2.训练配置
ch_PP-OCRv3_det_cml.yml
Global: character_dict_path: ../mb.txt #自定义字典 debug: false use_gpu: true epoch_num: 500 log_smooth_window: 20 print_batch_step: 10 save_model_dir: ./output/ch_PP-OCR_v3_det/ save_epoch_step: 100 eval_batch_step: - 0 - 400 cal_metric_during_train: false pretrained_model: null checkpoints: null save_inference_dir: null use_visualdl: false infer_img: doc/imgs_en/img_10.jpg save_res_path: ./checkpoints/det_db/predicts_db.txt distributed: true Architecture: name: DistillationModel algorithm: Distillation model_type: det Models: Student: pretrained: model_type: det algorithm: DB Transform: null Backbone: name: MobileNetV3 scale: 0.5 model_name: large disable_se: true Neck: name: RSEFPN out_channels: 96 shortcut: True Head: name: DBHead k: 50 Student2: pretrained: model_type: det algorithm: DB Transform: null Backbone: name: MobileNetV3 scale: 0.5 model_name: large disable_se: true Neck: name: RSEFPN out_channels: 96 shortcut: True Head: name: DBHead k: 50 Teacher: pretrained: freeze_params: true return_all_feats: false model_type: det algorithm: DB Backbone: name: ResNet_vd in_channels: 3 layers: 50 Neck: name: LKPAN out_channels: 256 Head: name: DBHead kernel_list: [7,2,2] k: 50 Loss: name: CombinedLoss loss_config_list: - DistillationDilaDBLoss: weight: 1.0 model_name_pairs: - ["Student", "Teacher"] - ["Student2", "Teacher"] key: maps balance_loss: true main_loss_type: DiceLoss alpha: 5 beta: 10 ohem_ratio: 3 - DistillationDMLLoss: model_name_pairs: - ["Student", "Student2"] maps_name: "thrink_maps" weight: 1.0 model_name_pairs: ["Student", "Student2"] key: maps - DistillationDBLoss: weight: 1.0 model_name_list: ["Student", "Student2"] 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: 5.0e-05 PostProcess: name: DistillationDBPostProcess model_name: ["Student"] key: head_out thresh: 0.3 box_thresh: 0.6 max_candidates: 1000 unclip_ratio: 1.5 Metric: name: DistillationMetric base_metric_name: DetMetric main_indicator: hmean key: "Student" # 数据集 Train: dataset: name: SimpleDataSet data_dir: /home/aistudio/dataset/train/image label_file_list: - /home/aistudio/dataset/train/label.txt ratio_list: [1.0] transforms: - DecodeImage: img_mode: BGR channel_first: false - DetLabelEncode: null - CopyPaste: - 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: null - KeepKeys: keep_keys: - image - threshold_map - threshold_mask - shrink_map - shrink_mask loader: shuffle: true drop_last: false batch_size_per_card: 12 num_workers: 4 # 数据集 Eval: dataset: name: SimpleDataSet data_dir: /home/aistudio/dataset/train/image label_file_list: - /home/aistudio/dataset/train/label.txt transforms: - DecodeImage: # load image img_mode: BGR channel_first: False - DetLabelEncode: # Class handling label - DetResizeForTest: - 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
# 拷贝配置到对应目录 !cp ~/ch_PP-OCRv3_det_cml.yml ~/PaddleOCR/configs/det/ch_PP-OCRv3/
%export CUDA_VISIBLE_DEVICES='0,1,2,3'
# !python tools/train.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml -o Optimizer.base_lr=0.0001 %cd ~/PaddleOCR/ !python3 -m paddle.distributed.launch --ips="localhost" --gpus '0,1,2,3' tools/train.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml -o Optimizer.base_lr=0.0001
[2022/09/13 20:07:57] ppocr INFO: epoch: [121/500], global_step: 2050, lr: 0.000869, dila_dbloss_Student_Teacher: 1.288603, dila_dbloss_Student2_Teacher: 1.283752, loss: 6.739372, dml_thrink_maps_0: 0.002895, db_Student_loss_shrink_maps: 1.410274, db_Student_loss_threshold_maps: 0.388946, db_Student_loss_binary_maps: 0.280807, db_Student2_loss_shrink_maps: 1.414650, db_Student2_loss_threshold_maps: 0.391234, db_Student2_loss_binary_maps: 0.281578, avg_reader_cost: 7.17047 s, avg_batch_cost: 11.48100 s, avg_samples: 12.0, ips: 1.04521 samples/s, eta: 20:44:22
五、识别数据集准备
把det的数据集转换为rec数据集，进行模型训练
# ppocr/utils/gen_label.py # convert the official gt to rec_gt_label.txt %cd ~/PaddleOCR !python ppocr/utils/gen_label.py --mode="rec" --input_path="../dataset/train/train.txt" --output_label="../dataset/train/train_rec_gt_label.txt" !python ppocr/utils/gen_label.py --mode="rec" --input_path="../dataset/train/eval.txt" --output_label="../dataset/train/eval_rec_gt_label.txt"
六、识别模型训练
1.预训练模型下载
%cd ~/PaddleOCR/pretrain_models !https://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_train.tar !tar -xvf ch_PP-OCRv3_rec_train.tar
2.配置训练参数
Global: debug: false use_gpu: true epoch_num: 800 log_smooth_window: 20 print_batch_step: 10 save_model_dir: ./output/rec_ppocr_v3_distillation save_epoch_step: 3 eval_batch_step: [0, 2000] cal_metric_during_train: true # 预训练模型 pretrained_model: pretrain_models/ch_PP-OCRv3_rec_train//best_accuracy.pdparams checkpoints: save_inference_dir: use_visualdl: false infer_img: doc/imgs_words/ch/word_1.jpg # 修改码表 character_dict_path: ../mb.txt max_text_length: &max_text_length 25 infer_mode: false use_space_char: true distributed: true save_res_path: ./output/rec/predicts_ppocrv3_distillation.txt Optimizer: name: Adam beta1: 0.9 beta2: 0.999 lr: name: Piecewise decay_epochs : [700, 800] values : [0.0005, 0.00005] warmup_epoch: 5 regularizer: name: L2 factor: 3.0e-05 Architecture: model_type: &model_type "rec" name: DistillationModel algorithm: Distillation Models: Teacher: pretrained: freeze_params: false return_all_feats: true model_type: *model_type algorithm: SVTR Transform: Backbone: name: MobileNetV1Enhance scale: 0.5 last_conv_stride: [1, 2] last_pool_type: avg Head: name: MultiHead head_list: - CTCHead: Neck: name: svtr dims: 64 depth: 2 hidden_dims: 120 use_guide: True Head: fc_decay: 0.00001 - SARHead: enc_dim: 512 max_text_length: *max_text_length Student: pretrained: freeze_params: false return_all_feats: true model_type: *model_type algorithm: SVTR Transform: Backbone: name: MobileNetV1Enhance scale: 0.5 last_conv_stride: [1, 2] last_pool_type: avg Head: name: MultiHead head_list: - CTCHead: Neck: name: svtr dims: 64 depth: 2 hidden_dims: 120 use_guide: True Head: fc_decay: 0.00001 - SARHead: enc_dim: 512 max_text_length: *max_text_length Loss: name: CombinedLoss loss_config_list: - DistillationDMLLoss: weight: 1.0 act: "softmax" use_log: true model_name_pairs: - ["Student", "Teacher"] key: head_out multi_head: True dis_head: ctc name: dml_ctc - DistillationDMLLoss: weight: 0.5 act: "softmax" use_log: true model_name_pairs: - ["Student", "Teacher"] key: head_out multi_head: True dis_head: sar name: dml_sar - DistillationDistanceLoss: weight: 1.0 mode: "l2" model_name_pairs: - ["Student", "Teacher"] key: backbone_out - DistillationCTCLoss: weight: 1.0 model_name_list: ["Student", "Teacher"] key: head_out multi_head: True - DistillationSARLoss: weight: 1.0 model_name_list: ["Student", "Teacher"] key: head_out multi_head: True PostProcess: name: DistillationCTCLabelDecode model_name: ["Student", "Teacher"] key: head_out multi_head: True Metric: name: DistillationMetric base_metric_name: RecMetric main_indicator: acc key: "Student" ignore_space: False # 修改数据及 Train: dataset: name: SimpleDataSet data_dir: /home/aistudio/dataset/train/image ext_op_transform_idx: 1 label_file_list: - /home/aistudio/dataset/train/train_rec_gt_label.txt transforms: - DecodeImage: img_mode: BGR channel_first: false - RecConAug: prob: 0.5 ext_data_num: 2 image_shape: [48, 320, 3] - RecAug: - MultiLabelEncode: - RecResizeImg: image_shape: [3, 48, 320] - KeepKeys: keep_keys: - image - label_ctc - label_sar - length - valid_ratio loader: shuffle: true batch_size_per_card: 128 drop_last: true num_workers: 4 # 修改数据及 Eval: dataset: name: SimpleDataSet data_dir: /home/aistudio/dataset/train/image ext_op_transform_idx: 1 label_file_list: - /home/aistudio/dataset/train/eval_rec_gt_label.txt transforms: - DecodeImage: img_mode: BGR channel_first: false - MultiLabelEncode: - RecResizeImg: image_shape: [3, 48, 320] - KeepKeys: keep_keys: - image - label_ctc - label_sar - length - valid_ratio loader: shuffle: false drop_last: false batch_size_per_card: 128 num_workers: 4
# 拷贝配置好的文件到指定位置 %cd ~ !cp ~/ch_PP-OCRv3_rec_distillation.yml ~/PaddleOCR/configs/rec/PP-OCRv3/
3.模型训练
%cd ~/PaddleOCR/ #多卡训练，通过--gpus参数指定卡号 !python -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml
七、联推理串
1.模型导出
# 导出检测模型 !python tools/export_model.py -c configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det_cml.yml -o Global.pretrained_model=./my_exps/det/best_accuracy Global.save_inference_dir=./inference/det
# 导出识别模型 !python tools/export_model.py -c configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml -o Global.pretrained_model=./my_exps/rec/best_accuracy Global.save_inference_dir=./inference/rec
2.联推理串
! python tools/infer/predict_system.py \ --det_model_dir=inference/det \ --rec_model_dir=inference/rec \ --image_dir="/home/aistudio/dataset/train/image/image_0.jpg" \ --rec_image_shape=3,48,320 # show img plt.figure(figsize=(10, 8)) img = plt.imread("./inference_results/test.jpg") plt.imshow(img)
