[ "MIT" ]

2

2020-03-17T19:11:19.000Z

2021-02-16T06:04:02.000Z

Lecture05_BayesianRegression/Lecture 05 - Conjugate Priors and Bayesian Linear Regression.ipynb

sawahashi3/LectureNotes

16a6bd2913cc3fd241032566ad46ff6084e4b235

[ "MIT" ]

2

2018-10-02T16:11:44.000Z

2018-10-23T17:05:00.000Z

Lecture05_BayesianRegression/Lecture 05 - Conjugate Priors and Bayesian Linear Regression.ipynb

sawahashi3/LectureNotes

16a6bd2913cc3fd241032566ad46ff6084e4b235

[ "MIT" ]

2

2020-03-18T00:29:54.000Z

2020-07-23T05:00:04.000Z

52.365439

373

0.516094

[ [ [ "# Sample Mean vs. Mode vs. Expected Value", "_____no_output_____" ] ], [ [ "import numpy as np\nimport matplotlib.pyplot as plt\nfrom scipy.stats import gamma\nfrom scipy.stats import multivariate_normal\nimport textwrap\nimport math\n%matplotlib inline \n\ndef sampleMeanEx():\n\t'''sampleMeanEx()'''\n\tnSamples = (5, 10, 100, 5000)\n\ta = 3\n\tb = 1\n\tfor i in range(len(nSamples)):\n\t\tfig = plt.figure()\n\t\tax = fig.add_subplot(*[1,1,1])\n\t\tdraws = np.random.gamma(shape=a,scale=b,size=nSamples[i])\n\t\tmode = (a-1)*b\n\t\texpectedv = a*b\n\t\tsamplemean = sum(draws)/len(draws)\n\t\tx = np.linspace(gamma.ppf(0.001, a, scale=b), gamma.ppf(0.999, a, scale=b), 100)\n\t\tax.plot(x, gamma.pdf(x, a, scale=b), 'r-', lw=5)\n\t\tax.scatter(draws, np.zeros(len(draws)), c='k')\n\t\tmyTitle = 'Num Points: ' + str(nSamples[i]) + ' Mode: ' + str(\"%.2f\"%mode) + ' E[x]: ' + str(\"%.2f\"%expectedv) + ' Sample Mean: ' + str(\"%.2f\"%samplemean)\n\t\tax.set_title(\"\\n\".join(textwrap.wrap(myTitle, 100)))\n\t\tplt.show()\n\nsampleMeanEx()", "_____no_output_____" ] ], [ [ "# Conjugate Priors\n\n* Last class we mentioned the concept of *conjugate priors*\n* Two distributions have a conjugate prior relationship when the form of the posterior is the same as the form of the prior. \n* For example, a Gaussian distribution is a conjugate prior for the mean of a Gaussian as shown in the following: ", "_____no_output_____" ], [ "\\begin{eqnarray}\np(\\mu |\\mathbf{X}) &\\propto& p(\\mathbf{X}|\\mu)p(\\mu)\\\\\n&=& \\prod_{i=1}^N\\mathscr{N}(x_i|\\mu, \\sigma^2)\\mathscr{N}(\\mu|\\mu_0, \\sigma_0^2)\\\\\n&=& \\prod_{i=1}^N \\frac{1}{\\sqrt{2\\pi\\sigma^2}}\\exp\\left\\{-\\frac{1}{2}\\frac{(x_i - \\mu)^2}{\\sigma^2} \\right\\}\\frac{1}{\\sqrt{2\\pi\\sigma_0^2}}\\exp\\left\\{-\\frac{1}{2}\\frac{(\\mu - \\mu_0)^2}{\\sigma_0^2} \\right\\}\\\\\n&=& \\frac{1}{\\sqrt{2\\pi\\sigma^2}}\\frac{1}{\\sqrt{2\\pi\\sigma_0^2}} \\exp\\left\\{ \\sum_{i=1}^N \\left(-\\frac{1}{2}\\frac{(x_i - \\mu)^2}{\\sigma^2}\\right) -\\frac{1}{2}\\frac{(\\mu - \\mu_0)^2}{\\sigma_0^2} \\right\\}\\\\\n&=& \\frac{1}{\\sqrt{2\\pi\\sigma^2}\\sqrt{2\\pi\\sigma_0^2}} \\exp\\left\\{ -\\frac{1}{2}\\left( \\sum_{i=1}^N\\frac{(x_i - \\mu)^2}{\\sigma^2} + \\frac{(\\mu - \\mu_0)^2}{\\sigma_0^2} \\right) \\right\\}\\\\\n&=& \\frac{1}{\\sqrt{2\\pi\\sigma^2}\\sqrt{2\\pi\\sigma_0^2}} \\exp\\left\\{ -\\frac{1}{2}\\left( \\frac{\\sum_{i=1}^Nx_i^2 - 2\\sum_{i=1}^Nx_i\\mu + \\mu^2N}{\\sigma^2} + \\frac{\\mu^2 - 2\\mu\\mu_0 + \\mu_0^2}{\\sigma_0^2} \\right) \\right\\}\\\\\n&=& \\frac{1}{\\sqrt{2\\pi\\sigma^2}\\sqrt{2\\pi\\sigma_0^2}} \\exp\\left\\{ -\\frac{1}{2}\\left( \\mu^2\\left( \\frac{N}{\\sigma^2} + \\frac{1}{\\sigma_0^2}\\right) - 2\\mu\\left( \\frac{\\sum_{i=1}^Nx_i}{\\sigma^2} + \\frac{\\mu_0}{\\sigma_0^2} \\right) + \\frac{\\sum_{i=1}^Nx_i^2 }{\\sigma^2} + \\frac{ \\mu_0^2}{\\sigma_0^2} \\right) \\right\\}\\\\\n&=& \\frac{1}{\\sqrt{2\\pi\\sigma^2}\\sqrt{2\\pi\\sigma_0^2}} \\exp\\left\\{ -\\frac{1}{2}\\left( \\frac{N}{\\sigma^2} + \\frac{1}{\\sigma_0^2}\\right)\\left( \\mu^2 - 2\\mu\\left( \\frac{\\sum_{i=1}^Nx_i}{\\sigma^2} + \\frac{\\mu_0}{\\sigma_0^2} \\right)\\left( \\frac{N}{\\sigma^2} + \\frac{1}{\\sigma_0^2}\\right)^{-1} \\right) \\right\\}\\\\\n& & \\exp\\left\\{ \\frac{\\sum_{i=1}^Nx_i^2 }{\\sigma^2} + \\frac{ \\mu_0^2}{\\sigma_0^2} \\right\\} \\nonumber\\\\\n&=& \\frac{1}{\\sqrt{2\\pi\\sigma^2}\\sqrt{2\\pi\\sigma_0^2}} \\exp\\left\\{ -\\frac{1}{2}\\left( \\frac{N}{\\sigma^2} + \\frac{1}{\\sigma_0^2}\\right)\\left( \\mu - \\left( \\frac{\\sum_{i=1}^Nx_i\\sigma_0^2 + \\mu_0\\sigma^2}{\\sigma^2\\sigma_0^2} \\right)\\left( \\frac{N\\sigma_0^2 + \\sigma^2}{\\sigma^2\\sigma_0^2}\\right)^{-1} \\right)^2 \\right.\\\\\n&+& \\left. \\frac{1}{2}\\left( \\frac{N}{\\sigma^2} + \\frac{1}{\\sigma_0^2}\\right)\n\\left(\n\\left( \\frac{\\sum_{i=1}^Nx_i\\sigma_0^2 + \\mu_0\\sigma^2}{\\sigma^2\\sigma_0^2} \\right)\\left( \\frac{N\\sigma_0^2 + \\sigma^2}{\\sigma^2\\sigma_0^2}\\right)^{-1} \\right)^2 \\right\\} \n\\exp\\left\\{ \\frac{\\sum_{i=1}^Nx_i^2 }{\\sigma^2} + \\frac{ \\mu_0^2}{\\sigma_0^2} \\right\\} \\nonumber\\\\\n%\n&=& \\frac{1}{\\sqrt{2\\pi\\sigma^2}\\sqrt{2\\pi\\sigma_0^2}} \\exp\\left\\{ -\\frac{1}{2}\\left( \\frac{N}{\\sigma^2} + \\frac{1}{\\sigma_0^2}\\right)\\left( \\mu - \\frac{\\sum_{i=1}^Nx_i\\sigma_0^2 + \\mu_0\\sigma^2}{N\\sigma_0^2 + \\sigma^2 } \\right)^2 \\right\\} \\\\\n& &\\exp\\left\\{ \\frac{1}{2}\\left( \\frac{N}{\\sigma^2} + \\frac{1}{\\sigma_0^2}\\right)\n\\left(\\frac{\\sum_{i=1}^Nx_i\\sigma_0^2 + \\mu_0\\sigma^2}{N\\sigma_0^2 + \\sigma^2 }\\right)^2 + \\frac{\\sum_{i=1}^Nx_i^2 }{\\sigma^2} + \\frac{ \\mu_0^2}{\\sigma_0^2} \\right\\} \\nonumber\\\\\n%\n&=& C \\exp\\left\\{ -\\frac{1}{2}\\left( \\frac{N}{\\sigma^2} + \\frac{1}{\\sigma_0^2}\\right)\\left( \\mu - \\frac{\\sum_{i=1}^Nx_i\\sigma_0^2 + \\mu_0\\sigma^2}{N\\sigma_0^2 + \\sigma^2 } \\right)^2 \\right\\} \\\\\n&\\propto& \\mathscr{N}\\left(\\mu\\left|\\frac{\\sum_{i=1}^N x_i\\sigma_0^2 + \\mu_0\\sigma^2}{N\\sigma_0^2 + \\sigma^2 }, \\left( \\frac{N}{\\sigma^2} + \\frac{1}{\\sigma_0^2}\\right)^{-1}\\right.\\right)\n\\end{eqnarray}", "_____no_output_____" ], [ "* So, as shown above, the form of the posterior is also a Gaussian distribution. \n\n* There are many conjugate prior relationships, e.g., Bernoulli-Beta, Gaussian-Gaussian, Gaussian-InverseWishart, Multinomial-Dirichlet\n", "_____no_output_____" ], [ "# Bayesian Regression\n\n* Look back our polynomial regression: \n\\begin{equation}\n\\min E^{\\ast}(\\mathbf{w}) = \\frac{1}{2}\\sum_{n=1}^N\\left( y(x_n, \\mathbf{w}) - t_n \\right)^2 + \\frac{\\lambda}{2}\\left\\| \\mathbf{w} \\right\\|_2^2\n\\end{equation}\nThis is equivalent to: \n\\begin{equation}\n\\max \\prod_{n=1}^N \\exp\\left\\{-\\frac{1}{2}\\left( y(x_n, \\mathbf{w}) - t_n \\right)^2 \\right\\}\\exp\\left\\{- \\frac{\\lambda}{2}\\left\\| \\mathbf{w} \\right\\|_2^2 \\right\\}\n\\end{equation}\n* As discussed, the first term is the likelihood and the second term is the prior on the weights\n* These are Gaussian distributions:\n\\begin{equation}\n\\mathscr{N}(x|\\mu, \\sigma^2) = \\frac{1}{\\sqrt{2\\pi \\sigma^2}}\\exp\\left\\{ -\\frac{1}{2}\\frac{(x-\\mu)^2}{\\sigma^2} \\right\\}\n\\end{equation}\n* $\\sigma^2$ is the variance OR $\\frac{1}{\\sigma^2}$ is the *precision*\n* So, as $\\lambda$ gets big, variance gets smaller/tighter. As $\\lambda$ gets small, variance gets larger/wider. \n", "_____no_output_____" ], [ "* Previously, we used: \n\\begin{equation}\ny = \\sum_{j=0}^M w_j x^j\n\\end{equation}\n* We can extend this, to make it more general and flexible: \n\\begin{equation}\ny = \\sum_{j=0}^M w_j \\phi_j(\\mathbf{x})\n\\end{equation}\nwhere $\\phi_j(\\mathbf{x})$ is a *basis function*\n* For example:\n * Basis function we were using previously: $\\phi_j(x) = x^j$ (for univariate $x$)\n * Linear Basis Function: $\\phi_j(\\mathbf{x}) = x_j$\n * Radial Basis Function: $\\phi_j(\\mathbf{x}) = \\exp\\left\\{ - \\frac{(x - \\mu_j)^2}{2s_j^2}\\right\\}$\n * Sigmoidal Basis Function: $\\phi_j(\\mathbf{x}) = \\frac{1}{1 + \\exp \\left\\{ \\frac{\\mathbf{x} - \\mu_j}{s}\\right\\}}$\n\n* As before:\n\\begin{equation}\nt = y(\\mathbf{x},\\mathbf{w}) + \\epsilon\n\\end{equation}\n* However, now: \n\\begin{equation}\ny = \\mathbf{w}^T\\boldsymbol{\\Phi}(\\mathbf{x}) = [w_0, w_1, \\ldots, w_M][\\phi_0(\\mathbf{x}), \\phi_1(\\mathbf{x}), \\ldots, \\phi_M(\\mathbf{x})]^T\n\\end{equation}\nwhere $\\epsilon \\sim \\mathscr{N}(\\cdot|0, \\beta^{-1})$\n", "_____no_output_____" ], [ "\\begin{equation}\np(t | \\mathbf{w}, \\beta) = \\prod_{n=1}^N \\mathscr{N}(t_n | \\mathbf{w}^T\\boldsymbol{\\Phi}(\\mathbf{x}_n), \\beta^{-1})\n\\end{equation}\n* So, what is the ``trick'' to use to maximize this? \n\\begin{eqnarray}\n\\mathscr{L} = \\frac{N}{2}\\ln\\beta - \\frac{N}{2}\\ln(2\\pi) - \\beta E(\\mathbf{w})\\\\\n\\frac{\\partial \\mathscr{L}}{\\partial \\mathbf{w}} = \\beta \\sum_{n=1}^N(t_n - \\mathbf{w}^T\\boldsymbol{\\Phi}(\\mathbf{x}_n))\\boldsymbol{\\Phi}(\\mathbf{x}_n)^T = 0 \n\\end{eqnarray}\n* This results in:\n\\begin{equation}\n\\mathbf{w}_{ML} = \\left(\\boldsymbol{\\Phi}^T\\boldsymbol{\\Phi} \\right)^{-1}\\boldsymbol{\\Phi}^T\\mathbf{t}\n\\end{equation}\nwhere \n\\begin{equation}\n\\boldsymbol{\\Phi} = \\left[ \\boldsymbol{\\Phi}(x_1), \\boldsymbol{\\Phi}(x_2), ...\\right]\n\\end{equation}\n", "_____no_output_____" ], [ "* What would you do if you want to include a prior? get the MAP solution? If assuming zero-mean Gaussian noise, then Regularized Least Squares!\n\n### Bayesian Linear Regression\n\n* Recall: $E_D(\\mathbf{w}) + \\lambda E_W(\\mathbf{w})$ where $\\lambda$ is the trade-off regularization parameter\n* A simple regularizer (and the one we used previously) is: $E_W(\\mathbf{w}) = \\frac{1}{2}\\mathbf{w}^T\\mathbf{w}$\n* If we assume zero-mean Gaussian noise: $E_D(\\mathbf{w}) = \\frac{1}{2}\\sum_{n=1}^N \\left\\{ t_n - \\mathbf{w}^T\\boldsymbol{\\phi}(\\mathbf{x}_n) \\right\\}^2$\n* Then, the total error becomes: $\\frac{1}{2}\\sum_{n=1}^N \\left\\{ t_n - \\mathbf{w}^T\\boldsymbol{\\phi}(\\mathbf{x}_n) \\right\\}^2 + \\frac{\\lambda}{2}\\mathbf{w}^T\\mathbf{w}$\n* We can take the derivative, set it equal to zero and solve for the weights. When we do, we get:\n\\begin{equation}\n\\mathbf{w} = \\left( \\lambda \\mathbf{I} + \\boldsymbol{\\Phi}^T\\boldsymbol{\\Phi}\\right)^{-1}\\boldsymbol{\\Phi}^T\\mathbf{t}\n\\end{equation}\n* Recall, we can interpret this as:\n\\begin{eqnarray}\n\\min_{\\mathbf{w}} E^{\\ast} &=& \\min_{\\mathbf{w}}\\left\\{ E_D(\\mathbf{w}) + \\lambda E_W(\\mathbf{w}) \\right\\}\\\\\n &=& \\max_{\\mathbf{w}}\\left\\{ -E_D(\\mathbf{w}) - \\lambda E_W(\\mathbf{w}) \\right\\}\\\\\n &=& \\max_{\\mathbf{w}}\\exp\\left\\{ -E_D(\\mathbf{w}) - \\lambda E_W(\\mathbf{w}) \\right\\}\\\\\n &=& \\max_{\\mathbf{w}}\\exp\\left\\{ -E_D(\\mathbf{w})\\right\\} \\exp\\left\\{- \\lambda E_W(\\mathbf{w}) \\right\\}\\\\\n &\\propto& \\max_{\\mathbf{w}} \\prod_{n=1}^N \\mathscr{N}\\left(t\\left|\\mathbf{w}^T\\mathbf{\\Phi}(\\mathbf{x}_n), \\beta\\mathbf{I}\\right.\\right) \\mathscr{N}\\left(\\mathbf{w}\\left|\\mathbf{m}_0, \\mathbf{S}_0\\right.\\right)\\\\\n &=& \\max_{\\mathbf{w}} p\\left(\\mathbf{t}\\left|\\mathbf{w}, \\mathbf{X}\\right.\\right)p\\left(\\mathbf{w}\\right)\\\\\n &\\propto& \\max_{\\mathbf{w}} p(\\mathbf{w}|\\mathbf{m}_N, \\mathbf{S}_N)= \\mathscr{N}(\\mathbf{w}|\\mathbf{m}_N, \\mathbf{S}_N)\n\\end{eqnarray}\nwhere $\\mathbf{m}_N = \\mathbf{S}_N\\left(\\mathbf{S}_0\\mathbf{m}_0 + \\beta\\boldsymbol{\\Phi}^T\\mathbf{t} \\right)$ and $\\mathbf{S}_N^{-1} = \\mathbf{S}_0^{-1} + \\beta\\boldsymbol{\\Phi}^T\\boldsymbol{\\Phi}$\n\n* What happens with different values of $\\beta$ and $\\mathbf{S}_0$? \n\n\n* To simplify, let us assume that $\\mathbf{S}_0 = \\alpha^{-1}\\mathbf{I}$ and $\\mathbf{m}_0 = \\mathbf{0}$, thus, $\\mathbf{m}_N = \\beta\\mathbf{S}_N\\boldsymbol{\\Phi}^T\\mathbf{t}$ and $\\mathbf{S}_N^{-1} = (\\alpha^{-1}\\mathbf{I})^{-1} + \\beta\\boldsymbol{\\Phi}^T\\boldsymbol{\\Phi} = \\alpha\\mathbf{I} + \\beta\\boldsymbol{\\Phi}^T\\boldsymbol{\\Phi}$\n* This results in the following Log Posterior: \n\\begin{eqnarray}\n\\ln p(\\mathbf{w}|\\mathbf{t}) = - \\frac{\\beta}{2}\\sum_{n=1}^N\\left( t_n - \\mathbf{w}^T\\mathbf{\\Phi}(\\mathbf{x}_n)\\right)^2 - \\frac{\\alpha}{2}\\mathbf{w}^T\\mathbf{w} + const\n\\end{eqnarray}\n\n* Let us suppose we are dealing with 1-D data, $\\mathbf{X} = \\left\\{ x_1, \\ldots, x_N \\right\\}$ and a linear form for y: $y(x, \\mathbf{w}) = w_0 + w_1x$\n\n* We are going to generate synthetic data from: $t = -0.3 + 0.5x + \\epsilon$ where $\\epsilon$ is from zero-mean Gaussian noise. The goal is to estimate the true values $w_0 = -0.3$ and $w_1 = 0.5$. \n\n", "_____no_output_____" ] ], [ [ "def innerBasisFunc(dataX):\n\treturn np.vstack([dataX[i,:]@dataX[i,:].T for i in range(dataX.shape[0])])\n\ndef prodBasisFunc(dataX):\n\treturn np.vstack([dataX[i,0]*dataX[i,1] for i in range(dataX.shape[0])])\n\ndef xorExample():\n\t'''xorExample()'''\n\tclass1X = np.vstack([np.array([-1,-1])+np.random.normal(0,.1,2 ) for i in range(100)])\n\tclass1X = np.vstack((class1X,np.vstack([np.array([1,1])+np.random.normal(0,.1,2 ) for i in range(100)])))\n\tclass2X = np.vstack([np.array([1,-1])+np.random.normal(0,.1,2 ) for i in range(100)])\n\tclass2X = np.vstack((class2X,np.vstack([np.array([-1,1])+np.random.normal(0,.1,2 ) for i in range(100)])))\n\tphi1X = prodBasisFunc(class1X)\n\tphi2X = prodBasisFunc(class2X)\n \n\tfig = plt.figure()\n\tax = fig.add_subplot(*[1,2,1])\n\tax.scatter(class1X[:,0], class1X[:,1], c='r') \n\tax.scatter(class2X[:,0], class2X[:,1]) \n\tax = fig.add_subplot(*[1,2,2])\n\tax.scatter(phi1X, np.zeros(phi1X.shape)+0.001, c='r') \n\tax.scatter(phi2X, np.zeros(phi2X.shape)) \n\tplt.show()\n \nxorExample()", "_____no_output_____" ], [ "def likelihood_prior_func(): \n fig = plt.figure(figsize=(18, 16), dpi= 80, facecolor='w', edgecolor='k')\n\n #set up variables\n a = -0.3\n b = 0.5\n rangeX = [-1, 1]\n step = 0.025\n X = np.mgrid[rangeX[0]:rangeX[1]:step]\n alpha = 30\n beta = 2\n S0 = (1/alpha)*np.eye(2)\n draw_num = (0,1,2,3,20)\n\n #initialize prior/posterior and sample data\n sigma = S0\n mean = [0,0]\n draws = np.random.uniform(rangeX[0],rangeX[1],size=draw_num[-1])\n T = a + b*draws + np.random.normal(loc=0, scale=math.sqrt(1/beta))\n\n for i in range(len(draw_num)):\n if draw_num[i]>0: #skip first image\n #Show data likelihood\n Phi = np.vstack((np.ones(draws[0:draw_num[i]].shape), draws[0:draw_num[i]]))\n t = T[0:draw_num[i]]\n sigma = np.linalg.inv(S0 + beta*[email protected])\n mean = beta*sigma@Phi@t\n\n w0, w1 = np.mgrid[rangeX[0]:rangeX[1]:step, rangeX[0]:rangeX[1]:step]\n p = multivariate_normal(t[draw_num[i]-1], 1/beta)\n out = np.empty(w0.shape)\n for j in range(len(w0)):\n out[j] = p.pdf(w0[j]+w1[j]*draws[draw_num[i]-1])\n\n ax = fig.add_subplot(*[len(draw_num),3,(i)*3+1])\n ax.pcolor(w0, w1, out)\n ax.scatter(a,b, c='c')\n myTitle = 'data likelihood'\n ax.set_title(\"\\n\".join(textwrap.wrap(myTitle, 100)))\n\n #Show prior/posterior\n w0, w1 = np.mgrid[rangeX[0]:rangeX[1]:step, rangeX[0]:rangeX[1]:step]\n pos = np.empty(w1.shape + (2,))\n pos[:, :, 0] = w0; pos[:, :, 1] = w1\n p = multivariate_normal(mean, sigma)\n\n ax = fig.add_subplot(*[len(draw_num),3,(i)*3+2])\n ax.pcolor(w0, w1, p.pdf(pos))\n ax.scatter(a,b, c='c')\n myTitle = 'Prior/Posterior'\n ax.set_title(\"\\n\".join(textwrap.wrap(myTitle, 100)))\n\n #Show data space\n for j in range(6):\n w0, w1 = np.random.multivariate_normal(mean, sigma)\n t = w0 + w1*X\n ax = fig.add_subplot(*[len(draw_num),3,(i)*3+3])\n ax.plot(X,t)\n if draw_num[i] > 0:\n ax.scatter(Phi[1,:], T[0:draw_num[i]])\n myTitle = 'data space'\n ax.set_title(\"\\n\".join(textwrap.wrap(myTitle, 100)))\n\n plt.show()\n\nlikelihood_prior_func()\n\n", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code" ]

[ [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code", "code" ] ]

ec7e427faa9f7a6c99f058b193204a7e3eba466d

613,531

ipynb

Jupyter Notebook

DV0101EN-2-3-1-Pie-Charts-Box-Plots-Scatter-Plots-and-Bubble-Plots-py-v2.0.ipynb

divyanshchoubisa/DATA-VISUALIZATION

3709e596c54ecc65f1986a8a7a5b14669d3bcc65

[ "MIT" ]

null

DV0101EN-2-3-1-Pie-Charts-Box-Plots-Scatter-Plots-and-Bubble-Plots-py-v2.0.ipynb

divyanshchoubisa/DATA-VISUALIZATION

3709e596c54ecc65f1986a8a7a5b14669d3bcc65

[ "MIT" ]

null

DV0101EN-2-3-1-Pie-Charts-Box-Plots-Scatter-Plots-and-Bubble-Plots-py-v2.0.ipynb

divyanshchoubisa/DATA-VISUALIZATION

3709e596c54ecc65f1986a8a7a5b14669d3bcc65

[ "MIT" ]

null

137.77925

83,016

0.840411

[ [ [ "<a href=\"https://cognitiveclass.ai\"><img src = \"https://ibm.box.com/shared/static/9gegpsmnsoo25ikkbl4qzlvlyjbgxs5x.png\" width = 400> </a>\n\n<h1 align=center><font size = 5>Pie Charts, Box Plots, Scatter Plots, and Bubble Plots</font></h1>", "_____no_output_____" ], [ "## Introduction\n\nIn this lab session, we continue exploring the Matplotlib library. More specificatlly, we will learn how to create pie charts, box plots, scatter plots, and bubble charts.", "_____no_output_____" ], [ "## Table of Contents\n\n<div class=\"alert alert-block alert-info\" style=\"margin-top: 20px\">\n\n1. [Exploring Datasets with *p*andas](#0)<br>\n2. [Downloading and Prepping Data](#2)<br>\n3. [Visualizing Data using Matplotlib](#4) <br>\n4. [Pie Charts](#6) <br>\n5. [Box Plots](#8) <br>\n6. [Scatter Plots](#10) <br>\n7. [Bubble Plots](#12) <br> \n</div>\n<hr>", "_____no_output_____" ], [ "# Exploring Datasets with *pandas* and Matplotlib<a id=\"0\"></a>\n\nToolkits: The course heavily relies on [*pandas*](http://pandas.pydata.org/) and [**Numpy**](http://www.numpy.org/) for data wrangling, analysis, and visualization. The primary plotting library we will explore in the course is [Matplotlib](http://matplotlib.org/).\n\nDataset: Immigration to Canada from 1980 to 2013 - [International migration flows to and from selected countries - The 2015 revision](http://www.un.org/en/development/desa/population/migration/data/empirical2/migrationflows.shtml) from United Nation's website.\n\nThe dataset contains annual data on the flows of international migrants as recorded by the countries of destination. The data presents both inflows and outflows according to the place of birth, citizenship or place of previous / next residence both for foreigners and nationals. In this lab, we will focus on the Canadian Immigration data.", "_____no_output_____" ], [ "# Downloading and Prepping Data <a id=\"2\"></a>", "_____no_output_____" ], [ "Import primary modules.", "_____no_output_____" ] ], [ [ "import numpy as np # useful for many scientific computing in Python\nimport pandas as pd # primary data structure library", "_____no_output_____" ] ], [ [ "Let's download and import our primary Canadian Immigration dataset using *pandas* `read_excel()` method. Normally, before we can do that, we would need to download a module which *pandas* requires to read in excel files. This module is **xlrd**. For your convenience, we have pre-installed this module, so you would not have to worry about that. Otherwise, you would need to run the following line of code to install the **xlrd** module:\n```\n!conda install -c anaconda xlrd --yes\n```", "_____no_output_____" ], [ "Download the dataset and read it into a *pandas* dataframe.", "_____no_output_____" ] ], [ [ "!conda install -c anaconda xlrd --yes", "Solving environment: done\n\n\n==> WARNING: A newer version of conda exists. <==\n current version: 4.5.11\n latest version: 4.7.10\n\nPlease update conda by running\n\n $ conda update -n base -c defaults conda\n\n\n\n## Package Plan ##\n\n environment location: /home/jupyterlab/conda/envs/python\n\n added / updated specs: \n - xlrd\n\n\nThe following packages will be downloaded:\n\n package | build\n ---------------------------|-----------------\n openssl-1.0.2s | h7b6447c_0 3.1 MB anaconda\n certifi-2019.6.16 | py36_0 154 KB anaconda\n xlrd-1.2.0 | py36_0 188 KB anaconda\n ------------------------------------------------------------\n Total: 3.5 MB\n\nThe following packages will be UPDATED:\n\n openssl: 1.0.2r-h14c3975_0 conda-forge --> 1.0.2s-h7b6447c_0 anaconda\n xlrd: 1.1.0-py37_1 --> 1.2.0-py36_0 anaconda\n\nThe following packages will be DOWNGRADED:\n\n certifi: 2019.6.16-py36_1 conda-forge --> 2019.6.16-py36_0 anaconda\n\n\nDownloading and Extracting Packages\nopenssl-1.0.2s | 3.1 MB | ##################################### | 100% \ncertifi-2019.6.16 | 154 KB | ##################################### | 100% \nxlrd-1.2.0 | 188 KB | ##################################### | 100% \nPreparing transaction: done\nVerifying transaction: done\nExecuting transaction: done\n" ], [ "df_can = pd.read_excel('https://s3-api.us-geo.objectstorage.softlayer.net/cf-courses-data/CognitiveClass/DV0101EN/labs/Data_Files/Canada.xlsx',\n sheet_name='Canada by Citizenship',\n skiprows=range(20),\n skipfooter=2\n )\n\nprint('Data downloaded and read into a dataframe!')", "Data downloaded and read into a dataframe!\n" ] ], [ [ "Let's take a look at the first five items in our dataset.", "_____no_output_____" ] ], [ [ "df_can.head()", "_____no_output_____" ] ], [ [ "Let's find out how many entries there are in our dataset.", "_____no_output_____" ] ], [ [ "# print the dimensions of the dataframe\nprint(df_can.shape)", "(195, 43)\n" ] ], [ [ "Clean up data. We will make some modifications to the original dataset to make it easier to create our visualizations. Refer to *Introduction to Matplotlib and Line Plots* and *Area Plots, Histograms, and Bar Plots* for a detailed description of this preprocessing.", "_____no_output_____" ] ], [ [ "# clean up the dataset to remove unnecessary columns (eg. REG) \ndf_can.drop(['AREA', 'REG', 'DEV', 'Type', 'Coverage'], axis=1, inplace=True)\n\n# let's rename the columns so that they make sense\ndf_can.rename(columns={'OdName':'Country', 'AreaName':'Continent','RegName':'Region'}, inplace=True)\n\n# for sake of consistency, let's also make all column labels of type string\ndf_can.columns = list(map(str, df_can.columns))\n\n# set the country name as index - useful for quickly looking up countries using .loc method\ndf_can.set_index('Country', inplace=True)\n\n# add total column\ndf_can['Total'] = df_can.sum(axis=1)\n\n# years that we will be using in this lesson - useful for plotting later on\nyears = list(map(str, range(1980, 2014)))\nprint('data dimensions:', df_can.shape)", "data dimensions: (195, 38)\n" ] ], [ [ "# Visualizing Data using Matplotlib<a id=\"4\"></a>", "_____no_output_____" ], [ "Import `Matplotlib`.", "_____no_output_____" ] ], [ [ "%matplotlib inline\n\nimport matplotlib as mpl\nimport matplotlib.pyplot as plt\n\nmpl.style.use('ggplot') # optional: for ggplot-like style\n\n# check for latest version of Matplotlib\nprint('Matplotlib version: ', mpl.__version__) # >= 2.0.0", "Matplotlib version: 3.1.1\n" ] ], [ [ "# Pie Charts <a id=\"6\"></a>\n\nA `pie chart` is a circualr graphic that displays numeric proportions by dividing a circle (or pie) into proportional slices. You are most likely already familiar with pie charts as it is widely used in business and media. We can create pie charts in Matplotlib by passing in the `kind=pie` keyword.\n\nLet's use a pie chart to explore the proportion (percentage) of new immigrants grouped by continents for the entire time period from 1980 to 2013. ", "_____no_output_____" ], [ "Step 1: Gather data. \n\nWe will use *pandas* `groupby` method to summarize the immigration data by `Continent`. The general process of `groupby` involves the following steps:\n\n1. **Split:** Splitting the data into groups based on some criteria.\n2. **Apply:** Applying a function to each group independently:\n .sum()\n .count()\n .mean() \n .std() \n .aggregate()\n .apply()\n .etc..\n3. **Combine:** Combining the results into a data structure.", "_____no_output_____" ], [ "<img src=\"https://s3-api.us-geo.objectstorage.softlayer.net/cf-courses-data/CognitiveClass/DV0101EN/labs/Images/Mod3Fig4SplitApplyCombine.png\" height=400 align=\"center\">", "_____no_output_____" ] ], [ [ "# group countries by continents and apply sum() function \ndf_continents = df_can.groupby('Continent', axis=0).sum()\n\n# note: the output of the groupby method is a `groupby' object. \n# we can not use it further until we apply a function (eg .sum())\nprint(type(df_can.groupby('Continent', axis=0)))\n\ndf_continents.head()", "<class 'pandas.core.groupby.generic.DataFrameGroupBy'>\n" ] ], [ [ "Step 2: Plot the data. We will pass in `kind = 'pie'` keyword, along with the following additional parameters:\n- `autopct` - is a string or function used to label the wedges with their numeric value. The label will be placed inside the wedge. If it is a format string, the label will be `fmt%pct`.\n- `startangle` - rotates the start of the pie chart by angle degrees counterclockwise from the x-axis.\n- `shadow` - Draws a shadow beneath the pie (to give a 3D feel).", "_____no_output_____" ] ], [ [ "# autopct create %, start angle represent starting point\ndf_continents['Total'].plot(kind='pie',\n figsize=(5, 6),\n autopct='%1.1f%%', # add in percentages\n startangle=90, # start angle 90° (Africa)\n shadow=True, # add shadow \n )\n\nplt.title('Immigration to Canada by Continent [1980 - 2013]')\nplt.axis('equal') # Sets the pie chart to look like a circle.\n\nplt.show()", "_____no_output_____" ] ], [ [ "The above visual is not very clear, the numbers and text overlap in some instances. Let's make a few modifications to improve the visuals:\n\n* Remove the text labels on the pie chart by passing in `legend` and add it as a seperate legend using `plt.legend()`.\n* Push out the percentages to sit just outside the pie chart by passing in `pctdistance` parameter.\n* Pass in a custom set of colors for continents by passing in `colors` parameter.\n* **Explode** the pie chart to emphasize the lowest three continents (Africa, North America, and Latin America and Carribbean) by pasing in `explode` parameter.\n", "_____no_output_____" ] ], [ [ "colors_list = ['gold', 'yellowgreen', 'lightcoral', 'lightskyblue', 'lightgreen', 'pink']\nexplode_list = [0.1, 0, 0, 0, 0.1, 0.1] # ratio for each continent with which to offset each wedge.\n\ndf_continents['Total'].plot(kind='pie',\n figsize=(15, 6),\n autopct='%1.1f%%', \n startangle=90, \n shadow=True, \n labels=None, # turn off labels on pie chart\n pctdistance=1.12, # the ratio between the center of each pie slice and the start of the text generated by autopct \n colors=colors_list, # add custom colors\n explode=explode_list # 'explode' lowest 3 continents\n )\n\n# scale the title up by 12% to match pctdistance\nplt.title('Immigration to Canada by Continent [1980 - 2013]', y=1.12) \n\nplt.axis('equal') \n\n# add legend\nplt.legend(labels=df_continents.index, loc='upper left') \n\nplt.show()", "_____no_output_____" ] ], [ [ "**Question:** Using a pie chart, explore the proportion (percentage) of new immigrants grouped by continents in the year 2013.\n\n**Note**: You might need to play with the explore values in order to fix any overlapping slice values.", "_____no_output_____" ] ], [ [ "### type your answer here\nexplode_list = [0.1, 0, 0, 0, 0.1, 0.2]\ndf_continents['2013'].plot(kind='pie',\n figsize=(15, 6),\n autopct='%1.1f%%', \n startangle=90, \n shadow=True, \n labels=None, \n pctdistance=1.12, \n explode=explode_list \n )\nplt.title('Immigration to Canada by Continent in 2013', y=1.12) \nplt.axis('equal') \nplt.legend(labels=df_continents.index, loc='upper left')\nplt.show()\n", "_____no_output_____" ] ], [ [ "Double-click __here__ for the solution.\n\n\n\n\n\n\n\n\n", "_____no_output_____" ], [ "# Box Plots <a id=\"8\"></a>\n\nA `box plot` is a way of statistically representing the *distribution* of the data through five main dimensions: \n\n- **Minimun:** Smallest number in the dataset.\n- **First quartile:** Middle number between the `minimum` and the `median`.\n- **Second quartile (Median):** Middle number of the (sorted) dataset.\n- **Third quartile:** Middle number between `median` and `maximum`.\n- **Maximum:** Highest number in the dataset.", "_____no_output_____" ], [ "<img src=\"https://s3-api.us-geo.objectstorage.softlayer.net/cf-courses-data/CognitiveClass/DV0101EN/labs/Images/boxplot_complete.png\" width=440, align=\"center\">", "_____no_output_____" ], [ "To make a `box plot`, we can use `kind=box` in `plot` method invoked on a *pandas* series or dataframe.\n\nLet's plot the box plot for the Japanese immigrants between 1980 - 2013.", "_____no_output_____" ], [ "Step 1: Get the dataset. Even though we are extracting the data for just one country, we will obtain it as a dataframe. This will help us with calling the `dataframe.describe()` method to view the percentiles.", "_____no_output_____" ] ], [ [ "# to get a dataframe, place extra square brackets around 'Japan'.\ndf_japan = df_can.loc[['Japan'], years].transpose()\ndf_japan.head()", "_____no_output_____" ] ], [ [ "Step 2: Plot by passing in `kind='box'`.", "_____no_output_____" ] ], [ [ "df_japan.plot(kind='box', figsize=(8, 6))\n\nplt.title('Box plot of Japanese Immigrants from 1980 - 2013')\nplt.ylabel('Number of Immigrants')\n\nplt.show()", "_____no_output_____" ] ], [ [ "We can immediately make a few key observations from the plot above:\n1. The minimum number of immigrants is around 200 (min), maximum number is around 1300 (max), and median number of immigrants is around 900 (median).\n2. 25% of the years for period 1980 - 2013 had an annual immigrant count of ~500 or fewer (First quartile).\n2. 75% of the years for period 1980 - 2013 had an annual immigrant count of ~1100 or fewer (Third quartile).\n\nWe can view the actual numbers by calling the `describe()` method on the dataframe.", "_____no_output_____" ] ], [ [ "df_japan.describe()", "_____no_output_____" ] ], [ [ "One of the key benefits of box plots is comparing the distribution of multiple datasets. In one of the previous labs, we observed that China and India had very similar immigration trends. Let's analyize these two countries further using box plots.\n\n**Question:** Compare the distribution of the number of new immigrants from India and China for the period 1980 - 2013.", "_____no_output_____" ], [ "Step 1: Get the dataset for China and India and call the dataframe **df_CI**.", "_____no_output_____" ] ], [ [ "### type your answer here\ndf_CI= df_can.loc[['China', 'India'], years].transpose()\ndf_CI.head()\n\n\n", "_____no_output_____" ] ], [ [ "Double-click __here__ for the solution.\n", "_____no_output_____" ], [ "Let's view the percentages associated with both countries using the `describe()` method.", "_____no_output_____" ] ], [ [ "### type your answer here\ndf_CI.describe()\n", "_____no_output_____" ] ], [ [ "Double-click __here__ for the solution.\n", "_____no_output_____" ], [ "Step 2: Plot data.", "_____no_output_____" ] ], [ [ "### type your answer here\ndf_CI.plot(kind='box', figsize=(10, 7))\nplt.title('Box plots of Immigrants from China and India (1980 - 2013)')\nplt.xlabel('Number of Immigrants')\nplt.show()\n", "_____no_output_____" ] ], [ [ "Double-click __here__ for the solution.\n\n\n\n\n", "_____no_output_____" ], [ "We can observe that, while both countries have around the same median immigrant population (~20,000), China's immigrant population range is more spread out than India's. The maximum population from India for any year (36,210) is around 15% lower than the maximum population from China (42,584).\n", "_____no_output_____" ], [ "If you prefer to create horizontal box plots, you can pass the `vert` parameter in the **plot** function and assign it to *False*. You can also specify a different color in case you are not a big fan of the default red color.", "_____no_output_____" ] ], [ [ "# horizontal box plots\ndf_CI.plot(kind='box', figsize=(10, 7), color='blue', vert=False)\n\nplt.title('Box plots of Immigrants from China and India (1980 - 2013)')\nplt.xlabel('Number of Immigrants')\n\nplt.show()", "_____no_output_____" ] ], [ [ "**Subplots**\n\nOften times we might want to plot multiple plots within the same figure. For example, we might want to perform a side by side comparison of the box plot with the line plot of China and India's immigration.\n\nTo visualize multiple plots together, we can create a **`figure`** (overall canvas) and divide it into **`subplots`**, each containing a plot. With **subplots**, we usually work with the **artist layer** instead of the **scripting layer**. \n\nTypical syntax is : <br>\n```python\n fig = plt.figure() # create figure\n ax = fig.add_subplot(nrows, ncols, plot_number) # create subplots\n```\nWhere\n- `nrows` and `ncols` are used to notionally split the figure into (`nrows` \\* `ncols`) sub-axes, \n- `plot_number` is used to identify the particular subplot that this function is to create within the notional grid. `plot_number` starts at 1, increments across rows first and has a maximum of `nrows` * `ncols` as shown below.\n\n<img src=\"https://s3-api.us-geo.objectstorage.softlayer.net/cf-courses-data/CognitiveClass/DV0101EN/labs/Images/Mod3Fig5Subplots_V2.png\" width=500 align=\"center\">", "_____no_output_____" ], [ "We can then specify which subplot to place each plot by passing in the `ax` paramemter in `plot()` method as follows:", "_____no_output_____" ] ], [ [ "fig = plt.figure() # create figure\n\nax0 = fig.add_subplot(1, 2, 1) # add subplot 1 (1 row, 2 columns, first plot)\nax1 = fig.add_subplot(1, 2, 2) # add subplot 2 (1 row, 2 columns, second plot). See tip below**\n\n# Subplot 1: Box plot\ndf_CI.plot(kind='box', color='blue', vert=False, figsize=(20, 6), ax=ax0) # add to subplot 1\nax0.set_title('Box Plots of Immigrants from China and India (1980 - 2013)')\nax0.set_xlabel('Number of Immigrants')\nax0.set_ylabel('Countries')\n\n# Subplot 2: Line plot\ndf_CI.plot(kind='line', figsize=(20, 6), ax=ax1) # add to subplot 2\nax1.set_title ('Line Plots of Immigrants from China and India (1980 - 2013)')\nax1.set_ylabel('Number of Immigrants')\nax1.set_xlabel('Years')\n\nplt.show()", "_____no_output_____" ] ], [ [ "** * Tip regarding subplot convention **\n\nIn the case when `nrows`, `ncols`, and `plot_number` are all less than 10, a convenience exists such that the a 3 digit number can be given instead, where the hundreds represent `nrows`, the tens represent `ncols` and the units represent `plot_number`. For instance,\n```python\n subplot(211) == subplot(2, 1, 1) \n```\nproduces a subaxes in a figure which represents the top plot (i.e. the first) in a 2 rows by 1 column notional grid (no grid actually exists, but conceptually this is how the returned subplot has been positioned).", "_____no_output_____" ], [ "Let's try something a little more advanced. \n\nPreviously we identified the top 15 countries based on total immigration from 1980 - 2013.\n\n**Question:** Create a box plot to visualize the distribution of the top 15 countries (based on total immigration) grouped by the *decades* `1980s`, `1990s`, and `2000s`.", "_____no_output_____" ], [ "Step 1: Get the dataset. Get the top 15 countries based on Total immigrant population. Name the dataframe **df_top15**.", "_____no_output_____" ] ], [ [ "### type your answer here\ndf_top15 = df_can.sort_values(['Total'], ascending=False, axis=0).head(15)\ndf_top15\n\n\n", "_____no_output_____" ] ], [ [ "Double-click __here__ for the solution.\n", "_____no_output_____" ], [ "Step 2: Create a new dataframe which contains the aggregate for each decade. One way to do that:\n 1. Create a list of all years in decades 80's, 90's, and 00's.\n 2. Slice the original dataframe df_can to create a series for each decade and sum across all years for each country.\n 3. Merge the three series into a new data frame. Call your dataframe **new_df**.", "_____no_output_____" ] ], [ [ "### type your answer here\n# create a list of all years in decades 80's, 90's, and 00's\nyears_80s = list(map(str, range(1980, 1990))) \nyears_90s = list(map(str, range(1990, 2000))) \nyears_00s = list(map(str, range(2000, 2010))) \n # slice the original dataframe df_can to create a series for each decade\ndf_80s = df_top15.loc[:, years_80s].sum(axis=1) \ndf_90s = df_top15.loc[:, years_90s].sum(axis=1) \ndf_00s = df_top15.loc[:, years_00s].sum(axis=1)\n# merge the three series into a new data frame\nnew_df = pd.DataFrame({'1980s': df_80s, '1990s': df_90s, '2000s':df_00s}) \n# display dataframe\nnew_df.head()", "_____no_output_____" ] ], [ [ "Double-click __here__ for the solution.\n\n\n\n\n\n\n", "_____no_output_____" ], [ "Let's learn more about the statistics associated with the dataframe using the `describe()` method.", "_____no_output_____" ] ], [ [ "### type your answer here\nnew_df.describe()\n", "_____no_output_____" ] ], [ [ "Double-click __here__ for the solution.\n", "_____no_output_____" ], [ "Step 3: Plot the box plots.", "_____no_output_____" ] ], [ [ "### type your answer here\nnew_df.plot(kind='box', figsize=(10, 6))\nplt.title('Immigration from top 15 countries for decades 80s, 90s and 2000s')\nplt.show()\n\n", "_____no_output_____" ] ], [ [ "Double-click __here__ for the solution.\n\n\n\n\n", "_____no_output_____" ], [ "Note how the box plot differs from the summary table created. The box plot scans the data and identifies the outliers. In order to be an outlier, the data value must be:<br>\n* larger than Q3 by at least 1.5 times the interquartile range (IQR), or,\n* smaller than Q1 by at least 1.5 times the IQR.\n\nLet's look at decade 2000s as an example: <br>\n* Q1 (25%) = 36,101.5 <br>\n* Q3 (75%) = 105,505.5 <br>\n* IQR = Q3 - Q1 = 69,404 <br>\n\nUsing the definition of outlier, any value that is greater than Q3 by 1.5 times IQR will be flagged as outlier.\n\nOutlier > 105,505.5 + (1.5 * 69,404) <br>\nOutlier > 209,611.5", "_____no_output_____" ] ], [ [ "# let's check how many entries fall above the outlier threshold \nnew_df[new_df['2000s']> 209611.5]", "_____no_output_____" ] ], [ [ "China and India are both considered as outliers since their population for the decade exceeds 209,611.5. \n\nThe box plot is an advanced visualizaiton tool, and there are many options and customizations that exceed the scope of this lab. Please refer to [Matplotlib documentation](http://matplotlib.org/api/pyplot_api.html#matplotlib.pyplot.boxplot) on box plots for more information.", "_____no_output_____" ], [ "# Scatter Plots <a id=\"10\"></a>\n\nA `scatter plot` (2D) is a useful method of comparing variables against each other. `Scatter` plots look similar to `line plots` in that they both map independent and dependent variables on a 2D graph. While the datapoints are connected together by a line in a line plot, they are not connected in a scatter plot. The data in a scatter plot is considered to express a trend. With further analysis using tools like regression, we can mathematically calculate this relationship and use it to predict trends outside the dataset.\n\nLet's start by exploring the following:\n\nUsing a `scatter plot`, let's visualize the trend of total immigrantion to Canada (all countries combined) for the years 1980 - 2013.", "_____no_output_____" ], [ "Step 1: Get the dataset. Since we are expecting to use the relationship betewen `years` and `total population`, we will convert `years` to `int` type.", "_____no_output_____" ] ], [ [ "# we can use the sum() method to get the total population per year\ndf_tot = pd.DataFrame(df_can[years].sum(axis=0))\n\n# change the years to type int (useful for regression later on)\ndf_tot.index = map(int, df_tot.index)\n\n# reset the index to put in back in as a column in the df_tot dataframe\ndf_tot.reset_index(inplace = True)\n\n# rename columns\ndf_tot.columns = ['year', 'total']\n\n# view the final dataframe\ndf_tot.head()", "_____no_output_____" ] ], [ [ "Step 2: Plot the data. In `Matplotlib`, we can create a `scatter` plot set by passing in `kind='scatter'` as plot argument. We will also need to pass in `x` and `y` keywords to specify the columns that go on the x- and the y-axis.", "_____no_output_____" ] ], [ [ "df_tot.plot(kind='scatter', x='year', y='total', figsize=(10, 6), color='darkblue')\n\nplt.title('Total Immigration to Canada from 1980 - 2013')\nplt.xlabel('Year')\nplt.ylabel('Number of Immigrants')\n\nplt.show()", "_____no_output_____" ] ], [ [ "Notice how the scatter plot does not connect the datapoints together. We can clearly observe an upward trend in the data: as the years go by, the total number of immigrants increases. We can mathematically analyze this upward trend using a regression line (line of best fit). ", "_____no_output_____" ], [ "So let's try to plot a linear line of best fit, and use it to predict the number of immigrants in 2015.\n\nStep 1: Get the equation of line of best fit. We will use **Numpy**'s `polyfit()` method by passing in the following:\n- `x`: x-coordinates of the data. \n- `y`: y-coordinates of the data. \n- `deg`: Degree of fitting polynomial. 1 = linear, 2 = quadratic, and so on.", "_____no_output_____" ] ], [ [ "x = df_tot['year'] # year on x-axis\ny = df_tot['total'] # total on y-axis\nfit = np.polyfit(x, y, deg=1)\n\nfit", "_____no_output_____" ] ], [ [ "The output is an array with the polynomial coefficients, highest powers first. Since we are plotting a linear regression `y= a*x + b`, our output has 2 elements `[5.56709228e+03, -1.09261952e+07]` with the the slope in position 0 and intercept in position 1. \n\nStep 2: Plot the regression line on the `scatter plot`.", "_____no_output_____" ] ], [ [ "df_tot.plot(kind='scatter', x='year', y='total', figsize=(10, 6), color='darkblue')\n\nplt.title('Total Immigration to Canada from 1980 - 2013')\nplt.xlabel('Year')\nplt.ylabel('Number of Immigrants')\n\n# plot line of best fit\nplt.plot(x, fit[0] * x + fit[1], color='red') # recall that x is the Years\nplt.annotate('y={0:.0f} x + {1:.0f}'.format(fit[0], fit[1]), xy=(2000, 150000))\n\nplt.show()\n\n# print out the line of best fit\n'No. Immigrants = {0:.0f} * Year + {1:.0f}'.format(fit[0], fit[1]) ", "_____no_output_____" ] ], [ [ "Using the equation of line of best fit, we can estimate the number of immigrants in 2015:\n```python\nNo. Immigrants = 5567 * Year - 10926195\nNo. Immigrants = 5567 * 2015 - 10926195\nNo. Immigrants = 291,310\n```\nWhen compared to the actuals from Citizenship and Immigration Canada's (CIC) [2016 Annual Report](http://www.cic.gc.ca/english/resources/publications/annual-report-2016/index.asp), we see that Canada accepted 271,845 immigrants in 2015. Our estimated value of 291,310 is within 7% of the actual number, which is pretty good considering our original data came from United Nations (and might differ slightly from CIC data).\n\nAs a side note, we can observe that immigration took a dip around 1993 - 1997. Further analysis into the topic revealed that in 1993 Canada introcuded Bill C-86 which introduced revisions to the refugee determination system, mostly restrictive. Further amendments to the Immigration Regulations cancelled the sponsorship required for \"assisted relatives\" and reduced the points awarded to them, making it more difficult for family members (other than nuclear family) to immigrate to Canada. These restrictive measures had a direct impact on the immigration numbers for the next several years.", "_____no_output_____" ], [ "**Question**: Create a scatter plot of the total immigration from Denmark, Norway, and Sweden to Canada from 1980 to 2013?", "_____no_output_____" ], [ "Step 1: Get the data:\n 1. Create a dataframe the consists of the numbers associated with Denmark, Norway, and Sweden only. Name it **df_countries**.\n 2. Sum the immigration numbers across all three countries for each year and turn the result into a dataframe. Name this new dataframe **df_total**.\n 3. Reset the index in place.\n 4. Rename the columns to **year** and **total**.\n 5. Display the resulting dataframe.", "_____no_output_____" ] ], [ [ "### type your answer here\ndf_countries = df_can.loc[['Denmark', 'Norway', 'Sweden'], years].transpose()\ndf_total = pd.DataFrame(df_countries.sum(axis=1))\ndf_total.reset_index(inplace=True)\ndf_total.columns = ['year', 'total']\ndf_total['year'] = df_total['year'].astype(int)\ndf_total.head()", "_____no_output_____" ] ], [ [ "Double-click __here__ for the solution.\n\n\n\n\n\n\n\n\n\n\n", "_____no_output_____" ], [ "Step 2: Generate the scatter plot by plotting the total versus year in **df_total**.", "_____no_output_____" ] ], [ [ "### type your answer here\ndf_total.plot(kind='scatter', x='year', y='total', figsize=(10, 6), color='darkblue')\n# add title and label to axes\nplt.title('Immigration from Denmark, Norway, and Sweden to Canada from 1980 - 2013')\nplt.xlabel('Year')\nplt.ylabel('Number of Immigrants')\n# show plot\nplt.show()\n\n\n", "_____no_output_____" ] ], [ [ "Double-click __here__ for the solution.\n\n\n\n\n", "_____no_output_____" ], [ "# Bubble Plots <a id=\"12\"></a>\n\nA `bubble plot` is a variation of the `scatter plot` that displays three dimensions of data (x, y, z). The datapoints are replaced with bubbles, and the size of the bubble is determined by the third variable 'z', also known as the weight. In `maplotlib`, we can pass in an array or scalar to the keyword `s` to `plot()`, that contains the weight of each point.\n\n**Let's start by analyzing the effect of Argentina's great depression**.\n\nArgentina suffered a great depression from 1998 - 2002, which caused widespread unemployment, riots, the fall of the government, and a default on the country's foreign debt. In terms of income, over 50% of Argentines were poor, and seven out of ten Argentine children were poor at the depth of the crisis in 2002. \n\nLet's analyze the effect of this crisis, and compare Argentina's immigration to that of it's neighbour Brazil. Let's do that using a `bubble plot` of immigration from Brazil and Argentina for the years 1980 - 2013. We will set the weights for the bubble as the *normalized* value of the population for each year.", "_____no_output_____" ], [ "Step 1: Get the data for Brazil and Argentina. Like in the previous example, we will convert the `Years` to type int and bring it in the dataframe.", "_____no_output_____" ] ], [ [ "df_can_t = df_can[years].transpose() # transposed dataframe\n\n# cast the Years (the index) to type int\ndf_can_t.index = map(int, df_can_t.index)\n\n# let's label the index. This will automatically be the column name when we reset the index\ndf_can_t.index.name = 'Year'\n\n# reset index to bring the Year in as a column\ndf_can_t.reset_index(inplace=True)\n\n# view the changes\ndf_can_t.head()", "_____no_output_____" ] ], [ [ "Step 2: Create the normalized weights. \n\nThere are several methods of normalizations in statistics, each with its own use. In this case, we will use [feature scaling](https://en.wikipedia.org/wiki/Feature_scaling) to bring all values into the range [0,1]. The general formula is:\n\n<img src=\"https://s3-api.us-geo.objectstorage.softlayer.net/cf-courses-data/CognitiveClass/DV0101EN/labs/Images/Mod3Fig3FeatureScaling.png\" align=\"center\">\n\nwhere *`X`* is an original value, *`X'`* is the normalized value. The formula sets the max value in the dataset to 1, and sets the min value to 0. The rest of the datapoints are scaled to a value between 0-1 accordingly.\n", "_____no_output_____" ] ], [ [ "# normalize Brazil data\nnorm_brazil = (df_can_t['Brazil'] - df_can_t['Brazil'].min()) / (df_can_t['Brazil'].max() - df_can_t['Brazil'].min())\n\n# normalize Argentina data\nnorm_argentina = (df_can_t['Argentina'] - df_can_t['Argentina'].min()) / (df_can_t['Argentina'].max() - df_can_t['Argentina'].min())", "_____no_output_____" ] ], [ [ "Step 3: Plot the data. \n- To plot two different scatter plots in one plot, we can include the axes one plot into the other by passing it via the `ax` parameter. \n- We will also pass in the weights using the `s` parameter. Given that the normalized weights are between 0-1, they won't be visible on the plot. Therefore we will:\n - multiply weights by 2000 to scale it up on the graph, and,\n - add 10 to compensate for the min value (which has a 0 weight and therefore scale with x2000).", "_____no_output_____" ] ], [ [ "# Brazil\nax0 = df_can_t.plot(kind='scatter',\n x='Year',\n y='Brazil',\n figsize=(14, 8),\n alpha=0.5, # transparency\n color='green',\n s=norm_brazil * 2000 + 10, # pass in weights \n xlim=(1975, 2015)\n )\n\n# Argentina\nax1 = df_can_t.plot(kind='scatter',\n x='Year',\n y='Argentina',\n alpha=0.5,\n color=\"blue\",\n s=norm_argentina * 2000 + 10,\n ax = ax0\n )\n\nax0.set_ylabel('Number of Immigrants')\nax0.set_title('Immigration from Brazil and Argentina from 1980 - 2013')\nax0.legend(['Brazil', 'Argentina'], loc='upper left', fontsize='x-large')", "_____no_output_____" ] ], [ [ "The size of the bubble corresponds to the magnitude of immigrating population for that year, compared to the 1980 - 2013 data. The larger the bubble, the more immigrants in that year.\n\nFrom the plot above, we can see a corresponding increase in immigration from Argentina during the 1998 - 2002 great depression. We can also observe a similar spike around 1985 to 1993. In fact, Argentina had suffered a great depression from 1974 - 1990, just before the onset of 1998 - 2002 great depression. \n\nOn a similar note, Brazil suffered the *Samba Effect* where the Brazilian real (currency) dropped nearly 35% in 1999. There was a fear of a South American financial crisis as many South American countries were heavily dependent on industrial exports from Brazil. The Brazilian government subsequently adopted an austerity program, and the economy slowly recovered over the years, culminating in a surge in 2010. The immigration data reflect these events.", "_____no_output_____" ], [ "**Question**: Previously in this lab, we created box plots to compare immigration from China and India to Canada. Create bubble plots of immigration from China and India to visualize any differences with time from 1980 to 2013. You can use **df_can_t** that we defined and used in the previous example.", "_____no_output_____" ], [ "Step 1: Normalize the data pertaining to China and India.", "_____no_output_____" ] ], [ [ "### type your answer here\n# normalize China data\nnorm_china = (df_can_t['China'] - df_can_t['China'].min()) / (df_can_t['China'].max() - df_can_t['China'].min())\n\n# normalize India data\nnorm_india = (df_can_t['India'] - df_can_t['India'].min()) / (df_can_t['India'].max() - df_can_t['India'].min())\n\n", "_____no_output_____" ] ], [ [ "Double-click __here__ for the solution.\n\n\n", "_____no_output_____" ], [ "Step 2: Generate the bubble plots.", "_____no_output_____" ] ], [ [ "### type your answer here\n# China\nax0 = df_can_t.plot(kind='scatter',\n x='Year',\n y='China',\n figsize=(14, 8),\n alpha=0.5, # transparency\n color='green',\n s=norm_china * 2000 + 10, # pass in weights \n xlim=(1975, 2015)\n )\n# India\nax1 = df_can_t.plot(kind='scatter',\n x='Year',\n y='India',\n alpha=0.5,\n color=\"blue\",\n s=norm_india * 2000 + 10,\n ax = ax0\n )\n\nax0.set_ylabel('Number of Immigrants')\nax0.set_title('Immigration from China and India from 1980 - 2013')\nax0.legend(['China', 'India'], loc='upper left', fontsize='x-large')\n\n\n", "_____no_output_____" ] ], [ [ "Double-click __here__ for the solution.\n\n\n\n\n", "_____no_output_____" ], [ "### Thank you for completing this lab!\n\nThis notebook was created by [Jay Rajasekharan](https://www.linkedin.com/in/jayrajasekharan) with contributions from [Ehsan M. Kermani](https://www.linkedin.com/in/ehsanmkermani), and [Slobodan Markovic](https://www.linkedin.com/in/slobodan-markovic).\n\nThis notebook was recently revamped by [Alex Aklson](https://www.linkedin.com/in/aklson/). I hope you found this lab session interesting. Feel free to contact me if you have any questions!", "_____no_output_____" ], [ "This notebook is part of a course on **Coursera** called *Data Visualization with Python*. If you accessed this notebook outside the course, you can take this course online by clicking [here](http://cocl.us/DV0101EN_Coursera_Week2_LAB2).", "_____no_output_____" ], [ "<hr>\n\nCopyright © 2019 [Cognitive Class](https://cognitiveclass.ai/?utm_source=bducopyrightlink&utm_medium=dswb&utm_campaign=bdu). This notebook and its source code are released under the terms of the [MIT License](https://bigdatauniversity.com/mit-license/).", "_____no_output_____" ] ] ]

[ "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown", "code", "markdown" ]

[ [ "markdown", "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code", "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown" ], [ "code" ], [ "markdown", "markdown", "markdown", "markdown" ] ]

ec7e56e5467a99776ae1ac241102e28d77629d89

29,860

ipynb

Jupyter Notebook

Udacity/HoughTransforms.ipynb

deepaksood619/DS_ML

2f5b374cddd11c8d5f7d4c980e62fb34c2beb6b5

[ "MIT" ]

null

Udacity/HoughTransforms.ipynb

deepaksood619/DS_ML

2f5b374cddd11c8d5f7d4c980e62fb34c2beb6b5

[ "MIT" ]

null

Udacity/HoughTransforms.ipynb

deepaksood619/DS_ML

2f5b374cddd11c8d5f7d4c980e62fb34c2beb6b5

[ "MIT" ]

null

168.700565

13,376

0.888212

[ [ [ "# Do relevant imports\nimport matplotlib.pyplot as plt\nimport matplotlib.image as mpimg\nimport numpy as np\nimport cv2\n\n# Read in and grayscale the image\nimage = mpimg.imread('exit-ramp.jpg')\ngray = cv2.cvtColor(image,cv2.COLOR_RGB2GRAY)\n\n# Define a kernel size and apply Gaussian smoothing\nkernel_size = 5\nblur_gray = cv2.GaussianBlur(gray,(kernel_size, kernel_size),0)\n\n# Define our parameters for Canny and apply\nlow_threshold = 50\nhigh_threshold = 150\nmasked_edges = cv2.Canny(blur_gray, low_threshold, high_threshold)\n\n# Define the Hough transform parameters\n# Make a blank the same size as our image to draw on\nrho = 1\ntheta = np.pi/180\nthreshold = 1\nmin_line_length = 10\nmax_line_gap = 1\nline_image = np.copy(image)*0 #creating a blank to draw lines on\n\n# Run Hough on edge detected image\nlines = cv2.HoughLinesP(masked_edges, rho, theta, threshold, np.array([]),\n min_line_length, max_line_gap)\n\n# Iterate over the output \"lines\" and draw lines on the blank\nfor line in lines:\n for x1,y1,x2,y2 in line:\n cv2.line(line_image,(x1,y1),(x2,y2),(255,0,0),10)\n\n# Create a \"color\" binary image to combine with line image\ncolor_edges = np.dstack((masked_edges, masked_edges, masked_edges)) \n\n# Draw the lines on the edge image\ncombo = cv2.addWeighted(color_edges, 0.8, line_image, 1, 0) \nplt.imshow(combo)\nplt.show()", "_____no_output_____" ], [ "import matplotlib.pyplot as plt\nimport matplotlib.image as mpimg\nimport numpy as np\nimport cv2\n\n\n# Read in and grayscale the image\nimage = mpimg.imread('exit-ramp.jpg')\ngray = cv2.cvtColor(image,cv2.COLOR_RGB2GRAY)\n\n# Define a kernel size and apply Gaussian smoothing\nkernel_size = 5\nblur_gray = cv2.GaussianBlur(gray,(kernel_size, kernel_size),0)\n\n# Define our parameters for Canny and apply\nlow_threshold = 50\nhigh_threshold = 150\nedges = cv2.Canny(blur_gray, low_threshold, high_threshold)\n\n# Next we'll create a masked edges image using cv2.fillPoly()\nmask = np.zeros_like(edges) \nignore_mask_color = 255 \n\n# This time we are defining a four sided polygon to mask\nimshape = image.shape\nvertices = np.array([[(45,imshape[0]),(450, 290), (490, 290), (imshape[1]-20,imshape[0])]], dtype=np.int32)\ncv2.fillPoly(mask, vertices, ignore_mask_color)\nmasked_edges = cv2.bitwise_and(edges, mask)\n\n# Define the Hough transform parameters\n# Make a blank the same size as our image to draw on\nrho = 2 # distance resolution in pixels of the Hough grid\ntheta = np.pi/180 # angular resolution in radians of the Hough grid\nthreshold = 15 # minimum number of votes (intersections in Hough grid cell)\nmin_line_length = 40 #minimum number of pixels making up a line\nmax_line_gap = 20 # maximum gap in pixels between connectable line segments\nline_image = np.copy(image)*0 # creating a blank to draw lines on\n\n# Run Hough on edge detected image\n# Output \"lines\" is an array containing endpoints of detected line segments\nlines = cv2.HoughLinesP(masked_edges, rho, theta, threshold, np.array([]),\n min_line_length, max_line_gap)\n\n# Iterate over the output \"lines\" and draw lines on a blank image\nfor line in lines:\n for x1,y1,x2,y2 in line:\n cv2.line(line_image,(x1,y1),(x2,y2),(255,0,0),10)\n\n# Create a \"color\" binary image to combine with line image\ncolor_edges = np.dstack((edges, edges, edges)) \n\n# Draw the lines on the edge image\nlines_edges = cv2.addWeighted(color_edges, 0.8, line_image, 1, 0) \nplt.imshow(lines_edges)\n\n", "_____no_output_____" ] ] ]

[ "code" ]

[ [ "code", "code" ] ]