The length of ratio_list should be the same as the file_list.

[bound123]

🔎 Search before asking

• I have searched the PaddleOCR Docs and found no similar bug report.
• I have searched the PaddleOCR Issues and found no similar bug report.
• I have searched the PaddleOCR Discussions and found no similar bug report.

🐛 Bug (问题描述)

Traceback (most recent call last):
File "D:\PaddleOCR-main\tools\train.py", line 269, in
main(config, device, logger, vdl_writer, seed)
File "D:\PaddleOCR-main\tools\train.py", line 55, in main
train_dataloader = build_dataloader(config, "Train", device, logger, seed)
File "D:\PaddleOCR-main\ppocr\data_init_.py", line 107, in build_dataloader
dataset = eval(module_name)(config, mode, logger, seed)
File "D:\PaddleOCR-main\ppocr\data\simple_dataset.py", line 42, in init
assert (
AssertionError: The length of ratio_list should be the same as the file_list.

🏃‍♂️ Environment (运行环境)

python 3.10

🌰 Minimal Reproducible Example (最小可复现问题的Demo)

配置文件如图
Global:
use_gpu: false
epoch_num: 1200
log_smooth_window: 20
print_batch_step: 2
save_model_dir: ./output/ch_db_res18/
save_epoch_step: 1200

evaluation is run every 5000 iterations after the 4000th iteration

eval_batch_step: [3000, 2000]
cal_metric_during_train: False
pretrained_model: D:\ancoda\anocada3\envs\paddle\Lib\site-packages\Preliminary_training\ch_ppocr_server_v2.0_det_train\best_accuracy.pdparams
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: detc
algorithm: DB
Transform:
Backbone:
name: ResNet_vd
layers: 18
disable_se: True
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:
name: Cosine
learning_rate: 0.001
warmup_epoch: 2
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: D:\ancoda\anocada3\envs\paddle\Lib\site-packages\train_data\det\train
label_file_list:
D:/ancoda/anocada3/envs/paddle/Lib/site-packages/train_data/det/train.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: [960, 960]
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: 8
num_workers: 4

Eval:
dataset:
name: SimpleDataSet
data_dir: D:\ancoda\anocada3\envs\paddle\Lib\site-packages\train_data\det\val
label_file_list:
D:/ancoda/anocada3/envs/paddle/Lib/site-packages/train_data/det/val.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: 2

[GreatV]

The error message AssertionError: The length of ratio_list should be the same as the file_list indicates that there is a mismatch between the number of elements in ratio_list and label_file_list in your configuration file.

Explanation

The ratio_list is used to specify the proportion of data to be used from each dataset listed in label_file_list. For every dataset in label_file_list, there must be a corresponding entry in ratio_list that defines its ratio. If the lengths of these two lists do not match, this error is raised.

In your configuration:

label_file_list:
  D:/ancoda/anocada3/envs/paddle/Lib/site-packages/train_data/det/train.txt
ratio_list: [1.0]

Here, the label_file_list contains one file, and the ratio_list correctly contains one ratio, so this part seems fine. However, this error can also occur if:

1. Your YAML formatting is incorrect. YAML is sensitive to formatting, so make sure there are no indentation or parsing issues.
2. The label_file_list is being parsed incorrectly. The file path may not be recognized as a valid list, or the configuration parser may be misinterpreting it.

Steps to Fix the Error

1. Ensure Proper YAML Syntax

   • Double-check the indentation and syntax in your configuration file. For example, ensure label_file_list is written as a proper list:

     label_file_list:
       - D:/ancoda/anocada3/envs/paddle/Lib/site-packages/train_data/det/train.txt
     ratio_list: [1.0]

     Note the use of - before the file path to explicitly define it as a list.

2. Verify the Dataset File

   • Ensure the file D:/ancoda/anocada3/envs/paddle/Lib/site-packages/train_data/det/train.txt exists and is accessible. If the file is missing, the label_file_list might be treated as empty.

3. Check for Multiple Datasets

   • If you plan to use multiple datasets in label_file_list, ensure the ratio_list has the same number of elements. For example:

     label_file_list:
       - path/to/dataset1.txt
       - path/to/dataset2.txt
     ratio_list: [0.6, 0.4]

     Here, ratio_list specifies that 60% of the data should come from dataset1 and 40% from dataset2.

4. Debug the Configuration Parsing

   • Insert debug statements or print the parsed configuration to verify what label_file_list and ratio_list contain at runtime. For example, you can add a print statement in the script (simple_dataset.py) where the assertion is failing:

     print(f"label_file_list: {self.label_file_list}")
     print(f"ratio_list: {self.ratio_list}")

5. Update PaddleOCR Version

   • Ensure you are using the latest stable version of PaddleOCR and PaddlePaddle. Bugs in earlier versions might cause unexpected parsing issues. To update:

     pip install --upgrade paddleocr paddlepaddle

6. Cross-Check with Related Discussions

   • The linked Discussion explains issues related to dataset configuration and shape mismatches, which might also cause similar errors. Ensure your dataset is properly pre-processed and the file paths are correct.

Additional Notes

• If you are still facing issues after following the steps above, try running a minimal configuration with a dummy dataset to isolate the problem. Test with:

  label_file_list:
    - dummy_dataset.txt
  ratio_list: [1.0]

  Ensure dummy_dataset.txt contains valid data in the expected format.

By properly aligning the lengths of label_file_list and ratio_list, and ensuring the dataset files are correctly specified, you should be able to resolve this error.

---

Response generated by feifei-bot | chatgpt-4o-latest