Datasets:

---

DanielPFlorian

/

Transformers-Github-Issues

like 0

[]

[]

[]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[]

[]

[]

[ "You're creating a configuration based on the size of `bert-base-uncased`, so yes, it will work for that checkpoint. For any other checkpoint, however, you would need to change the values which are different (e.g. vocab size, which is what's failing in your case). It is indicated in the [documentation](https://huggingface.co/transformers/model_doc/bert.html#bertconfig):\r\n\r\n> This is the configuration class to store the configuration of a BertModel. It is used to instantiate an BERT model according to the specified arguments, defining the model architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of the BERT bert-base-uncased architecture.\r\n\r\nIn order to load a configuration automatically from a checkpoint, you may use `from_pretrained` on the configuration as well. In your case, that would be:\r\n\r\n```py\r\nimport transformers\r\n\r\nconfig = transformers.BertConfig.from_pretrained(\"bert-base-multilingual-cased\", output_hidden_states=True)\r\nbert_model = transformers.AutoModel.from_pretrained('bert-base-multilingual-cased', config=config)\r\nprint(bert_model)\r\n```" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null }, { "id": 1771187924, "node_id": "MDU6TGFiZWwxNzcxMTg3OTI0", "url": "https://api.github.com/repos/huggingface/transformers/labels/Core:%20Pipeline", "name": "Core: Pipeline", "color": "FF7066", "default": false, "description": "Internals of the library; Pipeline." }, { "id": 1834052129, "node_id": "MDU6TGFiZWwxODM0MDUyMTI5", "url": "https://api.github.com/repos/huggingface/transformers/labels/High-Level%20feature", "name": "High-Level feature", "color": "f7c9a3", "default": false, "description": "" }, { "id": 1834052333, "node_id": "MDU6TGFiZWwxODM0MDUyMzMz", "url": "https://api.github.com/repos/huggingface/transformers/labels/Ex:%20Question%20Answering", "name": "Ex: Question Answering", "color": "86FFCF", "default": false, "description": "" } ]

[]

[ "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n", "I need a solution to this problem as well please help\r\n\r\nThanks.", "The solution is to use the `topk` parameter, for example, following the example in https://huggingface.co/transformers/task_summary.html about question answering, if you want 3 answer:\r\n\r\n` result = nlp(question=\"What is extractive question answering?\", context=context, topk = 3)`", "@fumpe Can this work for sagemaker? Actually I am trying to deploy this pipeline in sagemaker, how can we customize the topk parameter in that? (As Sagemaker works a bit differently), Sagemaker currently returns only 1 answer, and am unable to modify the parameters to increase the number of returned items. Thank you.\r\n\r\nThis is my model and pipeline setup in Sagemaker:\r\nhub = {\r\n'HF_MODEL_ID':'valhalla/t5-base-qa-qg-hl',\r\n'HF_TASK':'text2text-generation'\r\n}" ]

[]

[]

[ "Did you take a look at the [distillation README](https://github.com/huggingface/transformers/tree/master/examples/distillation)? It shows the command that was used to train the distilled model.", "Hey @silencio94, if you need help with distillation, drop me an email at clement [at] huggingface [dot] co.", "Thanks for replying. I'm planing to do some experiments on my DistilBERT after distillation. after Then I'll send an email or leave another issue. thanks!" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "Right here\r\n\r\nhttps://github.com/huggingface/transformers/blob/31f2437f07cf014a042789f52fa1519a485e8b2b/src/transformers/modeling_bert.py#L150\r\n\r\nIn the future, please make an effort to write a decent post that explains exactly what you need. This is very low quality.", "Sry for low quality. But how about its initialization？ In the paper， it uses sin and cos functions， while the code seems like using random initialization.\n\n\n\n\n| |\nWang\n|\n|\n邮箱：[email protected]\n|\n\nSignature is customized by Netease Mail Master\n\nOn 03/10/2020 22:15, Bram Vanroy wrote:\n\nRight here\n\nhttps://github.com/huggingface/transformers/blob/31f2437f07cf014a042789f52fa1519a485e8b2b/src/transformers/modeling_bert.py#L150\n\nIn the future, please make an effort to write a decent post that explains exactly what you need. This is very low quality.\n\n—\nYou are receiving this because you authored the thread.\nReply to this email directly, view it on GitHub, or unsubscribe.", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "Plus：I have seen a similiar issue in this project,however the problem in that issue is that he did not input the right pretrain_weights.But I do not think that will be the solution in here", "Similiarly,I aslo tried DistilBert,Roberta,XLMRoberta,these 3 models also cannot work for me,the error message is the same as the one I described above.", "I just tried this and cannot reproduce the behaviour that you indicate. Are you running this from a notebook? Try restarting your kernel and running it again.", "> I just tried this and cannot reproduce the behaviour that you indicate. Are you running this from a notebook? Try restarting your kernel and running it again.\r\n\r\nI run this programme on the linux GPU server,I tried restarting the python programme,however,the problem is still exsiting.Would this be the problem of downloading the model? ", "No. UnboundLocalError simply means that Python hasn't seen this variable before, which cannot occur in your code snippet. If the models were downloaded incorrectly, you'd get another error. Even if the `tokenizer` was initialized as `None` you'd get another error. \r\n\r\nAre you sure that is your _only_ code that is running? Please pos the full trace.", "> No. UnboundLocalError simply means that Python hasn't seen this variable before, which cannot occur in your code snippet. If the models were downloaded incorrectly, you'd get another error. Even if the `tokenizer` was initialized as `None` you'd get another error.\r\n> \r\n> Are you sure that is your _only_ code that is running? Please pos the full trace.\r\n\r\n```\r\nTraceback (most recent call last):\r\n File \"<stdin>\", line 3, in <module>\r\n File \"/users4/bwchen/anaconda3/lib/python3.7/site-packages/transformers/tokenization_utils.py\", line 302, in from_pretrained\r\n return cls._from_pretrained(*inputs, **kwargs)\r\n File \"/users4/bwchen/anaconda3/lib/python3.7/site-packages/transformers/tokenization_utils.py\", line 438, in _from_pretrained\r\n tokenizer = cls(*init_inputs, **init_kwargs)\r\n File \"/users4/bwchen/anaconda3/lib/python3.7/site-packages/transformers/tokenization_bert.py\", line 164, in __init__\r\n \"model use `tokenizer = BertTokenizer.from_pretrained(PRETRAINED_MODEL_NAME)`\".format(vocab_file))\r\nValueError: Can't find a vocabulary file at path '/users4/bwchen/.cache/torch/transformers/37cc1eaaea18a456726fc28ecb438852f0ca1d9e7d259e6e3747ee33065936f6'. To load the vocabulary from a Google pretrained model use `tokenizer = BertTokenizer.from_pretrained(PRETRAINED_MODEL_NAME)`\r\n```\r\nI am sure that is the only code I was running at that time , I am tring to reproduce this error.This time it is working properly when the model_class goes the aforementioned 'wrong' model XLMModel. However,when the model continues to run,I met another problem when the model was the DistillBert, does this error means that I have to use BertTokenizer instead of DistillBertTokenizer?", "I can also attest to this error.\r\n\r\nI am using a Kaggle notebook, and I get this error after running this in my first cell. Most of it is default code, bottom two lines are the key ones. \r\n\r\n```\r\n# This Python 3 environment comes with many helpful analytics libraries installed\r\n# It is defined by the kaggle/python docker image: https://github.com/kaggle/docker-python\r\n# For example, here's several helpful packages to load in \r\n\r\nimport numpy as np # linear algebra\r\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\r\n\r\n# Input data files are available in the \"../input/\" directory.\r\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\r\n\r\nimport os\r\nfor dirname, _, filenames in os.walk('/kaggle/input'):\r\n for filename in filenames:\r\n print(os.path.join(dirname, filename))\r\n\r\n# Any results you write to the current directory are saved as output.\r\nprint(os.getcwd(), os.listdir())\r\n\r\nfrom transformers import RobertaTokenizer\r\ntknzr = RobertaTokenizer.from_pretrained('roberta-large')\r\n```\r\n\r\nError thrown\r\n```\r\nUnboundLocalError Traceback (most recent call last)\r\n<ipython-input-1-7957db35f110> in <module>\r\n 19 from transformers import RobertaTokenizer\r\n 20 \r\n---> 21 tknzr = RobertaTokenizer.from_pretrained('roberta-large')\r\n\r\n/opt/conda/lib/python3.6/site-packages/transformers/tokenization_utils.py in from_pretrained(cls, *inputs, **kwargs)\r\n 300 \r\n 301 \"\"\"\r\n--> 302 return cls._from_pretrained(*inputs, **kwargs)\r\n 303 \r\n 304 \r\n\r\n/opt/conda/lib/python3.6/site-packages/transformers/tokenization_utils.py in _from_pretrained(cls, pretrained_model_name_or_path, *init_inputs, **kwargs)\r\n 442 \r\n 443 # Save inputs and kwargs for saving and re-loading with ``save_pretrained``\r\n--> 444 tokenizer.init_inputs = init_inputs\r\n 445 tokenizer.init_kwargs = init_kwargs\r\n 446 \r\n\r\nUnboundLocalError: local variable 'tokenizer' referenced before assignment\r\n```\r\n\r\nKaggle runs transformers version 2.3.0 by default. After updating to 2.5.1 it worked just fine. To update on Kaggle, turn the internet option on in the settings in the right side. Then do `!pip install -U transformers`", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[ { "id": 1834056635, "node_id": "MDU6TGFiZWwxODM0MDU2NjM1", "url": "https://api.github.com/repos/huggingface/transformers/labels/Core:%20Tokenization", "name": "Core: Tokenization", "color": "FF4446", "default": false, "description": "Internals of the library; Tokenization." } ]

[]

[ "**EDIT** : \r\nI've been accidentally testing this code on the 2.4.1 version in a different environment. Since I've updated to 2.5.1 the behavior is as expected." ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3202?src=pr&el=h1) Report\n> Merging [#3202](https://codecov.io/gh/huggingface/transformers/pull/3202?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/5ca356a464e98e065488205f3fcf9247f56c3832?src=pr&el=desc) will **increase** coverage by `<.01%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3202?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3202 +/- ##\n==========================================\n+ Coverage 77.96% 77.97% +<.01% \n==========================================\n Files 98 98 \n Lines 16668 16668 \n==========================================\n+ Hits 12996 12997 +1 \n+ Misses 3672 3671 -1\n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3202?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3202/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `94.72% <0%> (+0.15%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3202?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3202?src=pr&el=footer). Last update [5ca356a...bce6ca3](https://codecov.io/gh/huggingface/transformers/pull/3202?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[]

[ { "id": 1834056761, "node_id": "MDU6TGFiZWwxODM0MDU2NzYx", "url": "https://api.github.com/repos/huggingface/transformers/labels/Core:%20Modeling", "name": "Core: Modeling", "color": "FF8446", "default": false, "description": "Internals of the library; Models." } ]

[ { "login": "julien-c", "id": 326577, "node_id": "MDQ6VXNlcjMyNjU3Nw==", "avatar_url": "https://avatars.githubusercontent.com/u/326577?v=4", "gravatar_id": "", "url": "https://api.github.com/users/julien-c", "html_url": "https://github.com/julien-c", "followers_url": "https://api.github.com/users/julien-c/followers", "following_url": "https://api.github.com/users/julien-c/following{/other_user}", "gists_url": "https://api.github.com/users/julien-c/gists{/gist_id}", "starred_url": "https://api.github.com/users/julien-c/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/julien-c/subscriptions", "organizations_url": "https://api.github.com/users/julien-c/orgs", "repos_url": "https://api.github.com/users/julien-c/repos", "events_url": "https://api.github.com/users/julien-c/events{/privacy}", "received_events_url": "https://api.github.com/users/julien-c/received_events", "type": "User", "site_admin": false }, { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "For some reason there isn't a TF pretrained checkpoint for gpt2-xl [here](https://github.com/huggingface/transformers/blob/9499a3778e1b782f03bc3b15b2ae0cbd20b6391f/src/transformers/modeling_tf_gpt2.py#L39) but there is for Pytorch [here](https://github.com/huggingface/transformers/blob/4134100363e878693aa41f4a25a667ca46d80a9e/src/transformers/modeling_gpt2.py#L35)\r\n\r\nFixing this should only involve converting the pt checkpoint to a tf one. I'd be happy to do it myself if there is a conversion script that can convert Pytorch checkpoints to TF", "Converting a pytorch checkpoint to tf works with \r\n\r\n```python\r\nmodel = GPT2LMHeadModel.from_pretrained('gpt2-xl')\r\nmodel.save_pretrained('./')\r\nmodel = TFGPT2LMHeadModel.from_pretrained('./', from_pt=True)\r\nmodel.save_pretrained('./out')\r\n```\r\n\r\nIf you can tell me where to upload the TF checkpoint to, I'll open up a pull request", "Hi @bkkaggle thanks for pointing this out! @julien-c could you maybe help out here:\r\n\r\nWhile the model: \r\n\r\n\"gpt2-xl\": \"https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-xl-pytorch_model.bin\",\r\n\r\ndoes exist in PyTorch. It does not exist for TF 2. Could we add it as well? " ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3199?src=pr&el=h1) Report\n> Merging [#3199](https://codecov.io/gh/huggingface/transformers/pull/3199?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/49debe62fdc96e161f866dd8914d5915477bb742?src=pr&el=desc) will **increase** coverage by `0.01%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3199?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3199 +/- ##\n=========================================\n+ Coverage 77.98% 78% +0.01% \n=========================================\n Files 98 98 \n Lines 16645 16645 \n=========================================\n+ Hits 12981 12984 +3 \n+ Misses 3664 3661 -3\n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3199?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/file\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3199/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9maWxlX3V0aWxzLnB5) | `68% <0%> (-0.41%)` | :arrow_down: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3199/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `94.4% <0%> (-0.16%)` | :arrow_down: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3199/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.64% <0%> (+0.97%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3199?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3199?src=pr&el=footer). Last update [49debe6...f53348d](https://codecov.io/gh/huggingface/transformers/pull/3199?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "Thanks for sharing. [Model page](https://huggingface.co/twmkn9/albert-base-v2-squad2)" ]

[]

[]

[ "This solved a CUDA runtime error for me. Strange! Thanks for this PR!" ]

[]

[]

[]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3196?src=pr&el=h1) Report\n> Merging [#3196](https://codecov.io/gh/huggingface/transformers/pull/3196?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/3aca02efb3d4ff2d6d231c55d3b9367e61b7c0c4?src=pr&el=desc) will **decrease** coverage by `0.97%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3196?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3196 +/- ##\n==========================================\n- Coverage 77.98% 77.01% -0.98% \n==========================================\n Files 98 98 \n Lines 16660 16660 \n==========================================\n- Hits 12993 12831 -162 \n- Misses 3667 3829 +162\n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3196?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_tf\\_pytorch\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3196/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl9weXRvcmNoX3V0aWxzLnB5) | `8.72% <0%> (-81.21%)` | :arrow_down: |\n| [src/transformers/modeling\\_roberta.py](https://codecov.io/gh/huggingface/transformers/pull/3196/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19yb2JlcnRhLnB5) | `85.71% <0%> (-10%)` | :arrow_down: |\n| [src/transformers/modeling\\_xlnet.py](https://codecov.io/gh/huggingface/transformers/pull/3196/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ194bG5ldC5weQ==) | `73.48% <0%> (-2.3%)` | :arrow_down: |\n| [src/transformers/modeling\\_ctrl.py](https://codecov.io/gh/huggingface/transformers/pull/3196/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19jdHJsLnB5) | `96% <0%> (-2.23%)` | :arrow_down: |\n| [src/transformers/modeling\\_openai.py](https://codecov.io/gh/huggingface/transformers/pull/3196/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19vcGVuYWkucHk=) | `80.2% <0%> (-1.35%)` | :arrow_down: |\n| [src/transformers/modeling\\_transfo\\_xl.py](https://codecov.io/gh/huggingface/transformers/pull/3196/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190cmFuc2ZvX3hsLnB5) | `75.84% <0%> (+0.21%)` | :arrow_up: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3196/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `94.56% <0%> (+0.31%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3196?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3196?src=pr&el=footer). Last update [3aca02e...99b8533](https://codecov.io/gh/huggingface/transformers/pull/3196?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null }, { "id": 1834052847, "node_id": "MDU6TGFiZWwxODM0MDUyODQ3", "url": "https://api.github.com/repos/huggingface/transformers/labels/Ex:%20LM%20(Finetuning)", "name": "Ex: LM (Finetuning)", "color": "26FFF8", "default": false, "description": "Related to language modeling fine-tuning" } ]

[]

[ "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3194?src=pr&el=h1) Report\n> Merging [#3194](https://codecov.io/gh/huggingface/transformers/pull/3194?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/5164ea91a7b4d35cb03867233527fa383a651775?src=pr&el=desc) will **decrease** coverage by `1.09%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3194?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3194 +/- ##\n========================================\n- Coverage 78.09% 77% -1.1% \n========================================\n Files 98 98 \n Lines 16660 16660 \n========================================\n- Hits 13011 12829 -182 \n- Misses 3649 3831 +182\n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3194?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_tf\\_pytorch\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3194/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl9weXRvcmNoX3V0aWxzLnB5) | `8.72% <0%> (-81.21%)` | :arrow_down: |\n| [src/transformers/modeling\\_roberta.py](https://codecov.io/gh/huggingface/transformers/pull/3194/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19yb2JlcnRhLnB5) | `85.71% <0%> (-10%)` | :arrow_down: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3194/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `87.67% <0%> (-3.14%)` | :arrow_down: |\n| [src/transformers/modeling\\_xlnet.py](https://codecov.io/gh/huggingface/transformers/pull/3194/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ194bG5ldC5weQ==) | `73.48% <0%> (-2.3%)` | :arrow_down: |\n| [src/transformers/modeling\\_ctrl.py](https://codecov.io/gh/huggingface/transformers/pull/3194/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19jdHJsLnB5) | `96% <0%> (-2.23%)` | :arrow_down: |\n| [src/transformers/modeling\\_openai.py](https://codecov.io/gh/huggingface/transformers/pull/3194/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19vcGVuYWkucHk=) | `80.2% <0%> (-1.35%)` | :arrow_down: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3194/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `94.4% <0%> (-0.16%)` | :arrow_down: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3194?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3194?src=pr&el=footer). Last update [5164ea9...f3272af](https://codecov.io/gh/huggingface/transformers/pull/3194?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[]

[]

[ "It's in your torch home:\r\n\r\n```py\r\n>>> from torch.hub import _get_torch_home\r\n>>> _get_torch_home()\r\n'/home/<USER>/.cache/torch'\r\n```" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null }, { "id": 1834052847, "node_id": "MDU6TGFiZWwxODM0MDUyODQ3", "url": "https://api.github.com/repos/huggingface/transformers/labels/Ex:%20LM%20(Finetuning)", "name": "Ex: LM (Finetuning)", "color": "26FFF8", "default": false, "description": "Related to language modeling fine-tuning" }, { "id": 1834053007, "node_id": "MDU6TGFiZWwxODM0MDUzMDA3", "url": "https://api.github.com/repos/huggingface/transformers/labels/Ex:%20LM%20(Pretraining)", "name": "Ex: LM (Pretraining)", "color": "76FFAF", "default": false, "description": "Related to language modeling pre-training" } ]

[]

[ "Even I tried to follow the blog and train a LM from scratch but the instructions are ambiguous. Like for ex config file is passed as command line args but if its passed it tries to load it and throws error .", "I've covered some of the parts here: https://zablo.net/blog/post/training-roberta-from-scratch-the-missing-guide-polish-language-model/", "https://stackoverflow.com/questions/61232399/decoding-predictions-for-masked-language-modeling-task-using-custom-bpe\r\n\r\nI posted a question related to this on SO. Any help is appreciated! @marrrcin ", "bump!", "> I've covered some of the parts here: https://zablo.net/blog/post/training-roberta-from-scratch-the-missing-guide-polish-language-model/\r\n\r\nHey Marcin, Your post is very informative. Thanks for that. Could you say a few words on the reasoning for the vocab size being 32000 exactly? Are there any heuristics that helped your decision? (or) anyone here can say a few words on if there are any good heuristics you can follow to choose this hyperparameter? Thanks", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "Approx. how long do these tests take on a single V100?", "> Approx. how long do these tests take on a single V100?\r\n\r\nDon't know how long they take on a V100. On a cpu, all tests combined (7 model tests for PT and 6 model tests for TF) take less than 10min (whereas `test_modeling_tf_xlnet.py`, `test_modeling_xlnet.py` and `test_modeling_transfo_xl.py` combined take ca. 8min)", "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3191?src=pr&el=h1) Report\n> Merging [#3191](https://codecov.io/gh/huggingface/transformers/pull/3191?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/e03129ad447ad7670fcc6206e5eb27a5435d4d86?src=pr&el=desc) will **increase** coverage by `0.01%`.\n> The diff coverage is `100%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3191?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3191 +/- ##\n==========================================\n+ Coverage 78.15% 78.16% +0.01% \n==========================================\n Files 98 98 \n Lines 16641 16641 \n==========================================\n+ Hits 13006 13008 +2 \n+ Misses 3635 3633 -2\n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3191?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_tf\\_xlm.py](https://codecov.io/gh/huggingface/transformers/pull/3191/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl94bG0ucHk=) | `90.4% <100%> (ø)` | :arrow_up: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3191/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `94.56% <0%> (+0.15%)` | :arrow_up: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3191/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `93.89% <0%> (+0.19%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3191?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3191?src=pr&el=footer). Last update [e03129a...9050ffe](https://codecov.io/gh/huggingface/transformers/pull/3191?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "Good to merge for me" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3190?src=pr&el=h1) Report\n> Merging [#3190](https://codecov.io/gh/huggingface/transformers/pull/3190?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/b29fed790bdaa4be38b6d2c5de88e307474ea38d?src=pr&el=desc) will **increase** coverage by `0.1%`.\n> The diff coverage is `42.85%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3190?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3190 +/- ##\n=========================================\n+ Coverage 77.98% 78.09% +0.1% \n=========================================\n Files 98 98 \n Lines 16641 16645 +4 \n=========================================\n+ Hits 12978 12999 +21 \n+ Misses 3663 3646 -17\n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3190?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/configuration\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3190/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9jb25maWd1cmF0aW9uX3V0aWxzLnB5) | `96.74% <100%> (+0.02%)` | :arrow_up: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3190/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `91.38% <33.33%> (+3.59%)` | :arrow_up: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3190/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `94.56% <0%> (-0.16%)` | :arrow_down: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3190?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3190?src=pr&el=footer). Last update [b29fed7...847d370](https://codecov.io/gh/huggingface/transformers/pull/3190?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "Good to merge for me.", "> Small typo bug otherwise it's good to go\r\n\r\nThanks for spotting!" ]

[]

[]

[ "There are a lot of examples in [the documentation](https://huggingface.co/transformers/main_classes/model.html#transformers.PreTrainedModel)." ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[ { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "@Laksh1997 thanks a lot for reporting this error. Can you provide a code snippet and maybe a link to your data to easily reproduce this error? \r\n\r\nIn the meantime: \r\n\r\n- There has been a lot of changes recently in the beam search decoding -> I would recommend using the master branch of beam search decoding! \r\n- Beam search is not really made for top_p_top_k sampling, when using beam search I recommend setting do_sample=False", "Right, I'll try the master branch and inform any further problems.\r\n\r\nFor generation of samples (without any context or input), there is no point of having sample set to False, as one will always generate the same sample as a result.", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n", "I'm running into this issue somewhat commonly as well.\r\n\r\n- Python 3.7\r\n- transformers 3.0.2\r\n- torch 1.5.1\r\n- macOS 10.15\r\n- Running on CPU\r\n\r\nCode to reproduce error:\r\n```python\r\nimport torch\r\nfrom transformers import MarianMTModel, MarianTokenizer\r\n\r\ntorch.manual_seed(15)\r\nphrase = \"Ich verstehe nicht, was du sagen willst. Sprich doch Deutsch!\"\r\nmodel = MarianMTModel.from_pretrained(\"Helsinki-NLP/opus-mt-de-ZH\")\r\ntokenizer = MarianTokenizer.from_pretrained(\"Helsinki-NLP/opus-mt-de-ZH\")\r\n\r\n# Nucleus sampling as per https://github.com/huggingface/blog/blob/master/notebooks/02_how_to_generate.ipynb\r\ninput_ids = tokenizer.prepare_translation_batch([phrase])\r\ntoken_ids_p = model.generate(\r\n **input_ids,\r\n do_sample=True,\r\n top_p=0.9,\r\n)\r\n\r\ntranslated_p = [tokenizer.decode(string, skip_special_tokens=True) for string in token_ids_p]\r\nprint(translated_p)\r\n```\r\n\r\nError:\r\n```\r\nTraceback (most recent call last):\r\n File \"temp.py\", line 14, in <module>\r\n top_p=0.9,\r\n File \"/Users/kaz/envs/venv-3.7/lib/python3.7/site-packages/torch/autograd/grad_mode.py\", line 15, in decorate_context\r\n return func(*args, **kwargs)\r\n File \"/Users/kaz/envs/venv-3.7/lib/python3.7/site-packages/transformers/generation_utils.py\", line 459, in generate\r\n model_specific_kwargs=model_specific_kwargs,\r\n File \"/Users/kaz/envs/venv-3.7/lib/python3.7/site-packages/transformers/generation_utils.py\", line 757, in _generate_beam_search\r\n assert len(next_sent_beam) == num_beams, \"Beam should always be full\"\r\n```\r\n@patrickvonplaten Is it possible to revive this issue?" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null }, { "id": 1834052333, "node_id": "MDU6TGFiZWwxODM0MDUyMzMz", "url": "https://api.github.com/repos/huggingface/transformers/labels/Ex:%20Question%20Answering", "name": "Ex: Question Answering", "color": "86FFCF", "default": false, "description": "" }, { "id": 1838876023, "node_id": "MDU6TGFiZWwxODM4ODc2MDIz", "url": "https://api.github.com/repos/huggingface/transformers/labels/Distillation", "name": "Distillation", "color": "d4c5f9", "default": false, "description": "Related to model distillation" } ]

[]

[ "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3186?src=pr&el=h1) Report\n> Merging [#3186](https://codecov.io/gh/huggingface/transformers/pull/3186?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/e03129ad447ad7670fcc6206e5eb27a5435d4d86&el=desc) will **decrease** coverage by `0.50%`.\n> The diff coverage is `30.76%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3186?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3186 +/- ##\n==========================================\n- Coverage 78.15% 77.64% -0.51% \n==========================================\n Files 98 99 +1 \n Lines 16641 16795 +154 \n==========================================\n+ Hits 13006 13041 +35 \n- Misses 3635 3754 +119 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3186?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3186/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `91.34% <15.38%> (-3.07%)` | :arrow_down: |\n| [src/transformers/benchmark\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3186/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9iZW5jaG1hcmtfdXRpbHMucHk=) | `31.74% <31.74%> (ø)` | |\n| [src/transformers/\\_\\_init\\_\\_.py](https://codecov.io/gh/huggingface/transformers/pull/3186/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9fX2luaXRfXy5weQ==) | `98.92% <100.00%> (+0.01%)` | :arrow_up: |\n| [src/transformers/configuration\\_gpt2.py](https://codecov.io/gh/huggingface/transformers/pull/3186/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9jb25maWd1cmF0aW9uX2dwdDIucHk=) | `97.29% <100.00%> (+0.07%)` | :arrow_up: |\n| [src/transformers/modeling\\_gpt2.py](https://codecov.io/gh/huggingface/transformers/pull/3186/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19ncHQyLnB5) | `86.07% <100.00%> (ø)` | |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3186/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `91.53% <0.00%> (-2.17%)` | :arrow_down: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3186?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3186?src=pr&el=footer). Last update [e03129a...cb67ca6](https://codecov.io/gh/huggingface/transformers/pull/3186?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3185?src=pr&el=h1) Report\n> Merging [#3185](https://codecov.io/gh/huggingface/transformers/pull/3185?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/7420a6a9cc1750c2bd2c2c245d00048ec36d3bf0?src=pr&el=desc) will **decrease** coverage by `0.24%`.\n> The diff coverage is `80.42%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3185?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3185 +/- ##\n==========================================\n- Coverage 77.79% 77.55% -0.25% \n==========================================\n Files 100 100 \n Lines 17025 17105 +80 \n==========================================\n+ Hits 13245 13265 +20 \n- Misses 3780 3840 +60\n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3185?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/tokenization\\_bert.py](https://codecov.io/gh/huggingface/transformers/pull/3185/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy90b2tlbml6YXRpb25fYmVydC5weQ==) | `95.33% <ø> (-1.7%)` | :arrow_down: |\n| [src/transformers/tokenization\\_roberta.py](https://codecov.io/gh/huggingface/transformers/pull/3185/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy90b2tlbml6YXRpb25fcm9iZXJ0YS5weQ==) | `94.36% <100%> (-5.64%)` | :arrow_down: |\n| [src/transformers/pipelines.py](https://codecov.io/gh/huggingface/transformers/pull/3185/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9waXBlbGluZXMucHk=) | `74.51% <100%> (-0.28%)` | :arrow_down: |\n| [src/transformers/tokenization\\_transfo\\_xl.py](https://codecov.io/gh/huggingface/transformers/pull/3185/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy90b2tlbml6YXRpb25fdHJhbnNmb194bC5weQ==) | `40.67% <100%> (-0.43%)` | :arrow_down: |\n| [src/transformers/tokenization\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3185/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy90b2tlbml6YXRpb25fdXRpbHMucHk=) | `86.18% <78.61%> (-5.81%)` | :arrow_down: |\n| ... and [1 more](https://codecov.io/gh/huggingface/transformers/pull/3185/diff?src=pr&el=tree-more) | |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3185?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3185?src=pr&el=footer). Last update [7420a6a...860cf66](https://codecov.io/gh/huggingface/transformers/pull/3185?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "Really like the new typings!" ]

[]

[]

[ "`tokenizers` is only required when you wish to use the new, fast [tokenizers](https://github.com/huggingface/tokenizers). By default, though, the standard (slower) tokenizers are used. So you do not actually need the `tokenizers` library to run the `transformers` library.\r\n\r\nRelated:\r\nhttps://github.com/huggingface/transformers/issues/2980\r\nhttps://github.com/huggingface/transformers/issues/2831\r\n", "Can you please open an issue over at https://github.com/huggingface/tokenizers with your OS/Python details?", "while I am installing transformer==4.4, I am facing issue: Failed to build tokenizers.\r\n" ]

[]

[]

[ "Indeed, I think this is a mistake, I think it should be `bert-base-uncased` instead. I'm updating it." ]

[]

[]

[ "There are many examples in [the docs](https://huggingface.co/transformers/usage.html). The [pipeline reference](https://huggingface.co/transformers/main_classes/pipelines.html) will also be helpful." ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null }, { "id": 1834056761, "node_id": "MDU6TGFiZWwxODM0MDU2NzYx", "url": "https://api.github.com/repos/huggingface/transformers/labels/Core:%20Modeling", "name": "Core: Modeling", "color": "FF8446", "default": false, "description": "Internals of the library; Models." } ]

[ { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "Thanks for the bug report @xunzi2020! Did you by any chance measure the degree results would differ when replacing `-1e4` by `-1e10`? ", "Yes, I trained with chinese text (~30GB), it happened.\n\n\nPatrick von Platen <[email protected]> 于2020年3月9日周一 下午4:11写道：\n\n> Thanks for the bug report @xunzi2020 <https://github.com/xunzi2020>! Did\n> you by any chance measure the degree results would differ when replacing\n> -1e4 by -1e10?\n>\n> —\n> You are receiving this because you were mentioned.\n> Reply to this email directly, view it on GitHub\n> <https://github.com/huggingface/transformers/issues/3181?email_source=notifications&email_token=AOYV777JVZ6HL4BNOKYY7HDRGSQEXA5CNFSM4LEAB3C2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEOGCLSI#issuecomment-596387273>,\n> or unsubscribe\n> <https://github.com/notifications/unsubscribe-auth/AOYV777AZV3K5GBW4TFVFFDRGSQEXANCNFSM4LEAB3CQ>\n> .\n>\n", "Do you have a comparison such as: \r\n\r\n| - | GPT2 with -1e4 vs. GPT2 with -1e10 | \r\n| ------------- | ------------- | \r\n| abs mean(softmax(logits) - softmax(logits)) | ???? |\r\n| relative mean(softmax(logits) - softmax(logits)/softmax(logits)) | ???? |\r\n\r\nlets say averaged over 100 - 1000 input samples and all GPT2 logits (50256)? \r\n\r\nThat would be great to quantify the impact this change would have.\r\n", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "This looks great to me. Thanks @shubhamagarwal92. \r\n\r\nYou need to just run the black command, I believe to is `make style` to fix formatting. Otherwise it looks good.", "@srush I ran the `make style` which changed a lot of files, however, the `check_code_quality` is still failing! \r\nDo you want me to revert the last two commits and manually change the `'logs'` to `\"logs\"`\r\n\r\nBTW, `pip install -e \".[dev]\"` is failing on both mac and linux for `tensorflow` and on `sentencepiece` for mac. I had to manually install the `[\"black\", \"isort\", \"flake8\"]` packages. Python=3.8.2 in conda env. ", "> BTW, pip install -e \".[dev]\" is failing on both mac and linux for tensorflow and on sentencepiece for mac.\r\n\r\nThat would be because [TensorFlow only supports python 3.5-3.7](https://www.tensorflow.org/install/pip?lang=python3#system-requirements), unfortunately.", "> > BTW, pip install -e \".[dev]\" is failing on both mac and linux for tensorflow and on sentencepiece for mac.\r\n> \r\n> That would be because [TensorFlow only supports python 3.5-3.7](https://www.tensorflow.org/install/pip?lang=python3#system-requirements), unfortunately.\r\n\r\n@LysandreJik Thanks. Installs on ubuntu with python 3.6. \r\n\r\nHowever, on mac: \r\n```\r\nconda create -n transformers_dev python=3.6 -y\r\nconda activate transformers_dev\r\npip install -e \".[dev]\"\r\n\r\n\r\nFailed to build tokenizers\r\nERROR: Could not build wheels for tokenizers which use PEP 517 and cannot be installed directly\r\n```\r\n\r\nMac specs:\r\n```\r\nPython 3.6.10 | packaged by conda-forge | (default, Mar 5 2020, 09:56:10) \r\n[GCC Clang 9.0.1 ] on darwin\r\n```", "Hi @shubhamagarwal92 ,\r\n\r\nthanks for that PR and fixing the issues :+1: \r\n\r\nI just ran the `run_pl.sh` script, training and final testing run without errors now. However, during training precision, recall and F1 are always 0 -> final output on the test set shows:\r\n\r\n```bash\r\nTEST RESULTS\r\n{'val_loss': tensor(7.0679), 'precision': 0.0, 'recall': 0.0, 'f1': 0}\r\n----------------------------------------------------------------------------------------------------\r\nTesting: 200it [00:07, 27.98it/s]\r\n```\r\n\r\nLast lines of the prediction output:\r\n\r\n```bash\r\nder I-OTHderiv\r\nBibliothek I-OTHderiv\r\nberufen I-OTHderiv\r\nwurde I-OTHderiv\r\n, I-OTHderiv\r\nverließ I-OTHderiv\r\nGardthausen I-OTHderiv\r\nden I-OTHderiv\r\nBibliotheksdienst I-OTHderiv\r\n. I-OTHderiv\r\n```\r\n", "> I just ran the `run_pl.sh` script, training and final testing run without errors now. However, during training precision, recall and F1 are always 0 -> final output on the test set shows:\r\n> \r\n> ```shell\r\n> TEST RESULTS\r\n> {'val_loss': tensor(7.0679), 'precision': 0.0, 'recall': 0.0, 'f1': 0}\r\n> ----------------------------------------------------------------------------------------------------\r\n> Testing: 200it [00:07, 27.98it/s]\r\n> ```\r\n\r\nThanks for reporting this. Could you please verify the version of `pytorch-lightning`. For me it is\r\n`pytorch-lightning==0.7.1`, `transformers==2.5.1` and the results as reported in the [README](https://github.com/huggingface/transformers/pull/3180/commits/ed39624dd0d0f3bce55352a8c4c9a8f515793e29#diff-eb7fd389de7be266012669eab7db207bR119).\r\n\r\n\r\nAlso could you please check if the results in `${OUTPUT_DIR}/test_results.txt` mentioned [here](https://github.com/huggingface/transformers/pull/3180/commits/84ee92d47ee6d659edaf6a61d09b393ebeea4d5b#diff-5a6311e9856e7b0057d9c1b85cd85fadR27)\r\nalso correspond to 0. \r\n\r\nIt works for me as:\r\n\r\n\r\n", "Hi,\r\n\r\nI'm using the same versions of both `pytorch-lightning` and `transformers` 😂\r\n\r\nOutput of `test_results` is:\r\n\r\n```bash\r\n$ cat germeval-model/test_results.txt \r\nf1 = 0\r\nprecision = 0.0\r\nrecall = 0.0\r\nval_loss = tensor(9.4173)\r\n```\r\n\r\nBut I'm going to test it on another machine :)", "\r\n> But I'm going to test it on another machine :)\r\n\r\nI am also attaching my environment file via `pip freeze > requirements.txt`:\r\n[requirements.txt](https://github.com/huggingface/transformers/files/4307693/requirements.txt)\r\n\r\nPlease let me know if this doesn't work. ", "@shubhamagarwal92 I think somehow you have the wrong version of our style checks installed. \r\n\r\nCan you try running under this command?\r\n```\r\nsudo pip install git+git://github.com/timothycrosley/isort.git@e63ae06ec7d70b06df9e528357650281a3d3ec22#egg=isort\r\nsudo pip install .[tf,torch,quality]\r\n```\r\n\r\n@LysandreJik we have to fix this, it is really confusing...\r\n\r\n@stefan-it would love to see your log as well. Could you also try `rm cached*` I think maybe your feature cache got messed up?\r\n\r\n\r\n\r\n", "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3180?src=pr&el=h1) Report\n> Merging [#3180](https://codecov.io/gh/huggingface/transformers/pull/3180?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/e03129ad447ad7670fcc6206e5eb27a5435d4d86?src=pr&el=desc) will **increase** coverage by `0.01%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3180?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3180 +/- ##\n==========================================\n+ Coverage 78.15% 78.16% +0.01% \n==========================================\n Files 98 98 \n Lines 16641 16641 \n==========================================\n+ Hits 13006 13008 +2 \n+ Misses 3635 3633 -2\n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3180?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3180/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `94.72% <0%> (+0.31%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3180?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3180?src=pr&el=footer). Last update [e03129a...9f949d3](https://codecov.io/gh/huggingface/transformers/pull/3180?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "> ```\r\n> sudo pip install git+git://github.com/timothycrosley/isort.git@e63ae06ec7d70b06df9e528357650281a3d3ec22#egg=isort\r\n> sudo pip install .[tf,torch,quality]\r\n> ```\r\n\r\n@srush I reverted the last 3 style related commits, force-pushed and added a small commit to pass all the checks. Please merge if everything is fine. \r\n\r\nAlso, for `isort`, I guess you meant using `git+https://` \r\n```\r\npip install git+https://github.com/timothycrosley/isort.git@e63ae06ec7d70b06df9e528357650281a3d3ec22#egg=isort\r\n```\r\nThis link is also wrong at [contributing.md](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md#start-contributing-pull-requests). Could you please also state the python version in the md. \r\n\r\nThis command `pip install .[tf,torch,quality]` is still failing on Mac as mentioned in my previous comment.\r\n\r\n```\r\nFailed to build tokenizers\r\nERROR: Could not build wheels for tokenizers which use PEP 517 and cannot be installed directly\r\n```", "Thanks @shubhamagarwal92. Sorry for the annoyance.\r\n\r\n@LysandreJik this lgtm. ", "@srush Happy to help! :) \r\n\r\nThanks for approving the PR!", "@shashwath94 I think I've found the reason for the bad evaluation results: I'm using apex and the `--fp16` parameter in the `run_pl.sh` script!\r\n\r\nDo you have any idea, why it is not working using half precision 🤔", "Ah, I will check with pytorch-lightning. It is plausible we are not integrating correctly with them ", "@srush While you are it, could you please check the status of my PR in pl as well. \r\n\r\nhttps://github.com/PyTorchLightning/pytorch-lightning/pull/1094\r\n\r\nBasically, I was observing memory leak on GPU0 if other GPU id (eg. [1]) was provided when running the NER example.\r\n\r\nAFAIK, the solution is `torch.cuda.set_device()` " ]

[]

[]

[ "I have solved it by rebooting my computer and reinstall some modules that transformers can not found but have been installed", "This issue still exists. I have Tensorflow 2.1 with GPU and transformers version 2.8.0\r\n\r\nI tried with rebooting and creating new conda environments multiple times but still no success :(", "I can import it and my python version is 3.7.6 , 2.8.0 for transformers and 2.1 for tf. Maybe you could try to upgrade your python if its' version is under 3.7.", "> This issue still exists. I have Tensorflow 2.1 with GPU and transformers version 2.8.0\r\n> \r\n> I tried with rebooting and creating new conda environments multiple times but still no success :(\r\n\r\nI can import it and my python version is 3.7.6 , 2.8.0 for transformers and 2.1 for tf. Maybe you could try to upgrade your python if its' version is under 3.7.", "Hmm, the next day I tried it again with the same environment and it works xD Probably because of rebooting or something. But I also rebooted the first day, idk..." ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[ { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "Hi @BlackJack01 - thanks so much for this great contribution! We are still going through the notebook and discussing how to add it :-) Will let you know soon! ", "Hi @BlackJack01, sorry for the late answer. \r\nWe now have community notebooks here: https://github.com/huggingface/transformers/tree/master/notebooks#community-notebooks\r\nFeel free to open a PR to add it there :-) " ]

[]

[]

[ "hi @shoarora \r\n\r\ncan you share the script to to report performance on test set?", "@Mahmedturk you can check out the branch in PR #3405 \r\n\r\nIt's diverge pretty heavily from master and I haven't updated it yet, but you should still be able to run `run_glue.py` off that branch with the `--do_test` flag that I added and it should produce the `.tsv` files required for submission. ", "@shoarora I pulled the repo to update run_glue.py as I wanted to use this new feature. However, I now get an error when I run run_glue.py! Please see below the output of the error message. It looks like in previous versions, there weren't any keyword arguments named \"mode\" in GlueDataset() --possible? \r\n\r\n`Traceback (most recent call last):\r\n File \"./transformers/examples/text-classification/run_glue.py\", line 228, in <module>\r\n main()\r\n File \"./transformers/examples/text-classification/run_glue.py\", line 139, in main\r\n test_dataset = GlueDataset(data_args, tokenizer=tokenizer, mode=\"test\") if training_args.do_predict else None\r\nTypeError: __init__() got an unexpected keyword argument 'mode'`", "@AMChierici I didn't author #4463, which is what has made it to master to enable this feature. I haven't played with it yet so sorry I can't be of more help", "@AMChierici make sure you run from master, there's indeed a `mode` kwarg now.\r\n\r\n@shoarora Thanks for this first PR and I did check yours while merging the other (to make sure that the indices in csv parsing, etc. were correct)", "Thanks, @julien-c . Yes, solved.. In fact, I was not running from master.\r\n", "downloaded master right now.\r\n File \"examples/text-classification/run_glue.py\", line 143, in main\r\n if training_args.do_eval\r\nTypeError: __init__() got an unexpected keyword argument 'mode'" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3175?src=pr&el=h1) Report\n> Merging [#3175](https://codecov.io/gh/huggingface/transformers/pull/3175?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/e03129ad447ad7670fcc6206e5eb27a5435d4d86?src=pr&el=desc) will **not change** coverage.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3175?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3175 +/- ##\n=======================================\n Coverage 78.15% 78.15% \n=======================================\n Files 98 98 \n Lines 16641 16641 \n=======================================\n Hits 13006 13006 \n Misses 3635 3635\n```\n\n\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3175?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3175?src=pr&el=footer). Last update [e03129a...70d11c4](https://codecov.io/gh/huggingface/transformers/pull/3175?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "This might answer your question: https://stackoverflow.com/questions/39649102/how-do-i-select-which-gpu-to-run-a-job-on", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "Its a typo in the docs. I just used cnn.", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "Our models are TensorFlow 2.x only, while your notebook is in TensorFlow 1.x:\r\n\r\n```\r\nThe default version of TensorFlow in Colab will soon switch to TensorFlow 2.x.\r\nWe recommend you upgrade now or ensure your notebook will continue to use TensorFlow 1.x via the %tensorflow_version 1.x magic: more info.\r\n```\r\n\r\nYou can use the following command at the start of your notebook to use TensorFlow 2.x:\r\n\r\n```\r\n%tensorflow_version 2.x\r\n```", "@LysandreJik thank you." ]

[]

[]

[ "Yes, there are a few available on the hub, you can search for [whole word masking](https://huggingface.co/models?search=whole-word-masking) or [wwm](https://huggingface.co/models?search=wwm).\r\n\r\nFor your second question, you would have a better answer if you opened an issue on [huggingface/toeknizers](https://github.com/huggingface/tokenizers/issues) instead." ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "You might be interested in:\r\n- https://huggingface.co/huggingface/CodeBERTa-small-v1#codeberta and https://huggingface.co/huggingface/CodeBERTa-language-id\r\n- more generally, https://huggingface.co/blog/how-to-train", "Wow, it helped a lot @julien-c. Basically, did you train a language model using CodeSearchNet dataset?", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n", "@julien-c,\r\n\r\nIs there an approach where this tutorial https://huggingface.co/blog/how-to-train is trained over TPU cores?", "maybe @LysandreJik or @sgugger have a link to a notebook?", "I haven't tried training in notebooks on TPU, only with the example scripts.", "Unfortunately I only have notebooks that run the example scripts on TPU, nothing similar to the `how-to-train` blogpost.", "Thanks @julien-c, @sgugger and @LysandreJik,\r\n\r\nMaybe I can adapt the [Language Modeling example script](https://github.com/huggingface/transformers/tree/master/examples/language-modeling) applying [Pytorch Lightning](https://www.pytorchlightning.ai/) approach which easily support TPUs.", "If you're using the language modeling script, running it on TPUs is supported, just follow the instructions [here](https://github.com/huggingface/transformers/tree/master/examples#running-on-tpus).", "> If you're using the language modeling script, running it on TPUs is supported, just follow the instructions [here](https://github.com/huggingface/transformers/tree/master/examples#running-on-tpus).\r\n\r\nGreat, but it runs on Google Colab?", "Yes it does run on colab!", "You can find an example running the `run_glue.py` script [here](https://colab.research.google.com/drive/15q6UUzwNugWvVXNfOkWlGCaKIUvLZpxd?usp=sharing). You can do the same with the language modeling script! (Cloning the repository and running the script from there would be cleaner than `wget`ting all the files like it's done in this colab, though)", "> You can find an example running the `run_glue.py` script [here](https://colab.research.google.com/drive/15q6UUzwNugWvVXNfOkWlGCaKIUvLZpxd?usp=sharing). You can do the same with the language modeling script! (Cloning the repository and running the script from there would be cleaner than `wget`ting all the files like it's done in this colab, though)\r\n\r\nFantastic! That was of great help.", "@LysandreJik, @sgugger,\r\n\r\nUnfortunately, even using a small dataset (~400MB), the Colab killed the process due to the use of all available RAM (12.72GB)." ]

[]

[]

[ "That's more of a general Python question related to imports than an issue with the library, you would have more luck in trying with stack overflow.\r\n\r\nI believe the easiest way to modify files is to clone the repository and install it in your environment as an editable:\r\n\r\n```\r\ngit clone https://github.com/huggingface/transformers\r\ncd transformers\r\npip install -e .\r\n```\r\n\r\nEvery file modification will be directly reflected in your Python runtime (if you're on Jupyter you would need to restart your kernel for it to take effect).", "Thanks a lot. \r\nActually the problem was that I was appending the path (i.e src/transformers) at the end in PYTHON_PATH because of which it was loading modules from the transformers library. I added the source path at index 0 and now it is working the way I want it to work.\r\n\r\nand thanks @LysandreJik for the restart kernel trick. \r\n" ]

[ { "id": 1260952223, "node_id": "MDU6TGFiZWwxMjYwOTUyMjIz", "url": "https://api.github.com/repos/huggingface/transformers/labels/Discussion", "name": "Discussion", "color": "22870e", "default": false, "description": "Discussion on a topic (keep it focused or open a new issue though)" }, { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "GPT-2 and BERT are both transformer networks with very similar architectures. You can use the GPT-2 embeddings the same way you used BERT embeddings.\r\n\r\nAs you said, GPT-2 only handles left context. You can read [the paper](https://cdn.openai.com/better-language-models/language_models_are_unsupervised_multitask_learners.pdf) where the authors showcase results on several tasks in a zero-shot setting (section 3).", "I recently imported the GPT2Model and built a simple classifier. I think the model is too naive. And could use some improvements. If you notice any mistakes, please correct me. :) \r\n```\r\nclass SimpleGPT2SequenceClassifier(nn.Module):\r\n def __init__(self, hidden_size: int, num_classes:int ,max_seq_len:int, gpt_model_name:str, \r\n cache_dir:str):\r\n super(SimpleGPT2SequenceClassifier,self).__init__()\r\n self.gpt2model = GPT2Model.from_pretrained(\r\n gpt_model_name, cache_dir = cache_dir\r\n )\r\n self.fc1 = nn.Linear(hidden_size, num_classes)\r\n \r\n def forward(self, x_in):\r\n \"\"\"\r\n Args:\r\n x_in: encoded inputs ids of sent.\r\n \"\"\"\r\n \r\n gpt_out = self.gpt2model(x_in)[0] #returns tuple\r\n batch_size = gpt_out.shape[0]\r\n prediction_vector = self.fc1(gpt_out.view(batch_size,-1)) #(batch_size , max_len, num_classes)\r\n \r\n return prediction_vector\r\n```\r\nFor preprocessing the text before encoding them with the tokenizer.\r\n\r\n```\r\npunkt_sentence_detector = nltk.data.load('tokenizers/punkt/english.pickle')\r\nclass GPT2Preprocessor:\r\n def __init__(self, transformer_tokenizer, sentence_detector):\r\n self.transformer_tokenizer = transformer_tokenizer\r\n self.sentence_detector = sentence_detector\r\n\r\n def add_eos_tokens(self, text):\r\n eos_token = \" \" + self.transformer_tokenizer.eos_token + \" \"\r\n sentences = self.sentence_detector.tokenize(text)\r\n eos_added_text = (\r\n eos_token.join(sentences) + \" \" + self.transformer_tokenizer.eos_token\r\n )\r\n return eos_added_text\r\n```", "I tried GPT-2 embeddings and compare them with Roberta embeddings for the task of sentiment classification (both networks were frozen during the training). GPT-2 couldn't outperform the results of Roberta.", "@cozek from the code, it isn't obvious whether you've frozen gpt2 layers or not ?", "> @cozek from the code, it isn't obvious whether you've frozen gpt2 layers or not ?\r\n\r\nOf course, I have not frozen any layers. It is not always necessary to freeze the layers. If required you can easily freeze the layers as necessary. ", "> I tried GPT-2 embeddings and compare them with Roberta embeddings for the task of sentiment classification (both networks were frozen during the training). GPT-2 couldn't outperform the results of Roberta.\r\n\r\nDo you still have the notebooks? I would be interested to see how you implemented a classification head on top of gpt-2. ", "> > I tried GPT-2 embeddings and compare them with Roberta embeddings for the task of sentiment classification (both networks were frozen during the training). GPT-2 couldn't outperform the results of Roberta.\r\n> \r\n> Do you still have the notebooks? I would be interested to see how you implemented a classification head on top of gpt-2.\r\n\r\nhttps://github.com/cozek/OffensEval2020-code/blob/master/notebooks/Eng%20Task%20A%20-%20Ensemble%20DistilGPT2.ipynb\r\n\r\nHere you go. I used it for OffenEval 2020, Hate Speech Detection. I used the distilled version. Feel free to swap it out and take the full GPT-2. We got 0.90 Macro f1 with this model. ", "You can add a CLS token to the vocabulary \r\n\r\n`tokenizer.add_special_tokens({'cls_token': '[CLS]'})\r\n model.resize_token_embeddings(len(tokenizer))`\r\n\r\nThen append this CLS token at the end of your input \r\nThen use the representation of this CLS token for classification as done in BERT.\r\ncc @cozek ", "> > > I tried GPT-2 embeddings and compare them with Roberta embeddings for the task of sentiment classification (both networks were frozen during the training). GPT-2 couldn't outperform the results of Roberta.\r\n> > \r\n> > \r\n> > Do you still have the notebooks? I would be interested to see how you implemented a classification head on top of gpt-2.\r\n> \r\n> https://github.com/cozek/OffensEval2020-code/blob/master/notebooks/Eng%20Task%20A%20-%20Ensemble%20DistilGPT2.ipynb\r\n> \r\n> Here you go. I used it for OffenEval 2020, Hate Speech Detection. I used the distilled version. Feel free to swap it out and take the full GPT-2. We got 0.90 Macro f1 with this model.\r\n\r\nThanks a lot. Very helpful! ", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n", "@cozek I see in your code that you concatenate all the token embeddings together to produce the sentence representation and then pass that through `fc1`:\r\n\r\n\r\n```python\r\n gpt_out = self.gpt2model(x_in)[0] #returns tuple\r\n batch_size = gpt_out.shape[0]\r\n prediction_vector = self.fc1(gpt_out.view(batch_size,-1))\r\n```\r\n\r\nInstead of concatenating all the token embeddings, did you try:\r\n\r\n1. pooling over all the tokens to get the sentence representation? For example, max pooling or mean pooling?\r\n2. using the embedding of the last token?\r\n\r\n@AsmirMumin ", "> @cozek I see in your code that you concatenate all the token embeddings together to produce the sentence representation and then pass that through `fc1`:\r\n> \r\n> ```python\r\n> gpt_out = self.gpt2model(x_in)[0] #returns tuple\r\n> batch_size = gpt_out.shape[0]\r\n> prediction_vector = self.fc1(gpt_out.view(batch_size,-1))\r\n> ```\r\n> \r\n> Instead of concatenating all the token embeddings, did you try:\r\n> \r\n> 1. pooling over all the tokens to get the sentence representation? For example, max pooling or mean pooling?\r\n> 2. using the embedding of the last token?\r\n\r\nI did not try 1 or 2. Option 1 seems logical as it would reduce the size of the FC layer and increase training speed. \r\nI am not familiar with option 2.\r\n\r\n" ]

[]

[ { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "Hi mainulquraishi, \r\n\r\nAs mentioned in earlier issues #3031, #3021, #2975, #3069 it is not advised to use GPT2LMHeadModel in inference using a padded batch. To answer your two questions:\r\n\r\n1. You don't get the same output in your second code snippet as in your first code snippet because you \"argmax\" the next token from the logits corresponding to the masked token on position 3. \r\nYou would get the same output 'a' if you would take the \"argmax\" the next token from the **last non-padded token** which is at postion 2 in your example. \r\n\r\n2. The attention mask works as far as I can see. Using an attention mask means that logits at **other** positions than the masked position input are not influenced by the masked position input.\r\nThis means that if you mask position 3 you will see that changing the input for position 3 will not change the output for postion 4 - N, but changing the input for position 3 will surely influence the output of position 3 (A token cannot mask its own output).\r\n\r\nI would advise you to take a good look at Issue #3021 ", "Hi @patrickvonplaten I ran into the same issue that you described properly in your first point.\r\nSome questions for the record:\r\na) Could you please describe why position_ids argument is not required here? It's not clear for me why it was needed in https://github.com/huggingface/transformers/issues/3021 and not here.\r\nb) Any padded batch will likely have sentences with many different lengths. Is there a way that `GPT2LMHeadModel` is able to identify the last non-padded token for each sentence (maybe via the attention-mask) so we get the corresponding logits easily? Any function to do that filtering? If not, I guess we can do it from the client side via some tensor operation to discard the last padded tokens (we can infer the last padded tokens via the attention mask). Is this correct?\r\nc) Could we apply what we are discussing here in terms of padding to run the model with Torchscript? Any advice / warning here?", "The answer for (b) can be found in the code snippet that Patrick added in https://github.com/huggingface/transformers/issues/3021 . The following does the trick:\r\n```\r\nlast_non_masked_idx = torch.sum(attention_mask, dim=1) - 1\r\nstart_idx = inp_idx = (last_non_masked_idx).view(-1, 1).repeat(1, st.tokenizer.vocab_size).unsqueeze(1)\r\nlogits = logits.gather(1, start_idx).squeeze(1)\r\n```\r\n\r\n", "You just saved me @Damiox! Thank you so much :)" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n", "I have the same issue.. anyone has any inputs on this " ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "Depending on which version you are using, it might be that you are using the fast [tokenizers ](https://github.com/huggingface/tokenizers) library which offers a much improved tokenizer interface built on Rust. ", "Thanks for the reply. I don't think it is because of the tokenizer. I measured the encoder part and see huge improvement in the encoder speed. Still cannot figure out the difference. comparing pytorch-transformers 1.1 with transformers 2.5.1.", "It is unlikely that there have been architecture changes, but it might be caused by a different torch/tensorflow version? Are you testing both transformers versions on the same framework version? It is likely that this is caused by an optimisation of an activation function. ", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "Where did you get 32 from?\r\n\r\nThe [official file ](https://tfhub.dev/google/albert_xlarge/2)says 16.", "Well, the model configuration in tar file downloaded from TF Hub shows 32, which conflicts with the official definition.", "Indeed, that's right. I'll follow the issue you opened https://github.com/google-research/ALBERT/issues/180 and act accordingly.", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "https://github.com/huggingface/transformers/issues/2311", "@frankniujc it is helpful\r\nbut maybe a better way is take the all tokens in a whole, not prediction the next tokens", "The probability of a sentence P(s0s1s2s3s4...sn) = P(s1|s0) * P(s2|s0s1) * P(s3|s0s1s2) * ... * P(sn|s0s1s2...sn-1)\r\n\r\nSo you can do something like this\r\n```Python\r\ndef sentence_probability(sent):\r\n bos = tokenizer.encode('<|endoftext|>')\r\n tokens = tokenizer.encode(sent)\r\n tokens = bos + tokens\r\n input_ids = torch.tensor(tokens).unsqueeze(0).to('cuda')\r\n\r\n sent_probs = []\r\n\r\n for i, next_word in enumerate(tokens[1:]):\r\n next_word_logits = model(input_ids[:,:i+1])[0][0, -1].detach()\r\n next_word_prob = F.log_softmax(next_word_logits, dim=0)[next_word].item()\r\n\r\n sent_probs.append(next_word_prob)\r\n\r\n return sum(sent_probs)\r\n```", "@loveJasmine Have a look at [`lm-scorer`](https://github.com/simonepri/lm-scorer).\r\n\r\nIt is a tiny wrapper around `transformers` I wrote that allows you to get sentences probabilities using models that support it (only GPT2 models are implemented at the time of writing).", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "The conversion script should pretty much work out of the box, so feel free to do it and we welcome a PR (and we can upload the converted weights to our S3)", "which script you mean? sorry i did not get it. i think this is not possible\nto train it on huggingface is this?\n\nOn Fri, Mar 6, 2020, 10:12 PM Julien Chaumond <[email protected]>\nwrote:\n\n> The conversion script should pretty much work out of the box, so feel free\n> to do it and we welcome a PR (and we can upload the converted weights to\n> our S3)\n>\n> —\n> You are receiving this because you authored the thread.\n> Reply to this email directly, view it on GitHub\n> <https://github.com/huggingface/transformers/issues/3161?email_source=notifications&email_token=AODM7I44VQEQWKUWI5SYT3TRGFRMLA5CNFSM4LDD7F42YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEOC335Y#issuecomment-595967479>,\n> or unsubscribe\n> <https://github.com/notifications/unsubscribe-auth/AODM7I6VCDKQWMAKL4XDHETRGFRMLANCNFSM4LDD7F4Q>\n> .\n>\n", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3160?src=pr&el=h1) Report\n> Merging [#3160](https://codecov.io/gh/huggingface/transformers/pull/3160?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/6ffe03a0a1d472a4e5941793fd361d2b82c8be3f?src=pr&el=desc) will **increase** coverage by `0.01%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3160?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3160 +/- ##\n==========================================\n+ Coverage 78.11% 78.12% +0.01% \n==========================================\n Files 98 98 \n Lines 16651 16651 \n==========================================\n+ Hits 13007 13009 +2 \n+ Misses 3644 3642 -2\n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3160?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_bart.py](https://codecov.io/gh/huggingface/transformers/pull/3160/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19iYXJ0LnB5) | `93.41% <ø> (ø)` | :arrow_up: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3160/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `94.72% <0%> (+0.31%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3160?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3160?src=pr&el=footer). Last update [6ffe03a...0f206be](https://codecov.io/gh/huggingface/transformers/pull/3160?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "After fixing the these issues the output predictions seems very weird, maybe this is in an label 2 id bug?\r\n\r\n```bash\r\nNachdem I-OTH\r\ner I-PERpart\r\n1907 B-ORGderiv\r\nnicht I-PER\r\nzum I-PER\r\nDirektor I-PER\r\nder I-PERderiv\r\nBibliothek B-ORGpart\r\nberufen I-PER\r\nwurde I-PER\r\n, I-PER\r\nverließ B-PERderiv\r\nGardthausen I-PERderiv\r\nden I-ORG\r\nBibliotheksdienst B-ORGpart\r\n. I-PERderiv\r\n```\r\n\r\n😂", "cc @srush (via #3053)", "Thanks I will take a look. Can you verify you are on the latest pytorch-lightning? `prepare_data` was just added. \r\n\r\nAlso can you post your log. What was the Val accuracy? Are you on single GPU?", "@srush Please see the PR #3180 I have updated the bash script and README to run effortlessly. ", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "Thanks @tomhosking for noticing!", "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3158?src=pr&el=h1) Report\n> Merging [#3158](https://codecov.io/gh/huggingface/transformers/pull/3158?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/6ffe03a0a1d472a4e5941793fd361d2b82c8be3f?src=pr&el=desc) will **increase** coverage by `<.01%`.\n> The diff coverage is `100%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3158?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3158 +/- ##\n==========================================\n+ Coverage 78.11% 78.12% +<.01% \n==========================================\n Files 98 98 \n Lines 16651 16651 \n==========================================\n+ Hits 13007 13008 +1 \n+ Misses 3644 3643 -1\n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3158?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_bart.py](https://codecov.io/gh/huggingface/transformers/pull/3158/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19iYXJ0LnB5) | `93.57% <100%> (+0.16%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3158?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3158?src=pr&el=footer). Last update [6ffe03a...71e626b](https://codecov.io/gh/huggingface/transformers/pull/3158?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3157?src=pr&el=h1) Report\n> Merging [#3157](https://codecov.io/gh/huggingface/transformers/pull/3157?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/6ffe03a0a1d472a4e5941793fd361d2b82c8be3f?src=pr&el=desc) will **decrease** coverage by `0.15%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3157?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3157 +/- ##\n==========================================\n- Coverage 78.11% 77.96% -0.16% \n==========================================\n Files 98 98 \n Lines 16651 16651 \n==========================================\n- Hits 13007 12982 -25 \n- Misses 3644 3669 +25\n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3157?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3157/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.35% <0%> (-5.23%)` | :arrow_down: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3157/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `94.56% <0%> (+0.15%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3157?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3157?src=pr&el=footer). Last update [6ffe03a...491bea5](https://codecov.io/gh/huggingface/transformers/pull/3157?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "Thanks for sharing. I'll just switch the language tag to chinese (tags are case-sensitive)\r\n\r\nModel pages:\r\n[`voidful/albert_chinese_tiny`](https://huggingface.co/voidful/albert_chinese_tiny)\r\n[`voidful/albert_chinese_small`](https://huggingface.co/voidful/albert_chinese_small)\r\n[`voidful/albert_chinese_base`](https://huggingface.co/voidful/albert_chinese_base)\r\n[`voidful/albert_chinese_large`](https://huggingface.co/voidful/albert_chinese_large)\r\n[`voidful/albert_chinese_xlarge`](https://huggingface.co/voidful/albert_chinese_xlarge)\r\n[`voidful/albert_chinese_xxlarge`](https://huggingface.co/voidful/albert_chinese_xxlarge)" ]

[ { "id": 1834056635, "node_id": "MDU6TGFiZWwxODM0MDU2NjM1", "url": "https://api.github.com/repos/huggingface/transformers/labels/Core:%20Tokenization", "name": "Core: Tokenization", "color": "FF4446", "default": false, "description": "Internals of the library; Tokenization." } ]

[]

[]

[]

[]

[ "\r\n" ]

[]

[ { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "Hi @minimaxir - thanks for the Feature request! \r\n\r\nNot too sure about this though. Two arguments against it from my side:\r\n\r\n1. I feel like it's very easy to set the seed parameter before calling `generate()` without any real drawback. \r\n2. Also we want all our `generate()` arguments to have default values with a lot of them defined in `configuration_utils.py`. Adding a `seed` argument would either break this logic or set a default seed value in the `PretrainedConfig` class in `configuration_utils.py` which I definitely don't want to do.", "Same here, thanks for the proposition @minimaxir but I feel like there are many ways you can/should assigne seeds (e.g. [here](https://github.com/huggingface/transformers/blob/master/examples/run_glue.py#L104-L109)) and I don't think we would be comfortable with having this inside the model it-self.", "That's fair; for my purposes, it's not too difficult to wrap. (and I inadvertently realized it's better to wrap it for batch generation too)\r\n\r\nThanks!" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[ { "login": "sshleifer", "id": 6045025, "node_id": "MDQ6VXNlcjYwNDUwMjU=", "avatar_url": "https://avatars.githubusercontent.com/u/6045025?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sshleifer", "html_url": "https://github.com/sshleifer", "followers_url": "https://api.github.com/users/sshleifer/followers", "following_url": "https://api.github.com/users/sshleifer/following{/other_user}", "gists_url": "https://api.github.com/users/sshleifer/gists{/gist_id}", "starred_url": "https://api.github.com/users/sshleifer/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sshleifer/subscriptions", "organizations_url": "https://api.github.com/users/sshleifer/orgs", "repos_url": "https://api.github.com/users/sshleifer/repos", "events_url": "https://api.github.com/users/sshleifer/events{/privacy}", "received_events_url": "https://api.github.com/users/sshleifer/received_events", "type": "User", "site_admin": false } ]

[ "1) Identical to my benchmark for speed. Hadn't tested memory but I'm not surprised that their implementation is less.\r\n\r\nFor both memory and speed, they have a lot of clever tricks that we haven't implemented yet.\r\n\r\n4) Summarization Pipeline will not help, but I will take a longer look at this tomorrow and see if we can improve.\r\n", "On master, the gap has closed considerably!\r\n<16GB GPU RAM for fp16, bs=32, and timings much closer:\r\n\r\n\r\nMy numbers are a bit lower than yours because I am on an NVIDIA RTX GPU.\r\n", "I tested again and I have similar results ! Thanks for the fix.\r\n\r\nI now have exact same GPU memory utilization.\r\n\r\n---\r\n\r\nAbout the (now) small difference of inference time between implementations, do you know from where it comes from ?", "Haven't investigated. So far, I just investigated memory and the speed improvements were a happy side effect.", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "很简单哦。看我的代码：\r\n\r\n\r\n\"\"\"\r\nTraining the distilled model.\r\nSupported architectures include: BERT -> DistilBERT, RoBERTa -> DistilRoBERTa, GPT2 -> DistilGPT2.\r\n\"\"\"\r\nimport argparse\r\nimport json\r\nimport os\r\nimport pickle\r\nimport shutil\r\n\r\nimport numpy as np\r\nimport torch\r\n\r\nfrom distiller import Distiller\r\nfrom lm_seqs_dataset import LmSeqsDataset\r\nfrom transformers import (\r\n BertConfig,\r\n BertForMaskedLM,\r\n BertTokenizer,\r\n DistilBertConfig,\r\n DistilBertForMaskedLM,\r\n DistilBertTokenizer,\r\n GPT2Config,\r\n GPT2LMHeadModel,\r\n GPT2Tokenizer,\r\n RobertaConfig,\r\n RobertaForMaskedLM,\r\n RobertaTokenizer,\r\n)\r\nfrom utils import git_log, init_gpu_params, logger, set_seed\r\n\r\n\r\nMODEL_CLASSES = {\r\n \"distilbert\": (DistilBertConfig, DistilBertForMaskedLM, DistilBertTokenizer),\r\n \"roberta\": (RobertaConfig, RobertaForMaskedLM, RobertaTokenizer),\r\n \"bert\": (BertConfig, BertForMaskedLM, BertTokenizer),\r\n \"gpt2\": (GPT2Config, GPT2LMHeadModel, GPT2Tokenizer),\r\n}\r\n\r\n\r\ndef sanity_checks(args):\r\n \"\"\"\r\n A bunch of args sanity checks to perform even starting...\r\n \"\"\"\r\n assert (args.mlm and args.alpha_mlm > 0.0) or (not args.mlm and args.alpha_mlm == 0.0)\r\n assert (args.alpha_mlm > 0.0 and args.alpha_clm == 0.0) or (args.alpha_mlm == 0.0 and args.alpha_clm > 0.0)\r\n if args.mlm:\r\n assert os.path.isfile(args.token_counts)\r\n assert (args.student_type in [\"roberta\", \"distilbert\"]) and (args.teacher_type in [\"roberta\", \"bert\"])\r\n else:\r\n assert (args.student_type in [\"gpt2\"]) and (args.teacher_type in [\"gpt2\"])\r\n\r\n assert args.teacher_type == args.student_type or (\r\n args.student_type == \"distilbert\" and args.teacher_type == \"bert\"\r\n )\r\n assert os.path.isfile(args.student_config)\r\n if args.student_pretrained_weights is not None:\r\n assert os.path.isfile(args.student_pretrained_weights)\r\n\r\n if args.freeze_token_type_embds:\r\n assert args.student_type in [\"roberta\"]\r\n\r\n assert args.alpha_ce >= 0.0\r\n assert args.alpha_mlm >= 0.0\r\n assert args.alpha_clm >= 0.0\r\n assert args.alpha_mse >= 0.0\r\n assert args.alpha_cos >= 0.0\r\n assert args.alpha_ce + args.alpha_mlm + args.alpha_clm + args.alpha_mse + args.alpha_cos > 0.0\r\n\r\n\r\ndef freeze_pos_embeddings(student, args):\r\n if args.student_type == \"roberta\":\r\n student.roberta.embeddings.position_embeddings.weight.requires_grad = False\r\n elif args.student_type == \"gpt2\":\r\n student.transformer.wpe.weight.requires_grad = False\r\n\r\n\r\ndef freeze_token_type_embeddings(student, args):\r\n if args.student_type == \"roberta\":\r\n student.roberta.embeddings.token_type_embeddings.weight.requires_grad = False\r\n\r\n\r\ndef main():\r\n parser = argparse.ArgumentParser(description=\"Training\")\r\n parser.add_argument(\"--force\", \r\n action=\"store_true\", \r\n default=True,\r\n help=\"Overwrite dump_path if it already exists.\")\r\n\r\n parser.add_argument(\r\n \"--dump_path\", \r\n type=str, \r\n #required=True, \r\n default=r'D:\\2020.03.02distillgpt2' ,\r\n help=\"The output directory (log, checkpoints, parameters, etc.)\"\r\n )\r\n parser.add_argument(\r\n \"--data_file\",\r\n type=str,\r\n #required=True,\r\n default=r'scripts\\gpt2.pickle' ,\r\n help=\"The binarized file (tokenized + tokens_to_ids) and grouped by sequence.\",\r\n )\r\n\r\n parser.add_argument(\r\n \"--student_type\",\r\n type=str,\r\n choices=[\"distilbert\", \"roberta\", \"gpt2\"],\r\n #required=True,\r\n default='gpt2',\r\n help=\"The student type (DistilBERT, RoBERTa).\",\r\n )\r\n parser.add_argument(\"--student_config\", \r\n type=str, \r\n #required=True,\r\n default=r'training_configs\\distilgpt2.json',\r\n help=\"Path to the student configuration.\")\r\n parser.add_argument(\r\n \"--student_pretrained_weights\", default=None, type=str, help=\"Load student initialization checkpoint.\"\r\n )\r\n\r\n parser.add_argument(\r\n \"--teacher_type\", \r\n choices=[\"bert\", \"roberta\", \"gpt2\"], \r\n #required=True, \r\n default='gpt2',\r\n help=\"Teacher type (BERT, RoBERTa).\"\r\n )\r\n parser.add_argument(\"--teacher_name\", \r\n type=str, \r\n #required=True,\r\n default= r'D:\\checkpoint-652500',\r\n help=\"The teacher model.\")\r\n\r\n parser.add_argument(\"--temperature\", \r\n default=1.5, \r\n type=float, help=\"Temperature for the softmax temperature.\")\r\n \r\n parser.add_argument(\r\n \"--alpha_ce\", \r\n default=0.5,\r\n type=float, \r\n help=\"Linear weight for the distillation loss. Must be >=0.\"\r\n )\r\n parser.add_argument(\r\n \"--alpha_mlm\",\r\n default=0.0,\r\n type=float,\r\n help=\"Linear weight for the MLM loss. Must be >=0. Should be used in coonjunction with `mlm` flag.\",\r\n )\r\n parser.add_argument(\"--alpha_clm\", default=0.5, type=float, help=\"Linear weight for the CLM loss. Must be >=0.\")\r\n parser.add_argument(\"--alpha_mse\", default=0.0, type=float, help=\"Linear weight of the MSE loss. Must be >=0.\")\r\n parser.add_argument(\r\n \"--alpha_cos\", default=0.0, type=float, help=\"Linear weight of the cosine embedding loss. Must be >=0.\"\r\n )\r\n\r\n parser.add_argument(\r\n \"--mlm\", action=\"store_true\", help=\"The LM step: MLM or CLM. If `mlm` is True, the MLM is used over CLM.\"\r\n )\r\n parser.add_argument(\r\n \"--mlm_mask_prop\",\r\n default=0.15,\r\n type=float,\r\n help=\"Proportion of tokens for which we need to make a prediction.\",\r\n )\r\n parser.add_argument(\"--word_mask\", default=0.8, type=float, help=\"Proportion of tokens to mask out.\")\r\n parser.add_argument(\"--word_keep\", default=0.1, type=float, help=\"Proportion of tokens to keep.\")\r\n parser.add_argument(\"--word_rand\", default=0.1, type=float, help=\"Proportion of tokens to randomly replace.\")\r\n parser.add_argument(\r\n \"--mlm_smoothing\",\r\n default=0.7,\r\n type=float,\r\n help=\"Smoothing parameter to emphasize more rare tokens (see XLM, similar to word2vec).\",\r\n )\r\n parser.add_argument(\"--token_counts\", \r\n type=str, \r\n default=r'scripts\\gpt2_token_counts.pickle' ,\r\n help=\"The token counts in the data_file for MLM.\")\r\n\r\n parser.add_argument(\r\n \"--restrict_ce_to_mask\",\r\n action=\"store_true\",\r\n help=\"If true, compute the distilation loss only the [MLM] prediction distribution.\",\r\n )\r\n parser.add_argument(\r\n \"--freeze_pos_embs\",\r\n action=\"store_true\",\r\n help=\"Freeze positional embeddings during distillation. For student_type in ['roberta', 'gpt2'] only.\",\r\n )\r\n parser.add_argument(\r\n \"--freeze_token_type_embds\",\r\n action=\"store_true\",\r\n help=\"Freeze token type embeddings during distillation if existent. For student_type in ['roberta'] only.\",\r\n )\r\n\r\n parser.add_argument(\"--n_epoch\", type=int, default=3, help=\"Number of pass on the whole dataset.\")\r\n parser.add_argument(\"--batch_size\", type=int, default=4, help=\"Batch size (for each process).\")\r\n parser.add_argument(\r\n \"--group_by_size\",\r\n action=\"store_false\",\r\n help=\"If true, group sequences that have similar length into the same batch. Default is true.\",\r\n )\r\n\r\n parser.add_argument(\r\n \"--gradient_accumulation_steps\",\r\n type=int,\r\n default=50,\r\n help=\"Gradient accumulation for larger training batches.\",\r\n )\r\n parser.add_argument(\"--warmup_prop\", default=0.05, type=float, help=\"Linear warmup proportion.\")\r\n parser.add_argument(\"--weight_decay\", default=0.0, type=float, help=\"Weight deay if we apply some.\")\r\n parser.add_argument(\"--learning_rate\", default=5e-4, type=float, help=\"The initial learning rate for Adam.\")\r\n parser.add_argument(\"--adam_epsilon\", default=1e-6, type=float, help=\"Epsilon for Adam optimizer.\")\r\n parser.add_argument(\"--max_grad_norm\", default=5.0, type=float, help=\"Max gradient norm.\")\r\n parser.add_argument(\"--initializer_range\", default=0.02, type=float, help=\"Random initialization range.\")\r\n\r\n parser.add_argument(\r\n \"--fp16\",\r\n action=\"store_true\",\r\n help=\"Whether to use 16-bit (mixed) precision (through NVIDIA apex) instead of 32-bit\",\r\n )\r\n parser.add_argument(\r\n \"--fp16_opt_level\",\r\n type=str,\r\n default=\"O1\",\r\n help=\"For fp16: Apex AMP optimization level selected in ['O0', 'O1', 'O2', and 'O3'].\"\r\n \"See details at https://nvidia.github.io/apex/amp.html\",\r\n )\r\n parser.add_argument(\"--n_gpu\", type=int, default=1, help=\"Number of GPUs in the node.\")\r\n parser.add_argument(\"--local_rank\", type=int, default=-1, help=\"Distributed training - Local rank\")\r\n parser.add_argument(\"--seed\", type=int, default=2020, help=\"Random seed\")\r\n\r\n parser.add_argument(\"--log_interval\", type=int, default=500, help=\"Tensorboard logging interval.\")\r\n parser.add_argument(\"--checkpoint_interval\", type=int, default=1500, help=\"Checkpoint interval.\")\r\n args = parser.parse_args([])\r\n sanity_checks(args)\r\n\r\n # ARGS #\r\n init_gpu_params(args)\r\n set_seed(args)\r\n if args.is_master:\r\n if os.path.exists(args.dump_path):\r\n if not args.force:\r\n raise ValueError(\r\n f\"Serialization dir {args.dump_path} already exists, but you have not precised wheter to overwrite it\"\r\n \"Use `--force` if you want to overwrite it\"\r\n )\r\n else:\r\n shutil.rmtree(args.dump_path)\r\n\r\n if not os.path.exists(args.dump_path):\r\n os.makedirs(args.dump_path)\r\n logger.info(f\"Experiment will be dumped and logged in {args.dump_path}\")\r\n\r\n # SAVE PARAMS #\r\n logger.info(f\"Param: {args}\")\r\n with open(os.path.join(args.dump_path, \"parameters.json\"), \"w\",encoding = 'utf-8') as f:\r\n json.dump(vars(args), f, indent=4)\r\n #git_log(args.dump_path)\r\n\r\n student_config_class, student_model_class, _ = MODEL_CLASSES[args.student_type]\r\n teacher_config_class, teacher_model_class, teacher_tokenizer_class = MODEL_CLASSES[args.teacher_type]\r\n\r\n # TOKENIZER #\r\n from transformers import BertTokenizer \r\n tokenizer = BertTokenizer(\r\n vocab_file = r\"scripts\\vocab.txt\",\r\n unk_token='<unk>',\r\n sep_token='<sep>',\r\n pad_token='<pad>',\r\n cls_token='</s>',\r\n mask_token='<mask>') \r\n \r\n special_tokens_dict = {\"bos_token\": \"<s>\", \"eos_token\": \"</s>\"}\r\n tokenizer.add_special_tokens(special_tokens_dict)\r\n special_tok_ids = {}\r\n for tok_name, tok_symbol in tokenizer.special_tokens_map.items():\r\n idx = tokenizer.all_special_tokens.index(tok_symbol)\r\n special_tok_ids[tok_name] = tokenizer.all_special_ids[idx]\r\n logger.info(f\"Special tokens {special_tok_ids}\")\r\n args.special_tok_ids = special_tok_ids\r\n args.max_model_input_size = 512\r\n\r\n # DATA LOADER #\r\n logger.info(f\"Loading data from {args.data_file}\")\r\n with open(args.data_file, \"rb\") as fp:\r\n data = pickle.load(fp)\r\n\r\n if args.mlm:\r\n logger.info(f\"Loading token counts from {args.token_counts} (already pre-computed)\")\r\n with open(args.token_counts, \"rb\") as fp:\r\n counts = pickle.load(fp)\r\n\r\n token_probs = np.maximum(counts, 1) ** -args.mlm_smoothing\r\n for idx in special_tok_ids.values():\r\n token_probs[idx] = 0.0 # do not predict special tokens\r\n token_probs = torch.from_numpy(token_probs)\r\n else:\r\n token_probs = None\r\n\r\n train_lm_seq_dataset = LmSeqsDataset(params=args, data=data)\r\n logger.info(f\"Data loader created.\")\r\n\r\n # STUDENT #\r\n logger.info(f\"Loading student config from {args.student_config}\")\r\n stu_architecture_config = student_config_class.from_pretrained(args.student_config)\r\n stu_architecture_config.output_hidden_states = True\r\n\r\n if args.student_pretrained_weights is not None:\r\n logger.info(f\"Loading pretrained weights from {args.student_pretrained_weights}\")\r\n student = student_model_class.from_pretrained(args.student_pretrained_weights, config=stu_architecture_config)\r\n else:\r\n student = student_model_class(stu_architecture_config)\r\n\r\n if args.n_gpu > 0:\r\n student.to(f\"cuda:{args.local_rank}\")\r\n logger.info(f\"Student loaded.\")\r\n\r\n # TEACHER #\r\n teacher = teacher_model_class.from_pretrained(args.teacher_name, output_hidden_states=True)\r\n teacher.resize_token_embeddings(len(tokenizer)) # Update the model embeddings with the new vocabulary size\r\n teacher.to('cuda')\r\n teacher.eval()\r\n if args.n_gpu > 0:\r\n teacher.to(f\"cuda:{args.local_rank}\")\r\n logger.info(f\"Teacher loaded from {args.teacher_name}.\")\r\n\r\n # FREEZING #\r\n if args.freeze_pos_embs:\r\n freeze_pos_embeddings(student, args)\r\n if args.freeze_token_type_embds:\r\n freeze_token_type_embeddings(student, args)\r\n\r\n # SANITY CHECKS #\r\n assert student.config.vocab_size == teacher.config.vocab_size\r\n assert student.config.hidden_size == teacher.config.hidden_size\r\n assert student.config.max_position_embeddings == teacher.config.max_position_embeddings\r\n if args.mlm:\r\n assert token_probs.size(0) == stu_architecture_config.vocab_size\r\n\r\n # DISTILLER #\r\n torch.cuda.empty_cache()\r\n distiller = Distiller(\r\n params=args, dataset=train_lm_seq_dataset, token_probs=token_probs, student=student, teacher=teacher\r\n )\r\n distiller.train()\r\n logger.info(\"Let's go get some drinks.\")\r\n\r\n\r\nif __name__ == \"__main__\":\r\n main()\r\n", "Hi, not sure where to put this but there might be some small errors in the training code provided in this repo.\r\n1. `n_gpus` is used when initializing GPUs but the actual name of this variable is `gpus`\r\n2. It seems that now `return_dict` is default to `True`, so the function`step` fails because the results are unpacked to keys rather than values. The easiest fix I guess is to turn off `return_dict` in `train.py` like the following\r\n ```python\r\n student.config.update(dict(return_dict=False))\r\n teacher.config.update(dict(return_dict=False))\r\n ```", "I followed the guide in README to train a distill gpt2, but the performance of my distilgpt2 is not good as huggingface, actually, the performance is bad. Did you trained a nice performance distilgpt2?" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null }, { "id": 1834056761, "node_id": "MDU6TGFiZWwxODM0MDU2NzYx", "url": "https://api.github.com/repos/huggingface/transformers/labels/Core:%20Modeling", "name": "Core: Modeling", "color": "FF8446", "default": false, "description": "Internals of the library; Models." } ]

[]

[ "I tried to replicate the input as close as I could given the output you gave:\r\n\r\n```py\r\nfrom transformers import BertModel\r\nimport torch\r\n\r\ntoken_id = torch.tensor([[ 101, 5292, 3270, 102, 8638, 2060, 102]])\r\ntoken_types = torch.tensor([[0, 0, 0, 0, 1, 1, 1]])\r\nmask = torch.tensor([[1., 1., 1., 1., 1., 1., 1.]])\r\n\r\npadded_id = torch.tensor([[ 101, 5292, 3270, 102, 8638, 2060, 102, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]])\r\npadded_type = torch.tensor([[0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]])\r\npadded_mask = torch.tensor([[1., 1., 1., 1., 1., 1., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.]])\r\n\r\nbertModel = BertModel.from_pretrained('bert-base-uncased')\r\nbertModel = bertModel.eval()\r\n\r\noutput = bertModel(token_id, token_type_ids=token_types, attention_mask=mask)\r\npadded_output = bertModel(padded_id, token_type_ids=padded_type, attention_mask=padded_mask)\r\n```\r\n\r\nI then print the maximum difference between the output of the model with the non-padded input and the output with the padded input:\r\n\r\n```py\r\nprint(torch.max(output[0] - padded_output[0][:, :7]))\r\nprint(torch.max(output[1] - padded_output[1]))\r\n```\r\n\r\nWhich outputs the following (negligible) difference:\r\n\r\n```py\r\ntensor(3.6359e-06, grad_fn=<MaxBackward1>)\r\ntensor(5.9605e-07, grad_fn=<MaxBackward1>)\r\n```\r\n\r\nWould it be possible for you to give a completely reproducible script so that I may see where the issue lies?", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3149?src=pr&el=h1) Report\n> Merging [#3149](https://codecov.io/gh/huggingface/transformers/pull/3149?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/857e0a0d3ba39be6259961524a730d3f106cec9c?src=pr&el=desc) will **decrease** coverage by `0.05%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3149?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3149 +/- ##\n==========================================\n- Coverage 78.03% 77.97% -0.06% \n==========================================\n Files 98 98 \n Lines 16588 16588 \n==========================================\n- Hits 12944 12935 -9 \n- Misses 3644 3653 +9\n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3149?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3149/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `91.08% <0%> (-2.13%)` | :arrow_down: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3149/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `94.56% <0%> (+0.15%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3149?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3149?src=pr&el=footer). Last update [857e0a0...58fc8f9](https://codecov.io/gh/huggingface/transformers/pull/3149?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "Thanks!" ]

[]

[]

[ "good to merge for me", "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3148?src=pr&el=h1) Report\n> Merging [#3148](https://codecov.io/gh/huggingface/transformers/pull/3148?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/0001d056861bb1ec7bd6a825006f578629a101fc?src=pr&el=desc) will **decrease** coverage by `1.04%`.\n> The diff coverage is `95.83%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/3148?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #3148 +/- ##\n==========================================\n- Coverage 78.03% 76.98% -1.05% \n==========================================\n Files 98 98 \n Lines 16573 16583 +10 \n==========================================\n- Hits 12932 12766 -166 \n- Misses 3641 3817 +176\n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/3148?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3148/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `91.47% <95.83%> (-1.95%)` | :arrow_down: |\n| [src/transformers/modeling\\_tf\\_pytorch\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3148/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl9weXRvcmNoX3V0aWxzLnB5) | `8.72% <0%> (-81.21%)` | :arrow_down: |\n| [src/transformers/modeling\\_roberta.py](https://codecov.io/gh/huggingface/transformers/pull/3148/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19yb2JlcnRhLnB5) | `85.71% <0%> (-10%)` | :arrow_down: |\n| [src/transformers/modeling\\_xlnet.py](https://codecov.io/gh/huggingface/transformers/pull/3148/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ194bG5ldC5weQ==) | `73.48% <0%> (-2.3%)` | :arrow_down: |\n| [src/transformers/modeling\\_ctrl.py](https://codecov.io/gh/huggingface/transformers/pull/3148/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19jdHJsLnB5) | `96% <0%> (-2.23%)` | :arrow_down: |\n| [src/transformers/modeling\\_openai.py](https://codecov.io/gh/huggingface/transformers/pull/3148/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19vcGVuYWkucHk=) | `80.2% <0%> (-1.35%)` | :arrow_down: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/3148/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `94.56% <0%> (-0.16%)` | :arrow_down: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/3148?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/3148?src=pr&el=footer). Last update [0001d05...2861c9d](https://codecov.io/gh/huggingface/transformers/pull/3148?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]