connected to google colab with a t4 gpu: Traceback (most recent call last): File…
- User
connected to google colab with a t4 gpu:
Traceback (most recent call last): File "/content/PaddleOCR/tools/train.py", line 224, in config, device, logger, vdl_writer = program.preprocess(is_train=True) File "/content/PaddleOCR/tools/program.py", line 678, in preprocess .dev_id) if use_gpu else 'cpu' File "/usr/local/lib/python3.10/dist-packages/paddle/distributed/parallel.py", line 774, in device_id return self._device_id AttributeError: 'ParallelEnv' object has no attribute '_device_id'. Did you mean: 'device_id'? [28] 1 s from paddleocr import PaddleOCR ocr = PaddleOCR(use_angle_cls=True, lang='en')
Global: use_gpu: true use_xpu: false use_mlu: false epoch_num: 1200 log_smooth_window: 20 print_batch_step: 10 save_model_dir: ./output/db_mv3/ save_epoch_step: 1200
evaluation is run every 2000 iterations
eval_batch_step: [0, 2000] cal_metric_during_train: False pretrained_model: ./pretrain_models/MobileNetV3_large_x0_5_pretrained checkpoints: save_inference_dir: use_visualdl: False infer_img: doc/imgs_en/img_10.jpg save_res_path: ./output/det_db/predicts_db.txt
Architecture: model_type: det algorithm: DB Transform: Backbone: name: MobileNetV3 scale: 0.5 model_name: large Neck: name: DBFPN out_channels: 256 Head: name: DBHead k: 50
Loss: name: DBLoss balance_loss: true main_loss_type: DiceLoss alpha: 5 beta: 10 ohem_ratio: 3
Optimizer: name: Adam beta1: 0.9 beta2: 0.999 lr: learning_rate: 0.001 regularizer: name: 'L2' factor: 0
PostProcess: name: DBPostProcess thresh: 0.3 box_thresh: 0.6 max_candidates: 1000 unclip_ratio: 1.5
Metric: name: DetMetric main_indicator: hmean
Train: dataset: name: SimpleDataSet data_dir: /content/train/images/ label_file_list: - /content/train/labels/train_labels.txt ratio_list: [1.0] transforms: - DecodeImage: # load image img_mode: BGR channel_first: False - DetLabelEncode: # Class handling label - IaaAugment: augmenter_args: - { 'type': Fliplr, 'args': { 'p': 0.5 } } - { 'type': Affine, 'args': { 'rotate': [-10, 10] } } - { 'type': Resize, 'args': { 'size': [0.5, 3] } } - EastRandomCropData: size: [640, 640] max_tries: 50 keep_ratio: true - MakeBorderMap: shrink_ratio: 0.4 thresh_min: 0.3 thresh_max: 0.7 - MakeShrinkMap: shrink_ratio: 0.4 min_text_size: 8 - NormalizeImage: scale: 1./255. mean: [0.485, 0.456, 0.406] std: [0.229, 0.224, 0.225] order: 'hwc' - ToCHWImage: - KeepKeys: keep_keys: ['image', 'threshold_map', 'threshold_mask', 'shrink_map', 'shrink_mask'] # the order of the dataloader list loader: shuffle: True drop_last: False batch_size_per_card: 16 num_workers: 8 use_shared_memory: True
Eval: dataset: name: SimpleDataSet data_dir: /content/valid/images/ label_file_list: - /content/valid/labels/val_labels.txt transforms: - DecodeImage: # load image img_mode: BGR channel_first: False - DetLabelEncode: # Class handling label - DetResizeForTest: image_shape: [736, 1280] - NormalizeImage: scale: 1./255. mean: [0.485, 0.456, 0.406] std: [0.229, 0.224, 0.225] order: 'hwc' - ToCHWImage: - KeepKeys: keep_keys: ['image', 'shape', 'polys', 'ignore_tags'] loader: shuffle: False drop_last: False batch_size_per_card: 1 # must be 1 num_workers: 8 use_shared_memory: True
- Assistant
It appears that you're facing an issue with the PaddleOCR library while trying to train a model on Google Colab with a T4 GPU. The error message you're seeing is:
AttributeError: 'ParallelEnv' object has no attribute '_device_id'. Did you mean: 'device_id'?This type of error suggests that there might be a mismatch between the PaddleOCR code you are running and the version of PaddlePaddle installed, or there's a bug in the PaddlePaddle implementation.
Here are several steps you can take to resolve the issue:
- Check PaddlePaddle Version: Ensure that you have the correct version of PaddlePaddle installed that is compatible with the version of PaddleOCR you're using. You can install or upgrade PaddlePaddle in Colab with the following:
!pip install paddlepaddle-gpu==[version] -f https://www.paddlepaddle.org.cn/whl/mkl/stable.htmlReplace
[version]with the specific version compatible with your PaddleOCR. It's typically a good idea to consult the documentation or the GitHub repository to find the correct versions.- Update PaddleOCR: Ensure you have the latest version of PaddleOCR, or the version that is known to work with your current PaddlePaddle installation:
!pip install paddleocrOr if you cloned the repository:
cd PaddleOCR git pull-
Place the Code Inside the Main Guard: If you're running your script as a standalone program, ensure the main code is inside a
if __name__ == '__main__':guard when using multiprocessing to avoid potential issues in certain environments. -
Check PaddleOCR GitHub Issues: Search for any open or closed issues on the PaddleOCR GitHub repository that might describe similar problems and potential solutions or patches.
-
Report the Issue: If none of these steps work and you suspect a bug in the library, you could open a new issue on the PaddleOCR GitHub repository with all the details including the error message, the version of PaddleOCR, the version of PaddlePaddle, and the code that resulted in the error.
As a temporary workaround, you can also try using a CPU for training by setting
use_gpu: falsein your configuration file to see if the problem is specifically related to GPU parallel environment. However, be aware that training on a CPU will be significantly slower than training on a GPU.Please note that any adjustments should be consistent with the environment and dependencies of PaddleOCR and PaddlePaddle.