2020-06-24 18:13:03 +08:00
# OPTIONAL PARAMETERS LIST
2020-06-08 10:16:26 +08:00
The following list can be viewed via `--help`
| FLAG | Supported script | Use | Defaults | Note |
| :----------------------: | :------------: | :---------------: | :--------------: | :-----------------: |
| -c | ALL | Specify configuration file to use | None | **Please refer to the parameter introduction for configuration file usage** |
| -o | ALL | set configuration options | None | Configuration using -o has higher priority than the configuration file selected with -c. E.g: `-o Global.use_gpu=false` |
2020-06-24 18:13:03 +08:00
## INTRODUCTION TO GLOBAL PARAMETERS OF CONFIGURATION FILE
2020-06-08 10:16:26 +08:00
Take `rec_chinese_lite_train.yml` as an example
| Parameter | Use | Default | Note |
| :----------------------: | :---------------------: | :--------------: | :--------------------: |
| algorithm | Select algorithm to use | Synchronize with configuration file | For selecting model, please refer to the supported model [list ](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/README_en.md ) |
| use_gpu | Set using GPU or not | true | \ |
| epoch_num | Maximum training epoch number | 3000 | \ |
| log_smooth_window | Sliding window size | 20 | \ |
| print_batch_step | Set print log interval | 10 | \ |
| save_model_dir | Set model save path | output/{model_name} | \ |
| save_epoch_step | Set model save interval | 3 | \ |
| eval_batch_step | Set the model evaluation interval | 2000 | \ |
|train_batch_size_per_card | Set the batch size during training | 256 | \ |
| test_batch_size_per_card | Set the batch size during testing | 256 | \ |
| image_shape | Set input image size | [3, 32, 100] | \ |
| max_text_length | Set the maximum text length | 25 | \ |
| character_type | Set character type | ch | en/ch, the default dict will be used for en, and the custom dict will be used for ch|
| character_dict_path | Set dictionary path | ./ppocr/utils/ic15_dict.txt | \ |
| loss_type | Set loss type | ctc | Supports two types of loss: ctc / attention |
| reader_yml | Set the reader configuration file | ./configs/rec/rec_icdar15_reader.yml | \ |
| pretrain_weights | Load pre-trained model path | ./pretrain_models/CRNN/best_accuracy | \ |
| checkpoints | Load saved model path | None | Used to load saved parameters to continue training after interruption |
| save_inference_dir | path to save model for inference | None | Use to save inference model |
2020-06-24 18:13:03 +08:00
## INTRODUCTION TO READER PARAMETERS OF CONFIGURATION FILE
2020-06-08 10:16:26 +08:00
Take `rec_chinese_reader.yml` as an example:
| Parameter | Use | Default | Note |
| :----------------------: | :---------------------: | :--------------: | :--------------------: |
| reader_function | Select data reading method | ppocr.data.rec.dataset_traversal,SimpleReader | Support two data reading methods: SimpleReader / LMDBReader |
| num_workers | Set the number of data reading threads | 8 | \ |
| img_set_dir | Image folder path | ./train_data | \ |
| label_file_path | Groundtruth file path | ./train_data/rec_gt_train.txt| \ |
| infer_img | Result folder path | ./infer_img | \|
2020-06-12 12:09:55 +08:00
2020-06-24 18:13:03 +08:00
## INTRODUCTION TO OPTIMIZER PARAMETERS OF CONFIGURATION FILE
2020-06-12 12:09:55 +08:00
Take `rec_icdar15_train.yml` as an example:
| Parameter | Use | Default | None |
| :---------------------: | :---------------------: | :--------------: | :--------------------: |
| function | Select Optimizer function | pocr.optimizer,AdamDecay | Only support Adam |
| base_lr | Set the base lr | 0.0005 | \ |
| beta1 | Set the exponential decay rate for the 1st moment estimates | 0.9 | \ |
| beta2 | Set the exponential decay rate for the 2nd moment estimates | 0.999 | \ |
| decay | Whether to use decay | \ | \ |
2020-06-12 14:07:58 +08:00
| function(decay) | Set the decay function | cosine_decay | Only support cosine_decay |
2020-06-12 12:09:55 +08:00
| step_each_epoch | The number of steps in an epoch. | 20 | Calculation ：
| total_epoch | The number of epochs | 1000 | Consistent with Global.epoch_num |
