File size: 13,372 Bytes
e92402e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 |
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# YOLOv11 Training with Roboflow Dataset\n",
"\n",
"This notebook demonstrates how to train a YOLOv11 model using a dataset from Roboflow. It includes:\n",
"- Automatic GPU/CPU detection\n",
"- Configurable training parameters\n",
"- Training visualization and analysis\n",
"\n",
"## Step 1: Install Dependencies\n",
"First, we'll install the required packages."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# For Training\n",
"!pip install ultralytics roboflow \n",
"\n",
"# For Storage\n",
"!pip install minio"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Step 2: Import Libraries\n",
"Import all necessary libraries for training and analysis."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Common\n",
"import os\n",
"\n",
"# For Dataset manipulation\n",
"import yaml\n",
"from roboflow import Roboflow\n",
"\n",
"# For training\n",
"import torch\n",
"from ultralytics import YOLO\n",
"\n",
"# For Storage\n",
"from minio import Minio\n",
"from minio.error import S3Error"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Step 3: Download Dataset from Roboflow\n",
"Connect to Roboflow and download the dataset. Make sure to use your own API key and project details.\n",
"\n",
"**Remember to replace the placeholders with your values**."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"rf = Roboflow(api_key=\"xxxxxxxxxxxxxxxxx\") # Replace with your API key\n",
"project = rf.workspace(\"yyyyyyyyyyyyyy\").project(\"zzzzzzzzzzzzzzzzzzz\") # Replace with your workspace and project names\n",
"version = project.version(1111111111111111111111111111) # Replace with your version number\n",
"dataset = version.download(\"yolov11\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You'll need to explicitly specify the paths to each data split (training, validation, and test) in your configuration. This ensures YOLO can correctly locate and utilize your dataset files.\n",
"\n",
"This is done in the `data.yaml` file. If you open that file you will see these paths that you need to update:\n",
"\n",
"```\n",
"train: ../train/images\n",
"val: ../valid/images\n",
"test: ../test/images\n",
"```"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(f\"Dataset downloaded to: {dataset.location}\")\n",
"\n",
"dataset_yaml_path = f\"{dataset.location}/data.yaml\"\n",
"\n",
"with open(dataset_yaml_path, \"r\") as file:\n",
" data_config = yaml.safe_load(file)\n",
"\n",
"data_config[\"train\"] = f\"{dataset.location}/train/images\"\n",
"data_config[\"val\"] = f\"{dataset.location}/valid/images\"\n",
"data_config[\"test\"] = f\"{dataset.location}/test/images\"\n",
"\n",
"with open(dataset_yaml_path, \"w\") as file:\n",
" yaml.safe_dump(data_config, file)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Step 4: Configure Hyperparameters\n",
"Set up GPU/CPU detection (code automatically detects and use GPU if available)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
"print(f\"Using device: {device} ({'GPU' if device.type == 'cuda' else 'CPU'})\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Define all training parameters in a single configuration dictionary."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"\n",
"CONFIG = {\n",
" 'name': 'yolo_hardhat',\n",
" 'model': 'yolo11m.pt', # Model size options: n, s, m, l, x\n",
" 'data': dataset.location + \"/data.yaml\",\n",
" 'epochs': 1, # Set the number of epochs (keep 1 for Mock Training)\n",
" 'batch': 1 , # Adjust batch size based on device\n",
" 'imgsz': 640,\n",
" 'patience': 15,\n",
" 'device': device,\n",
" \n",
" # Optimizer settings\n",
" 'optimizer': 'SGD',\n",
" 'lr0': 0.001,\n",
" 'lrf': 0.005,\n",
" 'momentum': 0.9,\n",
" 'weight_decay': 0.0005,\n",
" 'warmup_epochs': 3,\n",
" 'warmup_bias_lr': 0.01,\n",
" 'warmup_momentum': 0.8,\n",
" 'amp': False,\n",
" \n",
" # Data augmentation settings\n",
" 'augment': True,\n",
" 'hsv_h': 0.015, # HSV-Hue augmentation\n",
" 'hsv_s': 0.7, # HSV-Saturation augmentation\n",
" 'hsv_v': 0.4, # HSV-Value augmentation\n",
" 'degrees': 10, # Image rotation (+/- deg)\n",
" 'translate': 0.1, # Image translation\n",
" 'scale': 0.3, # Image scale\n",
" 'shear': 0.0, # Image shear\n",
" 'perspective': 0.0, # Image perspective\n",
" 'flipud': 0.1, # Flip up-down\n",
" 'fliplr': 0.1, # Flip left-right\n",
" 'mosaic': 1.0, # Mosaic augmentation\n",
" 'mixup': 0.0, # Mixup augmentation\n",
"}\n",
"\n",
"# Configure PyTorch for GPU memory allocation\n",
"os.environ[\"PYTORCH_CUDA_ALLOC_CONF\"] = \"expandable_segments:True\""
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Step 5: Load Model\n",
"Initialize the YOLO model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"model = YOLO(CONFIG['model'])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Step 6: Start Training\n",
"\n",
"Begin the training process. By default, the `train` method handles both \"training\" and \"validation\" sets. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"results_train = model.train(\n",
" name=CONFIG['name'],\n",
" data=CONFIG['data'],\n",
" epochs=CONFIG['epochs'],\n",
" batch=CONFIG['batch'],\n",
" imgsz=CONFIG['imgsz'],\n",
" patience=CONFIG['patience'],\n",
" device=CONFIG['device'],\n",
" verbose=True,\n",
" \n",
" # Optimizer parameters\n",
" optimizer=CONFIG['optimizer'],\n",
" lr0=CONFIG['lr0'],\n",
" lrf=CONFIG['lrf'],\n",
" momentum=CONFIG['momentum'],\n",
" weight_decay=CONFIG['weight_decay'],\n",
" warmup_epochs=CONFIG['warmup_epochs'],\n",
" warmup_bias_lr=CONFIG['warmup_bias_lr'],\n",
" warmup_momentum=CONFIG['warmup_momentum'],\n",
" amp=CONFIG['amp'],\n",
" \n",
" # Augmentation parameters\n",
" augment=CONFIG['augment'],\n",
" hsv_h=CONFIG['hsv_h'],\n",
" hsv_s=CONFIG['hsv_s'],\n",
" hsv_v=CONFIG['hsv_v'],\n",
" degrees=CONFIG['degrees'],\n",
" translate=CONFIG['translate'],\n",
" scale=CONFIG['scale'],\n",
" shear=CONFIG['shear'],\n",
" perspective=CONFIG['perspective'],\n",
" flipud=CONFIG['flipud'],\n",
" fliplr=CONFIG['fliplr'],\n",
" mosaic=CONFIG['mosaic'],\n",
" mixup=CONFIG['mixup'],\n",
")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Step 7: Evaluate Model\n",
"\n",
" Evaluate the model on the test set."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"results_test = model.val(data=CONFIG['data'], split='test', device=CONFIG['device'], imgsz=CONFIG['imgsz'])\n",
"\n",
"#print(\"Test Results:\", results_test)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Step 8: (Optional) Model Export\n",
"\n",
"Export the trained YOLO model to ONNX format for deployment."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"model.export(format='onnx', imgsz=CONFIG['imgsz'])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Export the trained YOLO model to TorchScript"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#model.export(format=\"torchscript\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Step 9: Store the Model\n",
"\n",
"Save the trained model to the Object Storage system configured in your Workbench connection. \n",
"\n",
"Start by getting the credentials and configuring variables for accessing Object Storage."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"AWS_S3_ENDPOINT_NAME = os.getenv(\"AWS_S3_ENDPOINT\", \"\").replace('https://', '').replace('http://', '')\n",
"AWS_ACCESS_KEY_ID = os.getenv(\"AWS_ACCESS_KEY_ID\")\n",
"AWS_SECRET_ACCESS_KEY = os.getenv(\"AWS_SECRET_ACCESS_KEY\")\n",
"AWS_S3_BUCKET = os.getenv(\"AWS_S3_BUCKET\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Define the S3 client."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"client = Minio(\n",
" AWS_S3_ENDPOINT_NAME,\n",
" access_key=AWS_ACCESS_KEY_ID,\n",
" secret_key=AWS_SECRET_ACCESS_KEY,\n",
" secure=True\n",
")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Select files to be uploaded (files generated while training and validating the model)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"model_path_train = results_train.save_dir\n",
"weights_path = os.path.join(model_path_train, \"weights\")\n",
"model_path_test = results_test.save_dir\n",
"\n",
"files_train = [os.path.join(model_path_train, f) for f in os.listdir(model_path_train) if os.path.isfile(os.path.join(model_path_train, f))]\n",
"files_models = [os.path.join(weights_path, f) for f in os.listdir(weights_path) if os.path.isfile(os.path.join(weights_path, f))]\n",
"files_test = [os.path.join(model_path_test, f) for f in os.listdir(model_path_test) if os.path.isfile(os.path.join(model_path_test, f))]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Upload the files."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"directory_name= os.path.basename(model_path_train)\n",
"\n",
"for file_path_train in files_train:\n",
" try:\n",
" client.fput_object(AWS_S3_BUCKET, \"prototype/notebook/\" + directory_name + \"/train-val/\" + os.path.basename(file_path_train), file_path_train)\n",
" print(f\"'{os.path.basename(file_path_train)}' is successfully uploaded as object to bucket '{AWS_S3_BUCKET}'.\")\n",
" except S3Error as e:\n",
" print(\"Error occurred: \", e)\n",
"\n",
"for file_path_model in files_models:\n",
" try:\n",
" client.fput_object(AWS_S3_BUCKET, \"prototype/notebook/\" + directory_name + \"/\" + os.path.basename(file_path_model), file_path_model)\n",
" print(f\"'{os.path.basename(file_path_model)}' is successfully uploaded as object to bucket '{AWS_S3_BUCKET}'.\")\n",
" except S3Error as e:\n",
" print(\"Error occurred: \", e)\n",
"\n",
"for file_path_test in files_test:\n",
" try:\n",
" client.fput_object(AWS_S3_BUCKET, \"prototype/notebook/\" + directory_name + \"/test/\" + os.path.basename(file_path_test), file_path_test)\n",
" print(f\"'{os.path.basename(file_path_test)}' is successfully uploaded as object to bucket '{AWS_S3_BUCKET}'.\")\n",
" except S3Error as e:\n",
" print(\"Error occurred: \", e)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Step 10: Remove local files\n",
"\n",
"Once you uploaded the Model data to the Object Storage, you can remove the local files to save disk space."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"!rm -rf {model_path_train}\n",
"!rm -rf {model_path_test}"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.12.7"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
|