diff --git a/PaddleOCR2Pytorch/ch_ptocr_mobile_v2.0_cls_infer.pth b/PaddleOCR2Pytorch/ch_ptocr_mobile_v2.0_cls_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..437066c9b0454dbd48b677ca501bf84c9048139e --- /dev/null +++ b/PaddleOCR2Pytorch/ch_ptocr_mobile_v2.0_cls_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bfe13860824b3365c0c7f7ccfcddc8ff11645c60051739ff18bc9913f60c98e1 +size 588638 diff --git a/PaddleOCR2Pytorch/ch_ptocr_mobile_v2.0_det_infer.pth b/PaddleOCR2Pytorch/ch_ptocr_mobile_v2.0_det_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..b5d1bf8c23e6310f5ae066a78a200fd6380ae37e --- /dev/null +++ b/PaddleOCR2Pytorch/ch_ptocr_mobile_v2.0_det_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e3d3c418418221bb875bec3a1257c7a07d7f8094abeeed8a32a7eff46854e709 +size 2450576 diff --git a/PaddleOCR2Pytorch/ch_ptocr_mobile_v2.0_rec_infer.pth b/PaddleOCR2Pytorch/ch_ptocr_mobile_v2.0_rec_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..59cfdd0ffe1d7dca210992fda15a6854fc42303e --- /dev/null +++ b/PaddleOCR2Pytorch/ch_ptocr_mobile_v2.0_rec_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc4dbe1231deef508be154dfd882fa47c8856caae9dee534b6d7fb344c9a4abb +size 4406992 diff --git a/PaddleOCR2Pytorch/ch_ptocr_server_v2.0_det_infer.pth b/PaddleOCR2Pytorch/ch_ptocr_server_v2.0_det_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..257f2994b521c2bd097b11d2044430d25805422f --- /dev/null +++ b/PaddleOCR2Pytorch/ch_ptocr_server_v2.0_det_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b549b4ad0b49dde4f555869055bc2b974623fe2138faa544dbd8cf6f7c063c50 +size 49537024 diff --git a/PaddleOCR2Pytorch/ch_ptocr_server_v2.0_rec_infer.pth b/PaddleOCR2Pytorch/ch_ptocr_server_v2.0_rec_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..c0dbcec57c0dbed954b17a934d1067df0b581dba --- /dev/null +++ b/PaddleOCR2Pytorch/ch_ptocr_server_v2.0_rec_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0c315fdb59f16d6cf4250440036553d0036d3dc68fcaf7343f5998588cc42e4a +size 111568653 diff --git a/PaddleOCR2Pytorch/ch_ptocr_v2_det_infer.pth b/PaddleOCR2Pytorch/ch_ptocr_v2_det_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..a0490345f4bd1ef73760d22fc7e537b34cc9a338 --- /dev/null +++ b/PaddleOCR2Pytorch/ch_ptocr_v2_det_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:61d39ce0f175e81d54e6d07d5f9f85a4a6d34d77ca8a1ea18855a7a5eadcadb6 +size 2450576 diff --git a/PaddleOCR2Pytorch/ch_ptocr_v2_rec_infer.pth b/PaddleOCR2Pytorch/ch_ptocr_v2_rec_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..1469ae72dbd0431a9aae426c53dc1782ca772ac9 --- /dev/null +++ b/PaddleOCR2Pytorch/ch_ptocr_v2_rec_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:44eebdf924f59a510c38b17e723ff3740550d637b8022e8de41be8ebec71b2b1 +size 8353351 diff --git a/PaddleOCR2Pytorch/ch_ptocr_v3_det_infer.pth b/PaddleOCR2Pytorch/ch_ptocr_v3_det_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..55ea1a46ad9a5d778432e5ba7b0d04724099dd0a --- /dev/null +++ b/PaddleOCR2Pytorch/ch_ptocr_v3_det_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5e364ffd412f39417db2b4430098cfec7d0f8ed36c859224e3fc036186b91359 +size 2540826 diff --git a/PaddleOCR2Pytorch/ch_ptocr_v3_rec_infer.pth b/PaddleOCR2Pytorch/ch_ptocr_v3_rec_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..ca6b7a9f50833768599be9e2505a54a5fa80380f --- /dev/null +++ b/PaddleOCR2Pytorch/ch_ptocr_v3_rec_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8e172d862bc498de7d88aa8ab15ddcd72a29935861ac27b82328cbc84a268b70 +size 10661864 diff --git a/PaddleOCR2Pytorch/ch_ptocr_v4_det_infer.pth b/PaddleOCR2Pytorch/ch_ptocr_v4_det_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..c88cb73e528f5a46e250f3d96d02fb4598e9dfc1 --- /dev/null +++ b/PaddleOCR2Pytorch/ch_ptocr_v4_det_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:89622c3f3e76b3ac7d10d9434c1f117a7471dba44723885cc04b49932a740d5b +size 14506268 diff --git a/PaddleOCR2Pytorch/ch_ptocr_v4_det_server_infer.pth b/PaddleOCR2Pytorch/ch_ptocr_v4_det_server_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..46e92b2e6152086e0785e502ff98144110070ec9 --- /dev/null +++ b/PaddleOCR2Pytorch/ch_ptocr_v4_det_server_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2ccb02359eb1e56c72ba3cc21d166ed32a517863554f19fc6ec1b663fe4a6652 +size 114030092 diff --git a/PaddleOCR2Pytorch/ch_ptocr_v4_rec_infer.pth b/PaddleOCR2Pytorch/ch_ptocr_v4_rec_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..add292ac8f620c0bea8e9db8755037e1579d8921 --- /dev/null +++ b/PaddleOCR2Pytorch/ch_ptocr_v4_rec_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cb4265bb4300a2487e93e82ccfa1924bf9cd1194c1a202ab17a96b4911c27e0b +size 26919769 diff --git a/PaddleOCR2Pytorch/ch_ptocr_v4_rec_server_infer.pth b/PaddleOCR2Pytorch/ch_ptocr_v4_rec_server_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..c509a4312eefa4ea18613aa5a41a662d15995357 --- /dev/null +++ b/PaddleOCR2Pytorch/ch_ptocr_v4_rec_server_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2c0c9f5180ae3e4d8ea9d3830116ac49900abbb2af3985db02c2bbf484bb0bf9 +size 96808553 diff --git a/PaddleOCR2Pytorch/configs/cls/cls_mv3.yml b/PaddleOCR2Pytorch/configs/cls/cls_mv3.yml new file mode 100644 index 0000000000000000000000000000000000000000..b165bc4830f01f0e63c43b1d6a9635e432dc7605 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/cls/cls_mv3.yml @@ -0,0 +1,96 @@ +Global: + use_gpu: true + epoch_num: 100 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/cls/mv3/ + save_epoch_step: 3 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [0, 1000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + label_list: ['0','180'] + +Architecture: + model_type: cls + algorithm: CLS + Transform: + Backbone: + name: MobileNetV3 + scale: 0.35 + model_name: small + Neck: + Head: + name: ClsHead + class_dim: 2 + +Loss: + name: ClsLoss + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + regularizer: + name: 'L2' + factor: 0 + +PostProcess: + name: ClsPostProcess + +Metric: + name: ClsMetric + main_indicator: acc + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/cls + label_file_list: + - ./train_data/cls/train.txt + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - ClsLabelEncode: # Class handling label + - RecAug: + use_tia: False + - RandAugment: + - ClsResizeImg: + image_shape: [3, 48, 192] + - KeepKeys: + keep_keys: ['image', 'label'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 512 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/cls + label_file_list: + - ./train_data/cls/test.txt + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - ClsLabelEncode: # Class handling label + - ClsResizeImg: + image_shape: [3, 48, 192] + - KeepKeys: + keep_keys: ['image', 'label'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 512 + num_workers: 4 \ No newline at end of file diff --git a/PaddleOCR2Pytorch/configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det.yml b/PaddleOCR2Pytorch/configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det.yml new file mode 100644 index 0000000000000000000000000000000000000000..0e8af776479ea26f834ca9ddc169f80b3982e86d --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/ch_PP-OCRv3/ch_PP-OCRv3_det.yml @@ -0,0 +1,163 @@ +Global: + 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: + model_type: det + algorithm: DB + Transform: + 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 + +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: 5.0e-05 +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: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.txt + ratio_list: [1.0] + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - 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: 8 + num_workers: 4 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - DetResizeForTest: null + - 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 + - shape + - polys + - ignore_tags + loader: + shuffle: false + drop_last: false + batch_size_per_card: 1 + num_workers: 2 diff --git a/PaddleOCR2Pytorch/configs/det/ch_PP-OCRv4/ch_PP-OCRv4_det_cml.yml b/PaddleOCR2Pytorch/configs/det/ch_PP-OCRv4/ch_PP-OCRv4_det_cml.yml new file mode 100644 index 0000000000000000000000000000000000000000..fe582ba5b76208459a47563b9d02bf6aa2593b06 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/ch_PP-OCRv4/ch_PP-OCRv4_det_cml.yml @@ -0,0 +1,235 @@ +Global: + debug: false + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 20 + save_model_dir: ./output/ch_PP-OCRv4 + save_epoch_step: 50 + eval_batch_step: + - 0 + - 1000 + cal_metric_during_train: true + checkpoints: null + pretrained_model: 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: + model_type: det + algorithm: DB + Transform: null + Backbone: + name: PPLCNetNew + scale: 0.75 + pretrained: false + Neck: + name: RSEFPN + out_channels: 96 + shortcut: true + Head: + name: DBHead + k: 50 + Student2: + pretrained: null + model_type: det + algorithm: DB + Transform: null + Backbone: + name: PPLCNetNew + scale: 0.75 + pretrained: true + Neck: + name: RSEFPN + out_channels: 96 + shortcut: true + Head: + name: DBHead + k: 50 + Teacher: + pretrained: https://paddleocr.bj.bcebos.com/PP-OCRv4/chinese/ch_PP-OCRv4_det_cml_teacher_pretrained/teacher.pdparams + 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 + 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 +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.txt + ratio_list: [1.0] + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - 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 + total_epoch: 500 + - MakeShrinkMap: + shrink_ratio: 0.4 + min_text_size: 8 + total_epoch: 500 + - 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: 16 + num_workers: 8 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - DetResizeForTest: null + - 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 + - shape + - polys + - ignore_tags + loader: + shuffle: false + drop_last: false + batch_size_per_card: 1 + num_workers: 2 +profiler_options: null diff --git a/PaddleOCR2Pytorch/configs/det/ch_PP-OCRv4/ch_PP-OCRv4_det_student.yml b/PaddleOCR2Pytorch/configs/det/ch_PP-OCRv4/ch_PP-OCRv4_det_student.yml new file mode 100644 index 0000000000000000000000000000000000000000..39b260cf3ad0a0bb9664a5165f3262962bba072e --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/ch_PP-OCRv4/ch_PP-OCRv4_det_student.yml @@ -0,0 +1,171 @@ +Global: + debug: false + use_gpu: true + epoch_num: &epoch_num 500 + log_smooth_window: 20 + print_batch_step: 100 + save_model_dir: ./output/ch_PP-OCRv4 + save_epoch_step: 10 + eval_batch_step: + - 0 + - 1500 + cal_metric_during_train: false + checkpoints: + pretrained_model: https://paddleocr.bj.bcebos.com/pretrained/PPLCNetV3_x0_75_ocr_det.pdparams + 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: + model_type: det + algorithm: DB + Transform: null + Backbone: + name: PPLCNetV3 + scale: 0.75 + det: True + Neck: + name: RSEFPN + out_channels: 96 + shortcut: True + 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 #(8*8c) + warmup_epoch: 2 + regularizer: + name: L2 + factor: 5.0e-05 + +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: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.txt + ratio_list: [1.0] + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - CopyPaste: null + - 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 + total_epoch: *epoch_num + - MakeShrinkMap: + shrink_ratio: 0.4 + min_text_size: 8 + total_epoch: *epoch_num + - 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: 8 + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - DetResizeForTest: + - 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 + - shape + - polys + - ignore_tags + loader: + shuffle: false + drop_last: false + batch_size_per_card: 1 + num_workers: 2 +profiler_options: null diff --git a/PaddleOCR2Pytorch/configs/det/ch_PP-OCRv4/ch_PP-OCRv4_det_teacher.yml b/PaddleOCR2Pytorch/configs/det/ch_PP-OCRv4/ch_PP-OCRv4_det_teacher.yml new file mode 100644 index 0000000000000000000000000000000000000000..b58af1cc9be9c20372caa1fc94d233c7e14c0773 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/ch_PP-OCRv4/ch_PP-OCRv4_det_teacher.yml @@ -0,0 +1,172 @@ +Global: + debug: false + use_gpu: true + epoch_num: &epoch_num 500 + log_smooth_window: 20 + print_batch_step: 100 + save_model_dir: ./output/ch_PP-OCRv4 + save_epoch_step: 10 + eval_batch_step: + - 0 + - 1500 + cal_metric_during_train: false + checkpoints: + pretrained_model: https://paddleocr.bj.bcebos.com/pretrained/PPHGNet_small_ocr_det.pdparams + 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: + model_type: det + algorithm: DB + Transform: null + Backbone: + name: PPHGNet_small + det: True + Neck: + name: LKPAN + out_channels: 256 + intracl: true + Head: + name: PFHeadLocal + k: 50 + mode: "large" + + +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 #(8*8c) + warmup_epoch: 2 + regularizer: + name: L2 + factor: 1e-6 + +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: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.txt + ratio_list: [1.0] + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - CopyPaste: null + - 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 + total_epoch: *epoch_num + - MakeShrinkMap: + shrink_ratio: 0.4 + min_text_size: 8 + total_epoch: *epoch_num + - 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: 8 + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - DetResizeForTest: + - 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 + - shape + - polys + - ignore_tags + loader: + shuffle: false + drop_last: false + batch_size_per_card: 1 + num_workers: 2 +profiler_options: null diff --git a/PaddleOCR2Pytorch/configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml b/PaddleOCR2Pytorch/configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml new file mode 100644 index 0000000000000000000000000000000000000000..fd88495928b18c300386c6a9fd0cf57d840db21e --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml @@ -0,0 +1,134 @@ +Global: + use_gpu: true + epoch_num: 1200 + log_smooth_window: 20 + print_batch_step: 2 + save_model_dir: ./output/ch_db_mv3/ + save_epoch_step: 1200 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [3000, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + load_static_weights: True + 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 + disable_se: True + Neck: + name: DBFPN + out_channels: 96 + 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: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.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: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.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 diff --git a/PaddleOCR2Pytorch/configs/det/ch_ppocr_v2.0/ch_det_res18_db_v2.0.yml b/PaddleOCR2Pytorch/configs/det/ch_ppocr_v2.0/ch_det_res18_db_v2.0.yml new file mode 100644 index 0000000000000000000000000000000000000000..52d6d4ff3dde5f0ca01aff1f043e9a0d1f99fd17 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/ch_ppocr_v2.0/ch_det_res18_db_v2.0.yml @@ -0,0 +1,133 @@ +Global: + use_gpu: true + 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] + # if pretrained_model is saved in static mode, load_static_weights must set to True + load_static_weights: True + cal_metric_during_train: False + pretrained_model: ./pretrain_models/ResNet18_vd_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: 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: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.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: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.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 diff --git a/PaddleOCR2Pytorch/configs/det/det_mv3_db.yml b/PaddleOCR2Pytorch/configs/det/det_mv3_db.yml new file mode 100644 index 0000000000000000000000000000000000000000..00a16b5a38f2a271825a470684be401944eafec0 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/det_mv3_db.yml @@ -0,0 +1,133 @@ +Global: + use_gpu: true + 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] + # if pretrained_model is saved in static mode, load_static_weights must set to True + load_static_weights: True + 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: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.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: False + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.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: False \ No newline at end of file diff --git a/PaddleOCR2Pytorch/configs/det/det_mv3_east.yml b/PaddleOCR2Pytorch/configs/det/det_mv3_east.yml new file mode 100644 index 0000000000000000000000000000000000000000..05581a761cc91abce6d1e83bcdd7aacef1349148 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/det_mv3_east.yml @@ -0,0 +1,111 @@ +Global: + use_gpu: true + epoch_num: 10000 + log_smooth_window: 20 + print_batch_step: 2 + save_model_dir: ./output/east_mv3/ + save_epoch_step: 1000 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [4000, 5000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + load_static_weights: True + cal_metric_during_train: False + pretrained_model: ./pretrain_models/MobileNetV3_large_x0_5_pretrained + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: + save_res_path: ./output/det_east/predicts_east.txt + +Architecture: + model_type: det + algorithm: EAST + Transform: + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: large + Neck: + name: EASTFPN + model_name: small + Head: + name: EASTHead + model_name: small + +Loss: + name: EASTLoss + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + # name: Cosine + learning_rate: 0.001 + # warmup_epoch: 0 + regularizer: + name: 'L2' + factor: 0 + +PostProcess: + name: EASTPostProcess + score_thresh: 0.8 + cover_thresh: 0.1 + nms_thresh: 0.2 + +Metric: + name: DetMetric + main_indicator: hmean + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.txt + ratio_list: [1.0] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - DetLabelEncode: # Class handling label + - EASTProcessTrain: + image_shape: [512, 512] + background_ratio: 0.125 + min_crop_side_ratio: 0.1 + min_text_size: 10 + - KeepKeys: + keep_keys: ['image', 'score_map', 'geo_map', 'training_mask'] # dataloader will return list in this order + loader: + shuffle: True + drop_last: False + batch_size_per_card: 16 + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.txt + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - DetLabelEncode: # Class handling label + - DetResizeForTest: + limit_side_len: 2400 + limit_type: max + - 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 \ No newline at end of file diff --git a/PaddleOCR2Pytorch/configs/det/det_mv3_pse.yml b/PaddleOCR2Pytorch/configs/det/det_mv3_pse.yml new file mode 100644 index 0000000000000000000000000000000000000000..61ac24727acbd4f0b1eea15af08c0f9e71ce95a3 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/det_mv3_pse.yml @@ -0,0 +1,135 @@ +Global: + use_gpu: true + epoch_num: 600 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/det_mv3_pse/ + save_epoch_step: 600 + # evaluation is run every 63 iterations + eval_batch_step: [ 0,63 ] + cal_metric_during_train: False + pretrained_model: ./pretrain_models/MobileNetV3_large_x0_5_pretrained + checkpoints: #./output/det_r50_vd_pse_batch8_ColorJitter/best_accuracy + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_en/img_10.jpg + save_res_path: ./output/det_pse/predicts_pse.txt + +Architecture: + model_type: det + algorithm: PSE + Transform: null + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: large + Neck: + name: FPN + out_channels: 96 + Head: + name: PSEHead + hidden_dim: 96 + out_channels: 7 + +Loss: + name: PSELoss + alpha: 0.7 + ohem_ratio: 3 + kernel_sample_mask: pred + reduction: none + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Step + learning_rate: 0.001 + step_size: 200 + gamma: 0.1 + regularizer: + name: 'L2' + factor: 0.0005 + +PostProcess: + name: PSEPostProcess + thresh: 0 + box_thresh: 0.85 + min_area: 16 + box_type: box # 'box' or 'poly' + scale: 1 + +Metric: + name: DetMetric + main_indicator: hmean + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.txt + ratio_list: [ 1.0 ] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - DetLabelEncode: # Class handling label + - ColorJitter: + brightness: 0.12549019607843137 + saturation: 0.5 + - IaaAugment: + augmenter_args: + - { 'type': Resize, 'args': { 'size': [ 0.5, 3 ] } } + - { 'type': Fliplr, 'args': { 'p': 0.5 } } + - { 'type': Affine, 'args': { 'rotate': [ -10, 10 ] } } + - MakePseGt: + kernel_num: 7 + min_shrink_ratio: 0.4 + size: 640 + - RandomCropImgMask: + size: [ 640,640 ] + main_key: gt_text + crop_keys: [ 'image', 'gt_text', 'gt_kernels', 'mask' ] + - 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', 'gt_text', 'gt_kernels', 'mask' ] # the order of the dataloader list + loader: + shuffle: True + drop_last: False + batch_size_per_card: 16 + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.txt + ratio_list: [ 1.0 ] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - DetLabelEncode: # Class handling label + - DetResizeForTest: + limit_side_len: 736 + limit_type: min + - 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 \ No newline at end of file diff --git a/PaddleOCR2Pytorch/configs/det/det_ppocr_v3.yml b/PaddleOCR2Pytorch/configs/det/det_ppocr_v3.yml new file mode 100644 index 0000000000000000000000000000000000000000..0e8af776479ea26f834ca9ddc169f80b3982e86d --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/det_ppocr_v3.yml @@ -0,0 +1,163 @@ +Global: + 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: + model_type: det + algorithm: DB + Transform: + 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 + +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: 5.0e-05 +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: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.txt + ratio_list: [1.0] + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - 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: 8 + num_workers: 4 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - DetResizeForTest: null + - 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 + - shape + - polys + - ignore_tags + loader: + shuffle: false + drop_last: false + batch_size_per_card: 1 + num_workers: 2 diff --git a/PaddleOCR2Pytorch/configs/det/det_r50_db++_icdar15.yml b/PaddleOCR2Pytorch/configs/det/det_r50_db++_icdar15.yml new file mode 100644 index 0000000000000000000000000000000000000000..e0cd6012b660573a79ff013a1b6e2309074a3d86 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/det_r50_db++_icdar15.yml @@ -0,0 +1,163 @@ +Global: + debug: false + use_gpu: true + epoch_num: 1000 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/det_r50_icdar15/ + save_epoch_step: 200 + eval_batch_step: + - 0 + - 2000 + cal_metric_during_train: false + pretrained_model: ./pretrain_models/ResNet50_dcn_asf_synthtext_pretrained + 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 +Architecture: + model_type: det + algorithm: DB++ + Transform: null + Backbone: + name: ResNet + layers: 50 + dcn_stage: [False, True, True, True] + Neck: + name: DBFPN + out_channels: 256 + use_asf: True + Head: + name: DBHead + k: 50 +Loss: + name: DBLoss + balance_loss: true + main_loss_type: BCELoss + alpha: 5 + beta: 10 + ohem_ratio: 3 +Optimizer: + name: Momentum + momentum: 0.9 + lr: + name: DecayLearningRate + learning_rate: 0.007 + epochs: 1000 + factor: 0.9 + end_lr: 0 + weight_decay: 0.0001 +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: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.txt + ratio_list: + - 1.0 + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - 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: 10 + keep_ratio: true + - MakeShrinkMap: + shrink_ratio: 0.4 + min_text_size: 8 + - MakeBorderMap: + shrink_ratio: 0.4 + thresh_min: 0.3 + thresh_max: 0.7 + - NormalizeImage: + scale: 1./255. + mean: + - 0.48109378172549 + - 0.45752457890196 + - 0.40787054090196 + std: + - 1.0 + - 1.0 + - 1.0 + 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: 4 + num_workers: 8 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - DetResizeForTest: + image_shape: + - 1152 + - 2048 + - NormalizeImage: + scale: 1./255. + mean: + - 0.48109378172549 + - 0.45752457890196 + - 0.40787054090196 + std: + - 1.0 + - 1.0 + - 1.0 + order: hwc + - ToCHWImage: null + - KeepKeys: + keep_keys: + - image + - shape + - polys + - ignore_tags + loader: + shuffle: false + drop_last: false + batch_size_per_card: 1 + num_workers: 2 +profiler_options: null diff --git a/PaddleOCR2Pytorch/configs/det/det_r50_db++_td_tr.yml b/PaddleOCR2Pytorch/configs/det/det_r50_db++_td_tr.yml new file mode 100644 index 0000000000000000000000000000000000000000..65021bb66184381ba732980ac1b7a65d7bd3a355 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/det_r50_db++_td_tr.yml @@ -0,0 +1,166 @@ +Global: + debug: false + use_gpu: true + epoch_num: 1000 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/det_r50_td_tr/ + save_epoch_step: 200 + eval_batch_step: + - 0 + - 2000 + cal_metric_during_train: false + pretrained_model: ./pretrain_models/ResNet50_dcn_asf_synthtext_pretrained + 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 +Architecture: + model_type: det + algorithm: DB++ + Transform: null + Backbone: + name: ResNet + layers: 50 + dcn_stage: [False, True, True, True] + Neck: + name: DBFPN + out_channels: 256 + use_asf: True + Head: + name: DBHead + k: 50 +Loss: + name: DBLoss + balance_loss: true + main_loss_type: BCELoss + alpha: 5 + beta: 10 + ohem_ratio: 3 +Optimizer: + name: Momentum + momentum: 0.9 + lr: + name: DecayLearningRate + learning_rate: 0.007 + epochs: 1000 + factor: 0.9 + end_lr: 0 + weight_decay: 0.0001 +PostProcess: + name: DBPostProcess + thresh: 0.3 + box_thresh: 0.5 + max_candidates: 1000 + unclip_ratio: 1.5 +Metric: + name: DetMetric + main_indicator: hmean +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: + - ./train_data/TD_TR/TD500/train_gt_labels.txt + - ./train_data/TD_TR/TR400/gt_labels.txt + ratio_list: + - 1.0 + - 1.0 + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - 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: 10 + keep_ratio: true + - MakeShrinkMap: + shrink_ratio: 0.4 + min_text_size: 8 + - MakeBorderMap: + shrink_ratio: 0.4 + thresh_min: 0.3 + thresh_max: 0.7 + - NormalizeImage: + scale: 1./255. + mean: + - 0.48109378172549 + - 0.45752457890196 + - 0.40787054090196 + std: + - 1.0 + - 1.0 + - 1.0 + 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: 4 + num_workers: 8 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: + - ./train_data/TD_TR/TD500/test_gt_labels.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - DetLabelEncode: null + - DetResizeForTest: + image_shape: + - 736 + - 736 + keep_ratio: True + - NormalizeImage: + scale: 1./255. + mean: + - 0.48109378172549 + - 0.45752457890196 + - 0.40787054090196 + std: + - 1.0 + - 1.0 + - 1.0 + order: hwc + - ToCHWImage: null + - KeepKeys: + keep_keys: + - image + - shape + - polys + - ignore_tags + loader: + shuffle: false + drop_last: false + batch_size_per_card: 1 + num_workers: 2 +profiler_options: null diff --git a/PaddleOCR2Pytorch/configs/det/det_r50_vd_db.yml b/PaddleOCR2Pytorch/configs/det/det_r50_vd_db.yml new file mode 100644 index 0000000000000000000000000000000000000000..b1c4a04fa1b78d606e5235dc88feb7b6a087de01 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/det_r50_vd_db.yml @@ -0,0 +1,130 @@ +Global: + use_gpu: true + epoch_num: 1200 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/det_r50_vd/ + save_epoch_step: 1200 + # evaluation is run every 2000 iterations + eval_batch_step: [0,2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + load_static_weights: True + cal_metric_during_train: False + pretrained_model: ./pretrain_models/ResNet50_vd_ssld_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: ResNet_vd + layers: 50 + 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: 0.7 + max_candidates: 1000 + unclip_ratio: 1.5 + +Metric: + name: DetMetric + main_indicator: hmean + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.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 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.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 \ No newline at end of file diff --git a/PaddleOCR2Pytorch/configs/det/det_r50_vd_dcn_fce_ctw.yml b/PaddleOCR2Pytorch/configs/det/det_r50_vd_dcn_fce_ctw.yml new file mode 100644 index 0000000000000000000000000000000000000000..3a4075b322a173796f26ebdbe5b83ba98e98e72d --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/det_r50_vd_dcn_fce_ctw.yml @@ -0,0 +1,139 @@ +Global: + use_gpu: true + epoch_num: 1500 + log_smooth_window: 20 + print_batch_step: 20 + save_model_dir: ./output/det_r50_dcn_fce_ctw/ + save_epoch_step: 100 + # evaluation is run every 835 iterations + eval_batch_step: [0, 835] + cal_metric_during_train: False + pretrained_model: ./pretrain_models/ResNet50_vd_ssld_pretrained + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_en/img_10.jpg + save_res_path: ./output/det_fce/predicts_fce.txt + + +Architecture: + model_type: det + algorithm: FCE + Transform: + Backbone: + name: ResNet_vd + layers: 50 + dcn_stage: [False, True, True, True] + out_indices: [1,2,3] + Neck: + name: FCEFPN + out_channels: 256 + has_extra_convs: False + extra_stage: 0 + Head: + name: FCEHead + fourier_degree: 5 +Loss: + name: FCELoss + fourier_degree: 5 + num_sample: 50 + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + learning_rate: 0.0001 + regularizer: + name: 'L2' + factor: 0 + +PostProcess: + name: FCEPostProcess + scales: [8, 16, 32] + alpha: 1.0 + beta: 1.0 + fourier_degree: 5 + box_type: 'poly' + +Metric: + name: DetFCEMetric + main_indicator: hmean + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ctw1500/imgs/ + label_file_list: + - ./train_data/ctw1500/imgs/training.txt + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + ignore_orientation: True + - DetLabelEncode: # Class handling label + - ColorJitter: + brightness: 0.142 + saturation: 0.5 + contrast: 0.5 + - RandomScaling: + - RandomCropFlip: + crop_ratio: 0.5 + - RandomCropPolyInstances: + crop_ratio: 0.8 + min_side_ratio: 0.3 + - RandomRotatePolyInstances: + rotate_ratio: 0.5 + max_angle: 30 + pad_with_fixed_color: False + - SquareResizePad: + target_size: 800 + pad_ratio: 0.6 + - IaaAugment: + augmenter_args: + - { 'type': Fliplr, 'args': { 'p': 0.5 } } + - FCENetTargets: + fourier_degree: 5 + - 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', 'p3_maps', 'p4_maps', 'p5_maps'] # dataloader will return list in this order + loader: + shuffle: True + drop_last: False + batch_size_per_card: 6 + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ctw1500/imgs/ + label_file_list: + - ./train_data/ctw1500/imgs/test.txt + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + ignore_orientation: True + - DetLabelEncode: # Class handling label + - DetResizeForTest: + limit_type: 'min' + limit_side_len: 736 + - NormalizeImage: + scale: 1./255. + mean: [0.485, 0.456, 0.406] + std: [0.229, 0.224, 0.225] + order: 'hwc' + - Pad: + - 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 \ No newline at end of file diff --git a/PaddleOCR2Pytorch/configs/det/det_r50_vd_east.yml b/PaddleOCR2Pytorch/configs/det/det_r50_vd_east.yml new file mode 100644 index 0000000000000000000000000000000000000000..745a26691d89c1cb002f76c09396f5a58010b521 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/det_r50_vd_east.yml @@ -0,0 +1,110 @@ +Global: + use_gpu: true + epoch_num: 10000 + log_smooth_window: 20 + print_batch_step: 2 + save_model_dir: ./output/east_r50_vd/ + save_epoch_step: 1000 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [4000, 5000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + load_static_weights: True + cal_metric_during_train: False + pretrained_model: ./pretrain_models/ResNet50_vd_pretrained/ + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: + save_res_path: ./output/det_east/predicts_east.txt + +Architecture: + model_type: det + algorithm: EAST + Transform: + Backbone: + name: ResNet_vd + layers: 50 + Neck: + name: EASTFPN + model_name: large + Head: + name: EASTHead + model_name: large + +Loss: + name: EASTLoss + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + # name: Cosine + learning_rate: 0.001 + # warmup_epoch: 0 + regularizer: + name: 'L2' + factor: 0 + +PostProcess: + name: EASTPostProcess + score_thresh: 0.8 + cover_thresh: 0.1 + nms_thresh: 0.2 + +Metric: + name: DetMetric + main_indicator: hmean + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.txt + ratio_list: [1.0] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - DetLabelEncode: # Class handling label + - EASTProcessTrain: + image_shape: [512, 512] + background_ratio: 0.125 + min_crop_side_ratio: 0.1 + min_text_size: 10 + - KeepKeys: + keep_keys: ['image', 'score_map', 'geo_map', 'training_mask'] # dataloader will return list in this order + loader: + shuffle: True + drop_last: False + batch_size_per_card: 8 + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.txt + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - DetLabelEncode: # Class handling label + - DetResizeForTest: + limit_side_len: 2400 + limit_type: max + - 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 \ No newline at end of file diff --git a/PaddleOCR2Pytorch/configs/det/det_r50_vd_pse.yml b/PaddleOCR2Pytorch/configs/det/det_r50_vd_pse.yml new file mode 100644 index 0000000000000000000000000000000000000000..ac007183775e51b8d7cf55279adcc18fdbf442a4 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/det_r50_vd_pse.yml @@ -0,0 +1,134 @@ +Global: + use_gpu: true + epoch_num: 600 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/det_r50_vd_pse/ + save_epoch_step: 600 + # evaluation is run every 125 iterations + eval_batch_step: [ 0,125 ] + cal_metric_during_train: False + pretrained_model: ./pretrain_models/ResNet50_vd_ssld_pretrained + checkpoints: #./output/det_r50_vd_pse_batch8_ColorJitter/best_accuracy + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_en/img_10.jpg + save_res_path: ./output/det_pse/predicts_pse.txt + +Architecture: + model_type: det + algorithm: PSE + Transform: + Backbone: + name: ResNet_vd + layers: 50 + Neck: + name: FPN + out_channels: 256 + Head: + name: PSEHead + hidden_dim: 256 + out_channels: 7 + +Loss: + name: PSELoss + alpha: 0.7 + ohem_ratio: 3 + kernel_sample_mask: pred + reduction: none + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Step + learning_rate: 0.0001 + step_size: 200 + gamma: 0.1 + regularizer: + name: 'L2' + factor: 0.0005 + +PostProcess: + name: PSEPostProcess + thresh: 0 + box_thresh: 0.85 + min_area: 16 + box_type: box # 'box' or 'poly' + scale: 1 + +Metric: + name: DetMetric + main_indicator: hmean + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/train_icdar2015_label.txt + ratio_list: [ 1.0 ] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - DetLabelEncode: # Class handling label + - ColorJitter: + brightness: 0.12549019607843137 + saturation: 0.5 + - IaaAugment: + augmenter_args: + - { 'type': Resize, 'args': { 'size': [ 0.5, 3 ] } } + - { 'type': Fliplr, 'args': { 'p': 0.5 } } + - { 'type': Affine, 'args': { 'rotate': [ -10, 10 ] } } + - MakePseGt: + kernel_num: 7 + min_shrink_ratio: 0.4 + size: 640 + - RandomCropImgMask: + size: [ 640,640 ] + main_key: gt_text + crop_keys: [ 'image', 'gt_text', 'gt_kernels', 'mask' ] + - 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', 'gt_text', 'gt_kernels', 'mask' ] # the order of the dataloader list + loader: + shuffle: True + drop_last: False + batch_size_per_card: 8 + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.txt + ratio_list: [ 1.0 ] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - DetLabelEncode: # Class handling label + - DetResizeForTest: + limit_side_len: 736 + limit_type: min + - 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 \ No newline at end of file diff --git a/PaddleOCR2Pytorch/configs/det/det_r50_vd_sast_icdar15.yml b/PaddleOCR2Pytorch/configs/det/det_r50_vd_sast_icdar15.yml new file mode 100644 index 0000000000000000000000000000000000000000..a989bc8fc754ca88e3bff2de2a6db1060301fdd5 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/det_r50_vd_sast_icdar15.yml @@ -0,0 +1,110 @@ +Global: + use_gpu: true + epoch_num: 5000 + log_smooth_window: 20 + print_batch_step: 2 + save_model_dir: ./output/sast_r50_vd_ic15/ + save_epoch_step: 1000 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [4000, 5000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + load_static_weights: True + cal_metric_during_train: False + pretrained_model: ./pretrain_models/ResNet50_vd_ssld_pretrained/ + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: + save_res_path: ./output/sast_r50_vd_ic15/predicts_sast.txt + +Architecture: + model_type: det + algorithm: SAST + Transform: + Backbone: + name: ResNet_SAST + layers: 50 + Neck: + name: SASTFPN + with_cab: True + Head: + name: SASTHead + +Loss: + name: SASTLoss + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + # name: Cosine + learning_rate: 0.001 + # warmup_epoch: 0 + regularizer: + name: 'L2' + factor: 0 + +PostProcess: + name: SASTPostProcess + score_thresh: 0.5 + sample_pts_num: 2 + nms_thresh: 0.2 + expand_scale: 1.0 + shrink_ratio_of_width: 0.3 + +Metric: + name: DetMetric + main_indicator: hmean + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: [./train_data/icdar2013/train_label_json.txt, ./train_data/icdar2015/train_label_json.txt, ./train_data/icdar17_mlt_latin/train_label_json.txt, ./train_data/coco_text_icdar_4pts/train_label_json.txt] + ratio_list: [0.1, 0.45, 0.3, 0.15] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - DetLabelEncode: # Class handling label + - SASTProcessTrain: + image_shape: [512, 512] + min_crop_side_ratio: 0.3 + min_crop_size: 24 + min_text_size: 4 + max_text_size: 512 + - KeepKeys: + keep_keys: ['image', 'score_map', 'border_map', 'training_mask', 'tvo_map', 'tco_map'] # dataloader will return list in this order + loader: + shuffle: True + drop_last: False + batch_size_per_card: 4 + num_workers: 4 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/icdar2015/text_localization/ + label_file_list: + - ./train_data/icdar2015/text_localization/test_icdar2015_label.txt + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - DetLabelEncode: # Class handling label + - DetResizeForTest: + resize_long: 1536 + - 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 \ No newline at end of file diff --git a/PaddleOCR2Pytorch/configs/det/det_r50_vd_sast_totaltext.yml b/PaddleOCR2Pytorch/configs/det/det_r50_vd_sast_totaltext.yml new file mode 100644 index 0000000000000000000000000000000000000000..e040c4207e497a7bf237a84c9c8d1b7c33a2dde8 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/det/det_r50_vd_sast_totaltext.yml @@ -0,0 +1,110 @@ +Global: + use_gpu: true + epoch_num: 5000 + log_smooth_window: 20 + print_batch_step: 2 + save_model_dir: ./output/sast_r50_vd_tt/ + save_epoch_step: 1000 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [4000, 5000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + load_static_weights: True + cal_metric_during_train: False + pretrained_model: ./pretrain_models/ResNet50_vd_ssld_pretrained/ + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: + save_res_path: ./output/sast_r50_vd_tt/predicts_sast.txt + +Architecture: + model_type: det + algorithm: SAST + Transform: + Backbone: + name: ResNet_SAST + layers: 50 + Neck: + name: SASTFPN + with_cab: True + Head: + name: SASTHead + +Loss: + name: SASTLoss + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + # name: Cosine + learning_rate: 0.001 + # warmup_epoch: 0 + regularizer: + name: 'L2' + factor: 0 + +PostProcess: + name: SASTPostProcess + score_thresh: 0.5 + sample_pts_num: 6 + nms_thresh: 0.2 + expand_scale: 1.2 + shrink_ratio_of_width: 0.2 + +Metric: + name: DetMetric + main_indicator: hmean + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: [./train_data/art_latin_icdar_14pt/train_no_tt_test/train_label_json.txt, ./train_data/total_text_icdar_14pt/train_label_json.txt] + ratio_list: [0.5, 0.5] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - DetLabelEncode: # Class handling label + - SASTProcessTrain: + image_shape: [512, 512] + min_crop_side_ratio: 0.3 + min_crop_size: 24 + min_text_size: 4 + max_text_size: 512 + - KeepKeys: + keep_keys: ['image', 'score_map', 'border_map', 'training_mask', 'tvo_map', 'tco_map'] # dataloader will return list in this order + loader: + shuffle: True + drop_last: False + batch_size_per_card: 4 + num_workers: 4 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: + - ./train_data/total_text_icdar_14pt/test_label_json.txt + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - DetLabelEncode: # Class handling label + - DetResizeForTest: + resize_long: 768 + - 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 \ No newline at end of file diff --git a/PaddleOCR2Pytorch/configs/e2e/e2e_r50_vd_pg.yml b/PaddleOCR2Pytorch/configs/e2e/e2e_r50_vd_pg.yml new file mode 100644 index 0000000000000000000000000000000000000000..0a232f7a4f3b9ca214bbc6fd1840cec186c027e4 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/e2e/e2e_r50_vd_pg.yml @@ -0,0 +1,114 @@ +Global: + use_gpu: True + epoch_num: 600 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/pgnet_r50_vd_totaltext/ + save_epoch_step: 10 + # evaluation is run every 0 iterationss after the 1000th iteration + eval_batch_step: [ 0, 1000 ] + # 1. If pretrained_model is saved in static mode, such as classification pretrained model + # from static branch, load_static_weights must be set as True. + # 2. If you want to finetune the pretrained models we provide in the docs, + # you should set load_static_weights as False. + load_static_weights: False + cal_metric_during_train: False + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: + valid_set: totaltext # two mode: totaltext valid curved words, partvgg valid non-curved words + save_res_path: ./output/pgnet_r50_vd_totaltext/predicts_pgnet.txt + character_dict_path: ppocr/utils/ic15_dict.txt + character_type: EN + max_text_length: 50 # the max length in seq + max_text_nums: 30 # the max seq nums in a pic + tcl_len: 64 + +Architecture: + model_type: e2e + algorithm: PGNet + Transform: + Backbone: + name: ResNet + layers: 50 + Neck: + name: PGFPN + Head: + name: PGHead + +Loss: + name: PGLoss + tcl_bs: 64 + max_text_length: 50 # the same as Global: max_text_length + max_text_nums: 30 # the same as Global：max_text_nums + pad_num: 36 # the length of dict for pad + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + learning_rate: 0.001 + regularizer: + name: 'L2' + factor: 0 + + +PostProcess: + name: PGPostProcess + score_thresh: 0.5 +Metric: + name: E2EMetric + character_dict_path: ppocr/utils/ic15_dict.txt + main_indicator: f_score_e2e + +Train: + dataset: + name: PGDataSet + label_file_list: [.././train_data/total_text/train/] + ratio_list: [1.0] + data_format: icdar #two data format: icdar/textnet + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - PGProcessTrain: + batch_size: 14 # same as loader: batch_size_per_card + min_crop_size: 24 + min_text_size: 4 + max_text_size: 512 + - KeepKeys: + keep_keys: [ 'images', 'tcl_maps', 'tcl_label_maps', 'border_maps','direction_maps', 'training_masks', 'label_list', 'pos_list', 'pos_mask' ] # dataloader will return list in this order + loader: + shuffle: True + drop_last: True + batch_size_per_card: 14 + num_workers: 16 + +Eval: + dataset: + name: PGDataSet + data_dir: ./train_data/ + label_file_list: [./train_data/total_text/test/] + transforms: + - DecodeImage: # load image + img_mode: RGB + channel_first: False + - E2ELabelEncode: + - E2EResizeForTest: + max_side_len: 768 + - 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', 'strs', 'tags' ] + loader: + shuffle: False + drop_last: False + batch_size_per_card: 1 # must be 1 + num_workers: 2 \ No newline at end of file diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..2d4261f94f9d930f560df56abc8252947c22fb0f --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/ch_PP-OCRv3_rec.yml @@ -0,0 +1,126 @@ +Global: + debug: false + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_ppocr_v3 + save_epoch_step: 3 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/ppocr_keys_v1.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + Backbone: + name: MobileNetV1Enhance + scale: 0.5 + last_conv_stride: [1, 2] + last_pool_type: avg + Neck: + name: SequenceEncoder + encoder_type: svtr + dims: 64 + depth: 2 + hidden_dims: 120 + use_guide: True + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - SARLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + ignore_space: False + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.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 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.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 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml new file mode 100644 index 0000000000000000000000000000000000000000..e7cbae59a14af73639e1a74a14021b9b2ef60057 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/ch_PP-OCRv3_rec_distillation.yml @@ -0,0 +1,205 @@ +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 + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/ppocr_keys_v1.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 + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.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 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.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 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/en_PP-OCRv3_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/en_PP-OCRv3_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..af8b7ba093bec3368d473b8c44ea8fc6c8c6f66a --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/en_PP-OCRv3_rec.yml @@ -0,0 +1,126 @@ +Global: + debug: false + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/v3_en_mobile + save_epoch_step: 3 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/en_dict.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3_en.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + Backbone: + name: MobileNetV1Enhance + scale: 0.5 + last_conv_stride: [1, 2] + last_pool_type: avg + Neck: + name: SequenceEncoder + encoder_type: svtr + dims: 64 + depth: 2 + hidden_dims: 120 + use_guide: True + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - SARLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + ignore_space: False + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.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 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.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 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/.gitkeep b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/.gitkeep new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/arabic_PP-OCRv3_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/arabic_PP-OCRv3_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..55cee5b347e047739bdfbe104f1b9456a89a8b18 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/arabic_PP-OCRv3_rec.yml @@ -0,0 +1,126 @@ +Global: + debug: false + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/v3_arabic_mobile + save_epoch_step: 3 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/dict/arabic_dict.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3_arabic.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + Backbone: + name: MobileNetV1Enhance + scale: 0.5 + last_conv_stride: [1, 2] + last_pool_type: avg + Neck: + name: SequenceEncoder + encoder_type: svtr + dims: 64 + depth: 2 + hidden_dims: 120 + use_guide: True + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - SARLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + ignore_space: False + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.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 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.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 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/chinese_cht_PP-OCRv3_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/chinese_cht_PP-OCRv3_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..135a360f27be0576c338c2bad407779da9a0e33d --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/chinese_cht_PP-OCRv3_rec.yml @@ -0,0 +1,126 @@ +Global: + debug: false + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/v3_chinese_cht_mobile + save_epoch_step: 3 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/dict/chinese_cht_dict.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3_chinese_cht.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + Backbone: + name: MobileNetV1Enhance + scale: 0.5 + last_conv_stride: [1, 2] + last_pool_type: avg + Neck: + name: SequenceEncoder + encoder_type: svtr + dims: 64 + depth: 2 + hidden_dims: 120 + use_guide: True + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - SARLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + ignore_space: False + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.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 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.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 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/cyrillic_PP-OCRv3_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/cyrillic_PP-OCRv3_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..53651520bceed384bf5f04cd2d0a240983a1ab17 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/cyrillic_PP-OCRv3_rec.yml @@ -0,0 +1,126 @@ +Global: + debug: false + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/v3_cyrillic_mobile + save_epoch_step: 3 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/dict/cyrillic_dict.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3_cyrillic.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + Backbone: + name: MobileNetV1Enhance + scale: 0.5 + last_conv_stride: [1, 2] + last_pool_type: avg + Neck: + name: SequenceEncoder + encoder_type: svtr + dims: 64 + depth: 2 + hidden_dims: 120 + use_guide: True + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - SARLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + ignore_space: False + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.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 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.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 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/devanagari_PP-OCRv3_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/devanagari_PP-OCRv3_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..023b4afbb1e8939ce6acc06caec64859aac11a27 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/devanagari_PP-OCRv3_rec.yml @@ -0,0 +1,126 @@ +Global: + debug: false + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/v3_devanagari_mobile + save_epoch_step: 3 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/dict/devanagari_dict.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3_devanagari.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + Backbone: + name: MobileNetV1Enhance + scale: 0.5 + last_conv_stride: [1, 2] + last_pool_type: avg + Neck: + name: SequenceEncoder + encoder_type: svtr + dims: 64 + depth: 2 + hidden_dims: 120 + use_guide: True + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - SARLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + ignore_space: False + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.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 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.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 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/japan_PP-OCRv3_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/japan_PP-OCRv3_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..065b327793f26d1b5a47bb2d4ca9ca8c8eb1a2dd --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/japan_PP-OCRv3_rec.yml @@ -0,0 +1,126 @@ +Global: + debug: false + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/v3_japan_mobile + save_epoch_step: 3 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/dict/japan_dict.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3_japan.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + Backbone: + name: MobileNetV1Enhance + scale: 0.5 + last_conv_stride: [1, 2] + last_pool_type: avg + Neck: + name: SequenceEncoder + encoder_type: svtr + dims: 64 + depth: 2 + hidden_dims: 120 + use_guide: True + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - SARLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + ignore_space: False + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.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 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.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 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/ka_PP-OCRv3_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/ka_PP-OCRv3_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..5edb2980fdfd2d4d65533d6219d8514dbf858cf9 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/ka_PP-OCRv3_rec.yml @@ -0,0 +1,126 @@ +Global: + debug: false + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/v3_ka_mobile + save_epoch_step: 3 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/dict/ka_dict.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3_ka.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + Backbone: + name: MobileNetV1Enhance + scale: 0.5 + last_conv_stride: [1, 2] + last_pool_type: avg + Neck: + name: SequenceEncoder + encoder_type: svtr + dims: 64 + depth: 2 + hidden_dims: 120 + use_guide: True + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - SARLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + ignore_space: False + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.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 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.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 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/korean_PP-OCRv3_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/korean_PP-OCRv3_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..eebeb0970b159110b0edea632e65023272509630 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/korean_PP-OCRv3_rec.yml @@ -0,0 +1,126 @@ +Global: + debug: false + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/v3_korean_mobile + save_epoch_step: 3 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/dict/korean_dict.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3_korean.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + Backbone: + name: MobileNetV1Enhance + scale: 0.5 + last_conv_stride: [1, 2] + last_pool_type: avg + Neck: + name: SequenceEncoder + encoder_type: svtr + dims: 64 + depth: 2 + hidden_dims: 120 + use_guide: True + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - SARLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + ignore_space: False + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.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 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.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 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/latin_PP-OCRv3_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/latin_PP-OCRv3_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..4f8e25bab6bb3b33cf88d27ff777ed2c12fc8756 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/latin_PP-OCRv3_rec.yml @@ -0,0 +1,126 @@ +Global: + debug: false + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/v3_latin_mobile + save_epoch_step: 3 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/dict/latin_dict.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3_latin.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + Backbone: + name: MobileNetV1Enhance + scale: 0.5 + last_conv_stride: [1, 2] + last_pool_type: avg + Neck: + name: SequenceEncoder + encoder_type: svtr + dims: 64 + depth: 2 + hidden_dims: 120 + use_guide: True + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - SARLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + ignore_space: False + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.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 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.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 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/ta_PP-OCRv3_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/ta_PP-OCRv3_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..15b862492bf9af371b6e619900fb3e6485cad3e9 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/ta_PP-OCRv3_rec.yml @@ -0,0 +1,126 @@ +Global: + debug: false + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/v3_ta_mobile + save_epoch_step: 3 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/dict/ta_dict.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3_ta.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + Backbone: + name: MobileNetV1Enhance + scale: 0.5 + last_conv_stride: [1, 2] + last_pool_type: avg + Neck: + name: SequenceEncoder + encoder_type: svtr + dims: 64 + depth: 2 + hidden_dims: 120 + use_guide: True + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - SARLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + ignore_space: False + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.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 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.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 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/te_PP-OCRv3_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/te_PP-OCRv3_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..9074c8d5d9d35a1f51abedc1e35dd6bf512b044b --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv3/multi_language/te_PP-OCRv3_rec.yml @@ -0,0 +1,126 @@ +Global: + debug: false + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/v3_te_mobile + save_epoch_step: 3 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/dict/te_dict.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3_te.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + Backbone: + name: MobileNetV1Enhance + scale: 0.5 + last_conv_stride: [1, 2] + last_pool_type: avg + Neck: + name: SequenceEncoder + encoder_type: svtr + dims: 64 + depth: 2 + hidden_dims: 120 + use_guide: True + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - SARLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + ignore_space: False + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.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 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.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 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..158815e4aadc507934700de75807e9c8f012cd20 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec.yml @@ -0,0 +1,138 @@ +Global: + debug: false + use_gpu: true + epoch_num: 200 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_ppocr_v4 + save_epoch_step: 10 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: pytorchocr/utils/ppocr_keys_v1.txt + max_text_length: &max_text_length 25 + infer_mode: true + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR_LCNet + Transform: + Backbone: + name: PPLCNetV3 + scale: 0.95 + Head: + name: MultiHead + head_list: + - CTCHead: + Neck: + name: svtr + dims: 120 + depth: 2 + hidden_dims: 120 + kernel_size: [1, 3] + use_guide: True + Head: + fc_decay: 0.00001 + - NRTRHead: + nrtr_dim: 384 + max_text_length: *max_text_length + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - NRTRLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: MultiScaleDataSet + ds_width: false + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - RecConAug: + prob: 0.5 + ext_data_num: 2 + image_shape: [48, 320, 3] + max_text_length: *max_text_length + - RecAug: + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + sampler: + name: MultiScaleSampler + scales: [[320, 32], [320, 48], [320, 64]] + first_bs: &bs 192 + fix_bs: false + divided_factor: [8, 16] # w, h + is_training: True + loader: + shuffle: true + batch_size_per_card: *bs + drop_last: true + num_workers: 8 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - RecResizeImg: + image_shape: [3, 48, 320] + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + loader: + shuffle: false + drop_last: false + batch_size_per_card: 128 + num_workers: 4 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_ampO2_ultra.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_ampO2_ultra.yml new file mode 100644 index 0000000000000000000000000000000000000000..475c551689c3f1e7ba209a68685cb148291ff27a --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_ampO2_ultra.yml @@ -0,0 +1,140 @@ +Global: + debug: false + use_gpu: true + epoch_num: 200 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_ppocr_v4 + save_epoch_step: 10 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/ppocr_keys_v1.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3.txt + use_amp: True + amp_level: O2 + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR_LCNet + Transform: + Backbone: + name: PPLCNetV3 + scale: 0.95 + Head: + name: MultiHead + head_list: + - CTCHead: + Neck: + name: svtr + dims: 120 + depth: 2 + hidden_dims: 120 + kernel_size: [1, 3] + use_guide: True + Head: + fc_decay: 0.00001 + - NRTRHead: + nrtr_dim: 384 + max_text_length: *max_text_length + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - NRTRLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: MultiScaleDataSet + ds_width: false + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - RecConAug: + prob: 0.5 + ext_data_num: 2 + image_shape: [48, 320, 3] + max_text_length: *max_text_length + - RecAug: + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + sampler: + name: MultiScaleSampler + scales: [[320, 32], [320, 48], [320, 64]] + first_bs: &bs 384 + fix_bs: false + divided_factor: [8, 16] # w, h + is_training: True + loader: + shuffle: true + batch_size_per_card: *bs + drop_last: true + num_workers: 16 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - RecResizeImg: + image_shape: [3, 48, 320] + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + loader: + shuffle: false + drop_last: false + batch_size_per_card: 128 + num_workers: 16 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_ctc.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_ctc.yml new file mode 100644 index 0000000000000000000000000000000000000000..28901edb824741501a4ffb1ac53ab951ca9010d1 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_ctc.yml @@ -0,0 +1,132 @@ +Global: + debug: false + use_gpu: true + epoch_num: 200 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_ppocr_v4 + save_epoch_step: 10 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/ppocr_keys_v1.txt + max_text_length: &max_text_length 25 + infer_mode: true + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR_LCNet + Transform: + Backbone: + name: PPLCNetV3 + scale: 0.95 + Neck: + name: svtr + dims: 120 + depth: 2 + hidden_dims: 120 + kernel_size: [ 1, 3 ] + use_guide: True + Head: + name: CTCHead + fc_decay: 0.00004 + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - NRTRLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: MultiScaleDataSet + ds_width: false + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - RecConAug: + prob: 0.5 + ext_data_num: 2 + image_shape: [48, 320, 3] + max_text_length: *max_text_length + - RecAug: + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + sampler: + name: MultiScaleSampler + scales: [[320, 32], [320, 48], [320, 64]] + first_bs: &bs 192 + fix_bs: false + divided_factor: [8, 16] # w, h + is_training: True + loader: + shuffle: true + batch_size_per_card: *bs + drop_last: true + num_workers: 8 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - RecResizeImg: + image_shape: [3, 48, 320] + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + loader: + shuffle: false + drop_last: false + batch_size_per_card: 128 + num_workers: 4 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_distill.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_distill.yml new file mode 100644 index 0000000000000000000000000000000000000000..f613ee52b467f279e4bbbd33dca3c58862f4715a --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_distill.yml @@ -0,0 +1,231 @@ +Global: + debug: false + use_gpu: true + epoch_num: 200 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_dkd_400w_svtr_ctc_lcnet_blank_dkd0.1/ + save_epoch_step: 40 + eval_batch_step: + - 0 + - 2000 + cal_metric_during_train: true + pretrained_model: null + checkpoints: ./output/rec_dkd_400w_svtr_ctc_lcnet_blank_dkd0.1/latest + save_inference_dir: null + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/ppocr_keys_v1.txt + max_text_length: 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3.txt +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 2 + regularizer: + name: L2 + factor: 3.0e-05 +Architecture: + model_type: rec + name: DistillationModel + algorithm: Distillation + Models: + Teacher: + pretrained: + freeze_params: true + return_all_feats: true + model_type: rec + algorithm: SVTR + Transform: null + Backbone: + name: SVTRNet + img_size: + - 48 + - 320 + out_char_num: 40 + out_channels: 192 + patch_merging: Conv + embed_dim: + - 64 + - 128 + - 256 + depth: + - 3 + - 6 + - 3 + num_heads: + - 2 + - 4 + - 8 + mixer: + - Conv + - Conv + - Conv + - Conv + - Conv + - Conv + - Global + - Global + - Global + - Global + - Global + - Global + local_mixer: + - - 5 + - 5 + - - 5 + - 5 + - - 5 + - 5 + last_stage: false + prenorm: true + Head: + name: MultiHead + head_list: + - CTCHead: + Neck: + name: svtr + dims: 120 + depth: 2 + hidden_dims: 120 + kernel_size: [1, 3] + use_guide: True + Head: + fc_decay: 0.00001 + - NRTRHead: + nrtr_dim: 384 + max_text_length: *max_text_length + Student: + pretrained: + freeze_params: false + return_all_feats: true + model_type: rec + algorithm: SVTR + Transform: null + Backbone: + name: PPLCNetV3 + scale: 0.95 + Head: + name: MultiHead + head_list: + - CTCHead: + Neck: + name: svtr + dims: 120 + depth: 2 + hidden_dims: 120 + kernel_size: [1, 3] + use_guide: True + Head: + fc_decay: 0.00001 + - NRTRHead: + nrtr_dim: 384 + max_text_length: *max_text_length +Loss: + name: CombinedLoss + loss_config_list: + - DistillationDKDLoss: + weight: 0.1 + model_name_pairs: + - - Student + - Teacher + key: head_out + multi_head: true + alpha: 1.0 + beta: 2.0 + dis_head: gtc + name: dkd + - DistillationCTCLoss: + weight: 1.0 + model_name_list: + - Student + key: head_out + multi_head: true + - DistillationNRTRLoss: + weight: 1.0 + smoothing: false + model_name_list: + - Student + key: head_out + multi_head: true + - DistillCTCLogits: + weight: 1.0 + reduction: mean + model_name_pairs: + - - Student + - Teacher + key: head_out +PostProcess: + name: DistillationCTCLabelDecode + model_name: + - Student + key: head_out + multi_head: true +Metric: + name: DistillationMetric + base_metric_name: RecMetric + main_indicator: acc + key: Student + ignore_space: false +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: + - ./train_data/train_list.txt + ratio_list: + - 1.0 + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - RecAug: + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + loader: + shuffle: true + batch_size_per_card: 128 + drop_last: true + num_workers: 8 + use_shared_memory: true +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - RecResizeImg: + image_shape: [3, 48, 320] + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + loader: + shuffle: false + drop_last: false + batch_size_per_card: 128 + num_workers: 4 +profiler_options: null diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_fp32_ultra.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_fp32_ultra.yml new file mode 100644 index 0000000000000000000000000000000000000000..8c267309daefbe45929551ed186c7c5e7265a066 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_fp32_ultra.yml @@ -0,0 +1,138 @@ +Global: + debug: false + use_gpu: true + epoch_num: 200 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_ppocr_v4 + save_epoch_step: 10 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/ppocr_keys_v1.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR_LCNet + Transform: + Backbone: + name: PPLCNetV3 + scale: 0.95 + Head: + name: MultiHead + head_list: + - CTCHead: + Neck: + name: svtr + dims: 120 + depth: 2 + hidden_dims: 120 + kernel_size: [1, 3] + use_guide: True + Head: + fc_decay: 0.00001 + - NRTRHead: + nrtr_dim: 384 + max_text_length: *max_text_length + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - NRTRLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: MultiScaleDataSet + ds_width: false + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - RecConAug: + prob: 0.5 + ext_data_num: 2 + image_shape: [48, 320, 3] + max_text_length: *max_text_length + - RecAug: + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + sampler: + name: MultiScaleSampler + scales: [[320, 32], [320, 48], [320, 64]] + first_bs: &bs 192 + fix_bs: false + divided_factor: [8, 16] # w, h + is_training: True + loader: + shuffle: true + batch_size_per_card: *bs + drop_last: true + num_workers: 16 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - RecResizeImg: + image_shape: [3, 48, 320] + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + loader: + shuffle: false + drop_last: false + batch_size_per_card: 128 + num_workers: 16 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_hgnet.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_hgnet.yml new file mode 100644 index 0000000000000000000000000000000000000000..c181da6d64d0d78a8a0afc9ce680288c10c57b84 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_hgnet.yml @@ -0,0 +1,137 @@ +Global: + debug: false + use_gpu: true + epoch_num: 200 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_ppocr_v4_hgnet + save_epoch_step: 10 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: pytorchocr/utils/ppocr_keys_v1.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR_HGNet + Transform: + Backbone: + name: PPHGNet_small + Head: + name: MultiHead + head_list: + - CTCHead: + Neck: + name: svtr + dims: 120 + depth: 2 + hidden_dims: 120 + kernel_size: [1, 3] + use_guide: True + Head: + fc_decay: 0.00001 + - NRTRHead: + nrtr_dim: 384 + max_text_length: *max_text_length + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - NRTRLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: MultiScaleDataSet + ds_width: false + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - RecConAug: + prob: 0.5 + ext_data_num: 2 + image_shape: [48, 320, 3] + max_text_length: *max_text_length + - RecAug: + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + sampler: + name: MultiScaleSampler + scales: [[320, 32], [320, 48], [320, 64]] + first_bs: &bs 128 + fix_bs: false + divided_factor: [8, 16] # w, h + is_training: True + loader: + shuffle: true + batch_size_per_card: *bs + drop_last: true + num_workers: 8 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - RecResizeImg: + image_shape: [3, 48, 320] + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + loader: + shuffle: false + drop_last: false + batch_size_per_card: 128 + num_workers: 4 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_hgnet_ampO2_ultra.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_hgnet_ampO2_ultra.yml new file mode 100644 index 0000000000000000000000000000000000000000..43035216c3cee092855e5140454c75381f1ee7b5 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_hgnet_ampO2_ultra.yml @@ -0,0 +1,139 @@ +Global: + debug: false + use_gpu: true + epoch_num: 200 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_ppocr_v4_hgnet + save_epoch_step: 10 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/ppocr_keys_v1.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3.txt + use_amp: True + amp_level: O2 + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR_HGNet + Transform: + Backbone: + name: PPHGNet_small + Head: + name: MultiHead + head_list: + - CTCHead: + Neck: + name: svtr + dims: 120 + depth: 2 + hidden_dims: 120 + kernel_size: [1, 3] + use_guide: True + Head: + fc_decay: 0.00001 + - NRTRHead: + nrtr_dim: 384 + max_text_length: *max_text_length + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - NRTRLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: MultiScaleDataSet + ds_width: false + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - RecConAug: + prob: 0.5 + ext_data_num: 2 + image_shape: [48, 320, 3] + max_text_length: *max_text_length + - RecAug: + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + sampler: + name: MultiScaleSampler + scales: [[320, 32], [320, 48], [320, 64]] + first_bs: &bs 256 + fix_bs: false + divided_factor: [8, 16] # w, h + is_training: True + loader: + shuffle: true + batch_size_per_card: *bs + drop_last: true + num_workers: 16 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - RecResizeImg: + image_shape: [3, 48, 320] + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + loader: + shuffle: false + drop_last: false + batch_size_per_card: 128 + num_workers: 16 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_hgnet_fp32_ultra.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_hgnet_fp32_ultra.yml new file mode 100644 index 0000000000000000000000000000000000000000..ee9ebcafd39339c7584e5122361804995834ae8f --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/ch_PP-OCRv4_rec_hgnet_fp32_ultra.yml @@ -0,0 +1,137 @@ +Global: + debug: false + use_gpu: true + epoch_num: 200 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_ppocr_v4_hgnet + save_epoch_step: 10 + eval_batch_step: [0, 2000] + cal_metric_during_train: true + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: ppocr/utils/ppocr_keys_v1.txt + max_text_length: &max_text_length 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3.txt + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 + + +Architecture: + model_type: rec + algorithm: SVTR_HGNet + Transform: + Backbone: + name: PPHGNet_small + Head: + name: MultiHead + head_list: + - CTCHead: + Neck: + name: svtr + dims: 120 + depth: 2 + hidden_dims: 120 + kernel_size: [1, 3] + use_guide: True + Head: + fc_decay: 0.00001 + - NRTRHead: + nrtr_dim: 384 + max_text_length: *max_text_length + +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: + - NRTRLoss: + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: MultiScaleDataSet + ds_width: false + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - RecConAug: + prob: 0.5 + ext_data_num: 2 + image_shape: [48, 320, 3] + max_text_length: *max_text_length + - RecAug: + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + sampler: + name: MultiScaleSampler + scales: [[320, 32], [320, 48], [320, 64]] + first_bs: &bs 256 + fix_bs: false + divided_factor: [8, 16] # w, h + is_training: True + loader: + shuffle: true + batch_size_per_card: *bs + drop_last: true + num_workers: 16 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - RecResizeImg: + image_shape: [3, 48, 320] + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + loader: + shuffle: false + drop_last: false + batch_size_per_card: 128 + num_workers: 16 diff --git a/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/en_PP-OCRv4_rec.yml b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/en_PP-OCRv4_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..b355fb4e4d2f0fc108ca6292f2b4dc344ac71ec4 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/PP-OCRv4/en_PP-OCRv4_rec.yml @@ -0,0 +1,150 @@ +Global: + debug: false + use_gpu: true + epoch_num: 50 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_ppocr_v4 + save_epoch_step: 10 + eval_batch_step: + - 0 + - 2000 + cal_metric_during_train: true + pretrained_model: refactor + checkpoints: null + save_inference_dir: null + use_visualdl: false + infer_img: doc/imgs_words/ch/word_1.jpg + character_dict_path: pytorchocr/utils/en_dict.txt + max_text_length: 25 + infer_mode: false + use_space_char: true + distributed: true + save_res_path: ./output/rec/predicts_ppocrv3.txt +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.0005 + warmup_epoch: 5 + regularizer: + name: L2 + factor: 3.0e-05 +Architecture: + model_type: rec + algorithm: SVTR_LCNet + Transform: null + Backbone: + name: PPLCNetV3 + scale: 0.95 + Head: + name: MultiHead + head_list: + - CTCHead: + Neck: + name: svtr + dims: 120 + depth: 2 + hidden_dims: 120 + kernel_size: + - 1 + - 3 + use_guide: true + Head: + fc_decay: 1.0e-05 + - NRTRHead: + nrtr_dim: 384 + max_text_length: 25 +Loss: + name: MultiLoss + loss_config_list: + - CTCLoss: null + - NRTRLoss: null +PostProcess: + name: CTCLabelDecode +Metric: + name: RecMetric + main_indicator: acc + ignore_space: false +Train: + dataset: + name: MultiScaleDataSet + ds_width: false + data_dir: ./train_data/ + ext_op_transform_idx: 1 + label_file_list: + - ./train_data/train_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - RecConAug: + prob: 0.5 + ext_data_num: 2 + image_shape: + - 48 + - 320 + - 3 + max_text_length: 25 + - RecAug: null + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + sampler: + name: MultiScaleSampler + scales: + - - 320 + - 32 + - - 320 + - 48 + - - 320 + - 64 + first_bs: 96 + fix_bs: false + divided_factor: + - 8 + - 16 + is_training: true + loader: + shuffle: true + batch_size_per_card: 96 + drop_last: true + num_workers: 8 +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: + - ./train_data/val_list.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - MultiLabelEncode: + gtc_encode: NRTRLabelEncode + - RecResizeImg: + image_shape: + - 3 + - 48 + - 320 + - KeepKeys: + keep_keys: + - image + - label_ctc + - label_gtc + - length + - valid_ratio + loader: + shuffle: false + drop_last: false + batch_size_per_card: 128 + num_workers: 4 +profiler_options: null diff --git a/PaddleOCR2Pytorch/configs/rec/ch_ppocr_v2.0/rec_chinese_common_train_v2.0.yml b/PaddleOCR2Pytorch/configs/rec/ch_ppocr_v2.0/rec_chinese_common_train_v2.0.yml new file mode 100644 index 0000000000000000000000000000000000000000..1db3e1cb8633f03f91d1d44064a19f7661e57b12 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/ch_ppocr_v2.0/rec_chinese_common_train_v2.0.yml @@ -0,0 +1,100 @@ +Global: + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_chinese_common_v2.0 + save_epoch_step: 3 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words/ch/word_1.jpg + # for data or label process + character_dict_path: ppocr/utils/ppocr_keys_v1.txt + character_type: ch + max_text_length: 25 + infer_mode: False + use_space_char: True + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + regularizer: + name: 'L2' + factor: 0.00004 + +Architecture: + model_type: rec + algorithm: CRNN + Transform: + Backbone: + name: ResNet + layers: 34 + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 256 + Head: + name: CTCHead + fc_decay: 0.00004 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/train_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - RecAug: + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/val_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml b/PaddleOCR2Pytorch/configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml new file mode 100644 index 0000000000000000000000000000000000000000..dc9d650f30f3d1086616a81c27aaf5db389a1fe7 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml @@ -0,0 +1,102 @@ +Global: + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_chinese_lite_v2.0 + save_epoch_step: 3 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words/ch/word_1.jpg + # for data or label process + character_dict_path: ppocr/utils/ppocr_keys_v1.txt + character_type: ch + max_text_length: 25 + infer_mode: False + use_space_char: True + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + regularizer: + name: 'L2' + factor: 0.00001 + +Architecture: + model_type: rec + algorithm: CRNN + Transform: + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: small + small_stride: [1, 2, 2, 2] + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 48 + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/train_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - RecAug: + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data + label_file_list: ["./train_data/val_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/multi_language/generate_multi_language_configs.py b/PaddleOCR2Pytorch/configs/rec/multi_language/generate_multi_language_configs.py new file mode 100644 index 0000000000000000000000000000000000000000..7a65401d22ec71d69f702dc2ab9d82c24723786f --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/multi_language/generate_multi_language_configs.py @@ -0,0 +1,200 @@ +# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +import yaml +from argparse import ArgumentParser, RawDescriptionHelpFormatter +import os.path +import logging +logging.basicConfig(level=logging.INFO) + +support_list = { + 'it': 'italian', + 'es': 'spanish', + 'pt': 'portuguese', + 'ru': 'russian', + 'ar': 'arabic', + 'ta': 'tamil', + 'ug': 'uyghur', + 'fa': 'persian', + 'ur': 'urdu', + 'rs_latin': 'serbian latin', + 'oc': 'occitan', + 'rs_cyrillic': 'serbian cyrillic', + 'bg': 'bulgarian', + 'uk': 'ukranian', + 'be': 'belarusian', + 'te': 'telugu', + 'kn': 'kannada', + 'ch_tra': 'chinese tradition', + 'hi': 'hindi', + 'mr': 'marathi', + 'ne': 'nepali', +} +assert (os.path.isfile("./rec_multi_language_lite_train.yml") + ), "Loss basic configuration file rec_multi_language_lite_train.yml.\ +You can download it from \ +https://github.com/PaddlePaddle/PaddleOCR/tree/dygraph/configs/rec/multi_language/" + +global_config = yaml.load( + open("./rec_multi_language_lite_train.yml", 'rb'), Loader=yaml.Loader) +project_path = os.path.abspath(os.path.join(os.getcwd(), "../../../")) + + +class ArgsParser(ArgumentParser): + def __init__(self): + super(ArgsParser, self).__init__( + formatter_class=RawDescriptionHelpFormatter) + self.add_argument( + "-o", "--opt", nargs='+', help="set configuration options") + self.add_argument( + "-l", + "--language", + nargs='+', + help="set language type, support {}".format(support_list)) + self.add_argument( + "--train", + type=str, + help="you can use this command to change the train dataset default path" + ) + self.add_argument( + "--val", + type=str, + help="you can use this command to change the eval dataset default path" + ) + self.add_argument( + "--dict", + type=str, + help="you can use this command to change the dictionary default path" + ) + self.add_argument( + "--data_dir", + type=str, + help="you can use this command to change the dataset default root path" + ) + + def parse_args(self, argv=None): + args = super(ArgsParser, self).parse_args(argv) + args.opt = self._parse_opt(args.opt) + args.language = self._set_language(args.language) + return args + + def _parse_opt(self, opts): + config = {} + if not opts: + return config + for s in opts: + s = s.strip() + k, v = s.split('=') + config[k] = yaml.load(v, Loader=yaml.Loader) + return config + + def _set_language(self, type): + assert (type), "please use -l or --language to choose language type" + assert( + type[0] in support_list.keys() + ),"the sub_keys(-l or --language) can only be one of support list: \n{},\nbut get: {}, " \ + "please check your running command".format(support_list, type) + global_config['Global'][ + 'character_dict_path'] = 'ppocr/utils/dict/{}_dict.txt'.format(type[ + 0]) + global_config['Global'][ + 'save_model_dir'] = './output/rec_{}_lite'.format(type[0]) + global_config['Train']['dataset'][ + 'label_file_list'] = ["train_data/{}_train.txt".format(type[0])] + global_config['Eval']['dataset'][ + 'label_file_list'] = ["train_data/{}_val.txt".format(type[0])] + global_config['Global']['character_type'] = type[0] + assert ( + os.path.isfile( + os.path.join(project_path, global_config['Global'][ + 'character_dict_path'])) + ), "Loss default dictionary file {}_dict.txt.You can download it from \ +https://github.com/PaddlePaddle/PaddleOCR/tree/dygraph/ppocr/utils/dict/".format( + type[0]) + return type[0] + + +def merge_config(config): + """ + Merge config into global config. + Args: + config (dict): Config to be merged. + Returns: global config + """ + for key, value in config.items(): + if "." not in key: + if isinstance(value, dict) and key in global_config: + global_config[key].update(value) + else: + global_config[key] = value + else: + sub_keys = key.split('.') + assert ( + sub_keys[0] in global_config + ), "the sub_keys can only be one of global_config: {}, but get: {}, please check your running command".format( + global_config.keys(), sub_keys[0]) + cur = global_config[sub_keys[0]] + for idx, sub_key in enumerate(sub_keys[1:]): + if idx == len(sub_keys) - 2: + cur[sub_key] = value + else: + cur = cur[sub_key] + + +def loss_file(path): + assert ( + os.path.exists(path) + ), "There is no such file:{},Please do not forget to put in the specified file".format( + path) + + +if __name__ == '__main__': + FLAGS = ArgsParser().parse_args() + merge_config(FLAGS.opt) + save_file_path = 'rec_{}_lite_train.yml'.format(FLAGS.language) + if os.path.isfile(save_file_path): + os.remove(save_file_path) + + if FLAGS.train: + global_config['Train']['dataset']['label_file_list'] = [FLAGS.train] + train_label_path = os.path.join(project_path, FLAGS.train) + loss_file(train_label_path) + if FLAGS.val: + global_config['Eval']['dataset']['label_file_list'] = [FLAGS.val] + eval_label_path = os.path.join(project_path, FLAGS.val) + loss_file(eval_label_path) + if FLAGS.dict: + global_config['Global']['character_dict_path'] = FLAGS.dict + dict_path = os.path.join(project_path, FLAGS.dict) + loss_file(dict_path) + if FLAGS.data_dir: + global_config['Eval']['dataset']['data_dir'] = FLAGS.data_dir + global_config['Train']['dataset']['data_dir'] = FLAGS.data_dir + data_dir = os.path.join(project_path, FLAGS.data_dir) + loss_file(data_dir) + + with open(save_file_path, 'w') as f: + yaml.dump( + dict(global_config), f, default_flow_style=False, sort_keys=False) + logging.info("Project path is :{}".format(project_path)) + logging.info("Train list path set to :{}".format(global_config['Train'][ + 'dataset']['label_file_list'][0])) + logging.info("Eval list path set to :{}".format(global_config['Eval'][ + 'dataset']['label_file_list'][0])) + logging.info("Dataset root path set to :{}".format(global_config['Eval'][ + 'dataset']['data_dir'])) + logging.info("Dict path set to :{}".format(global_config['Global'][ + 'character_dict_path'])) + logging.info("Config file set to :configs/rec/multi_language/{}". + format(save_file_path)) diff --git a/PaddleOCR2Pytorch/configs/rec/multi_language/rec_arabic_lite_train.yml b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_arabic_lite_train.yml new file mode 100644 index 0000000000000000000000000000000000000000..6dcfd1b69988b09c7dfc05cdbacce9756ea1f7cb --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_arabic_lite_train.yml @@ -0,0 +1,111 @@ +Global: + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_arabic_lite + save_epoch_step: 3 + eval_batch_step: + - 0 + - 2000 + cal_metric_during_train: true + pretrained_model: null + checkpoints: null + save_inference_dir: null + use_visualdl: false + infer_img: null + character_dict_path: ppocr/utils/dict/arabic_dict.txt + character_type: arabic + max_text_length: 25 + infer_mode: false + use_space_char: true +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + regularizer: + name: L2 + factor: 1.0e-05 +Architecture: + model_type: rec + algorithm: CRNN + Transform: null + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: small + small_stride: + - 1 + - 2 + - 2 + - 2 + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 48 + Head: + name: CTCHead + fc_decay: 1.0e-05 +Loss: + name: CTCLoss +PostProcess: + name: CTCLabelDecode +Metric: + name: RecMetric + main_indicator: acc +Train: + dataset: + name: SimpleDataSet + data_dir: train_data/ + label_file_list: + - train_data/arabic_train.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - RecAug: null + - CTCLabelEncode: null + - RecResizeImg: + image_shape: + - 3 + - 32 + - 320 + - KeepKeys: + keep_keys: + - image + - label + - length + loader: + shuffle: true + batch_size_per_card: 256 + drop_last: true + num_workers: 8 +Eval: + dataset: + name: SimpleDataSet + data_dir: train_data/ + label_file_list: + - train_data/arabic_val.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - CTCLabelEncode: null + - RecResizeImg: + image_shape: + - 3 + - 32 + - 320 + - KeepKeys: + keep_keys: + - image + - label + - length + loader: + shuffle: false + drop_last: false + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/multi_language/rec_cyrillic_lite_train.yml b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_cyrillic_lite_train.yml new file mode 100644 index 0000000000000000000000000000000000000000..52527c1dfb9a306429bbab9241c623581d546e45 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_cyrillic_lite_train.yml @@ -0,0 +1,111 @@ +Global: + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_cyrillic_lite + save_epoch_step: 3 + eval_batch_step: + - 0 + - 2000 + cal_metric_during_train: true + pretrained_model: null + checkpoints: null + save_inference_dir: null + use_visualdl: false + infer_img: null + character_dict_path: ppocr/utils/dict/cyrillic_dict.txt + character_type: cyrillic + max_text_length: 25 + infer_mode: false + use_space_char: true +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + regularizer: + name: L2 + factor: 1.0e-05 +Architecture: + model_type: rec + algorithm: CRNN + Transform: null + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: small + small_stride: + - 1 + - 2 + - 2 + - 2 + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 48 + Head: + name: CTCHead + fc_decay: 1.0e-05 +Loss: + name: CTCLoss +PostProcess: + name: CTCLabelDecode +Metric: + name: RecMetric + main_indicator: acc +Train: + dataset: + name: SimpleDataSet + data_dir: train_data/ + label_file_list: + - train_data/cyrillic_train.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - RecAug: null + - CTCLabelEncode: null + - RecResizeImg: + image_shape: + - 3 + - 32 + - 320 + - KeepKeys: + keep_keys: + - image + - label + - length + loader: + shuffle: true + batch_size_per_card: 256 + drop_last: true + num_workers: 8 +Eval: + dataset: + name: SimpleDataSet + data_dir: train_data/ + label_file_list: + - train_data/cyrillic_val.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - CTCLabelEncode: null + - RecResizeImg: + image_shape: + - 3 + - 32 + - 320 + - KeepKeys: + keep_keys: + - image + - label + - length + loader: + shuffle: false + drop_last: false + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/multi_language/rec_devanagari_lite_train.yml b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_devanagari_lite_train.yml new file mode 100644 index 0000000000000000000000000000000000000000..e1a7c829c3e6d3c3a57f1d501cdd80a560703ec7 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_devanagari_lite_train.yml @@ -0,0 +1,111 @@ +Global: + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_devanagari_lite + save_epoch_step: 3 + eval_batch_step: + - 0 + - 2000 + cal_metric_during_train: true + pretrained_model: null + checkpoints: null + save_inference_dir: null + use_visualdl: false + infer_img: null + character_dict_path: ppocr/utils/dict/devanagari_dict.txt + character_type: devanagari + max_text_length: 25 + infer_mode: false + use_space_char: true +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + regularizer: + name: L2 + factor: 1.0e-05 +Architecture: + model_type: rec + algorithm: CRNN + Transform: null + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: small + small_stride: + - 1 + - 2 + - 2 + - 2 + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 48 + Head: + name: CTCHead + fc_decay: 1.0e-05 +Loss: + name: CTCLoss +PostProcess: + name: CTCLabelDecode +Metric: + name: RecMetric + main_indicator: acc +Train: + dataset: + name: SimpleDataSet + data_dir: train_data/ + label_file_list: + - train_data/devanagari_train.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - RecAug: null + - CTCLabelEncode: null + - RecResizeImg: + image_shape: + - 3 + - 32 + - 320 + - KeepKeys: + keep_keys: + - image + - label + - length + loader: + shuffle: true + batch_size_per_card: 256 + drop_last: true + num_workers: 8 +Eval: + dataset: + name: SimpleDataSet + data_dir: train_data/ + label_file_list: + - train_data/devanagari_val.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - CTCLabelEncode: null + - RecResizeImg: + image_shape: + - 3 + - 32 + - 320 + - KeepKeys: + keep_keys: + - image + - label + - length + loader: + shuffle: false + drop_last: false + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/multi_language/rec_en_number_lite_train.yml b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_en_number_lite_train.yml new file mode 100644 index 0000000000000000000000000000000000000000..13eda8481cad8ca308cd0629214b52146c3ebf13 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_en_number_lite_train.yml @@ -0,0 +1,102 @@ +Global: + use_gpu: True + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_en_number_lite + save_epoch_step: 3 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: + # for data or label process + character_dict_path: ppocr/utils/dict/en_dict.txt + character_type: EN + max_text_length: 25 + infer_mode: False + use_space_char: False + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + regularizer: + name: 'L2' + factor: 0.00001 + +Architecture: + model_type: rec + algorithm: CRNN + Transform: + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: small + small_stride: [1, 2, 2, 2] + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 48 + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/train_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - RecAug: + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/eval_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/multi_language/rec_french_lite_train.yml b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_french_lite_train.yml new file mode 100644 index 0000000000000000000000000000000000000000..63378d38a0d31fc77c33173e0ed864f28c5c3a8b --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_french_lite_train.yml @@ -0,0 +1,102 @@ +Global: + use_gpu: True + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_french_lite + save_epoch_step: 3 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: + # for data or label process + character_dict_path: ppocr/utils/dict/french_dict.txt + character_type: french + max_text_length: 25 + infer_mode: False + use_space_char: False + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + regularizer: + name: 'L2' + factor: 0.00001 + +Architecture: + model_type: rec + algorithm: CRNN + Transform: + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: small + small_stride: [1, 2, 2, 2] + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 48 + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/train_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - RecAug: + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/eval_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/multi_language/rec_german_lite_train.yml b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_german_lite_train.yml new file mode 100644 index 0000000000000000000000000000000000000000..1651510c5e4597e82298135d2f6c64aa747cf961 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_german_lite_train.yml @@ -0,0 +1,102 @@ +Global: + use_gpu: True + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_german_lite + save_epoch_step: 3 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: + # for data or label process + character_dict_path: ppocr/utils/dict/german_dict.txt + character_type: german + max_text_length: 25 + infer_mode: False + use_space_char: False + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + regularizer: + name: 'L2' + factor: 0.00001 + +Architecture: + model_type: rec + algorithm: CRNN + Transform: + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: small + small_stride: [1, 2, 2, 2] + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 48 + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/train_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - RecAug: + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/eval_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/multi_language/rec_japan_lite_train.yml b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_japan_lite_train.yml new file mode 100644 index 0000000000000000000000000000000000000000..bb47584edbc70f68d8d2d89dced3ec9b12f0e1cb --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_japan_lite_train.yml @@ -0,0 +1,102 @@ +Global: + use_gpu: True + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_japan_lite + save_epoch_step: 3 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: + # for data or label process + character_dict_path: ppocr/utils/dict/japan_dict.txt + character_type: japan + max_text_length: 25 + infer_mode: False + use_space_char: False + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + regularizer: + name: 'L2' + factor: 0.00001 + +Architecture: + model_type: rec + algorithm: CRNN + Transform: + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: small + small_stride: [1, 2, 2, 2] + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 48 + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/train_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - RecAug: + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/eval_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/multi_language/rec_korean_lite_train.yml b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_korean_lite_train.yml new file mode 100644 index 0000000000000000000000000000000000000000..77f15524f78cd7f1c3dcf4988960e718422f5d89 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_korean_lite_train.yml @@ -0,0 +1,102 @@ +Global: + use_gpu: True + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_korean_lite + save_epoch_step: 3 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: + # for data or label process + character_dict_path: ppocr/utils/dict/korean_dict.txt + character_type: korean + max_text_length: 25 + infer_mode: False + use_space_char: False + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + regularizer: + name: 'L2' + factor: 0.00001 + +Architecture: + model_type: rec + algorithm: CRNN + Transform: + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: small + small_stride: [1, 2, 2, 2] + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 48 + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/train_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - RecAug: + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/eval_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/multi_language/rec_latin_lite_train.yml b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_latin_lite_train.yml new file mode 100644 index 0000000000000000000000000000000000000000..e71112b4b4f0afd3ceab9f10078bc5d518ee9e59 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_latin_lite_train.yml @@ -0,0 +1,111 @@ +Global: + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_latin_lite + save_epoch_step: 3 + eval_batch_step: + - 0 + - 2000 + cal_metric_during_train: true + pretrained_model: null + checkpoints: null + save_inference_dir: null + use_visualdl: false + infer_img: null + character_dict_path: ppocr/utils/dict/latin_dict.txt + character_type: latin + max_text_length: 25 + infer_mode: false + use_space_char: true +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + regularizer: + name: L2 + factor: 1.0e-05 +Architecture: + model_type: rec + algorithm: CRNN + Transform: null + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: small + small_stride: + - 1 + - 2 + - 2 + - 2 + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 48 + Head: + name: CTCHead + fc_decay: 1.0e-05 +Loss: + name: CTCLoss +PostProcess: + name: CTCLabelDecode +Metric: + name: RecMetric + main_indicator: acc +Train: + dataset: + name: SimpleDataSet + data_dir: train_data/ + label_file_list: + - train_data/latin_train.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - RecAug: null + - CTCLabelEncode: null + - RecResizeImg: + image_shape: + - 3 + - 32 + - 320 + - KeepKeys: + keep_keys: + - image + - label + - length + loader: + shuffle: true + batch_size_per_card: 256 + drop_last: true + num_workers: 8 +Eval: + dataset: + name: SimpleDataSet + data_dir: train_data/ + label_file_list: + - train_data/latin_val.txt + transforms: + - DecodeImage: + img_mode: BGR + channel_first: false + - CTCLabelEncode: null + - RecResizeImg: + image_shape: + - 3 + - 32 + - 320 + - KeepKeys: + keep_keys: + - image + - label + - length + loader: + shuffle: false + drop_last: false + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/multi_language/rec_multi_language_lite_train.yml b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_multi_language_lite_train.yml new file mode 100644 index 0000000000000000000000000000000000000000..c42a3d1a3a5971365967effd4fdb2cc43725ef75 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/multi_language/rec_multi_language_lite_train.yml @@ -0,0 +1,103 @@ +Global: + use_gpu: True + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec_multi_language_lite + save_epoch_step: 3 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: + # for data or label process + character_dict_path: + # Set the language of training, if set, select the default dictionary file + character_type: + max_text_length: 25 + infer_mode: False + use_space_char: True + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Cosine + learning_rate: 0.001 + regularizer: + name: 'L2' + factor: 0.00001 + +Architecture: + model_type: rec + algorithm: CRNN + Transform: + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: small + small_stride: [1, 2, 2, 2] + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 48 + Head: + name: CTCHead + fc_decay: 0.00001 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: SimpleDataSet + data_dir: train_data/ + label_file_list: ["./train_data/train_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - RecAug: + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: train_data/ + label_file_list: ["./train_data/val_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 320] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_d28_can.yml b/PaddleOCR2Pytorch/configs/rec/rec_d28_can.yml new file mode 100644 index 0000000000000000000000000000000000000000..b6aa26b19afa544105da4af50aa004310f22fedd --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_d28_can.yml @@ -0,0 +1,122 @@ +Global: + use_gpu: True + epoch_num: 240 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/can/ + save_epoch_step: 1 + # evaluation is run every 1105 iterations (1 epoch)(batch_size = 8) + eval_batch_step: [0, 1105] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/datasets/crohme_demo/hme_00.jpg + # for data or label process + character_dict_path: ppocr/utils/dict/latex_symbol_dict.txt + max_text_length: 36 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_can.txt + +Optimizer: + name: Momentum + momentum: 0.9 + clip_norm_global: 100.0 + lr: + name: TwoStepCosine + learning_rate: 0.01 + warmup_epoch: 1 + weight_decay: 0.0001 + +Architecture: + model_type: rec + algorithm: CAN + in_channels: 1 + Transform: + Backbone: + name: DenseNet + growthRate: 24 + reduction: 0.5 + bottleneck: True + use_dropout: True + input_channel: 1 + Head: + name: CANHead + in_channel: 684 + out_channel: 111 + max_text_length: 36 + ratio: 16 + attdecoder: + is_train: False + input_size: 256 + hidden_size: 256 + encoder_out_channel: 684 + dropout: True + dropout_ratio: 0.5 + word_num: 111 + counting_decoder_out_channel: 111 + attention: + attention_dim: 512 + word_conv_kernel: 1 + +Loss: + name: CANLoss + +PostProcess: + name: CANLabelDecode + +Metric: + name: CANMetric + main_indicator: exp_rate + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/CROHME/training/images/ + label_file_list: ["./train_data/CROHME/training/labels.txt"] + transforms: + - DecodeImage: + channel_first: False + - NormalizeImage: + mean: [0,0,0] + std: [1,1,1] + order: 'hwc' + - GrayImageChannelFormat: + inverse: True + - CANLabelEncode: + lower: False + - KeepKeys: + keep_keys: ['image', 'label'] + loader: + shuffle: True + batch_size_per_card: 8 + drop_last: False + num_workers: 4 + collate_fn: DyMaskCollator + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/CROHME/evaluation/images/ + label_file_list: ["./train_data/CROHME/evaluation/labels.txt"] + transforms: + - DecodeImage: + channel_first: False + - NormalizeImage: + mean: [0,0,0] + std: [1,1,1] + order: 'hwc' + - GrayImageChannelFormat: + inverse: True + - CANLabelEncode: + lower: False + - KeepKeys: + keep_keys: ['image', 'label'] + loader: + shuffle: False + drop_last: False + batch_size_per_card: 1 + num_workers: 4 + collate_fn: DyMaskCollator diff --git a/PaddleOCR2Pytorch/configs/rec/rec_efficientb3_fpn_pren.yml b/PaddleOCR2Pytorch/configs/rec/rec_efficientb3_fpn_pren.yml new file mode 100644 index 0000000000000000000000000000000000000000..0fac6a7a8abcc5c1e4301b1040e6e98df74abed9 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_efficientb3_fpn_pren.yml @@ -0,0 +1,92 @@ +Global: + use_gpu: True + epoch_num: 8 + log_smooth_window: 20 + print_batch_step: 5 + save_model_dir: ./output/rec/pren_new + save_epoch_step: 3 + # evaluation is run every 2000 iterations after the 4000th iteration + eval_batch_step: [4000, 2000] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words/ch/word_1.jpg + # for data or label process + character_dict_path: + max_text_length: &max_text_length 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_pren.txt + +Optimizer: + name: Adadelta + lr: + name: Piecewise + decay_epochs: [2, 5, 7] + values: [0.5, 0.1, 0.01, 0.001] + +Architecture: + model_type: rec + algorithm: PREN + in_channels: 3 + Backbone: + name: EfficientNetb3_PREN + Neck: + name: PRENFPN + n_r: 5 + d_model: 384 + max_len: *max_text_length + dropout: 0.1 + Head: + name: PRENHead + +Loss: + name: PRENLoss + +PostProcess: + name: PRENLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: + img_mode: BGR + channel_first: False + - PRENLabelEncode: + - RecAug: + - PRENResizeImg: + image_shape: [64, 256] # h,w + - KeepKeys: + keep_keys: ['image', 'label'] + loader: + shuffle: True + batch_size_per_card: 128 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/validation/ + transforms: + - DecodeImage: + img_mode: BGR + channel_first: False + - PRENLabelEncode: + - PRENResizeImg: + image_shape: [64, 256] # h,w + - KeepKeys: + keep_keys: ['image', 'label'] + loader: + shuffle: False + drop_last: False + batch_size_per_card: 64 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_icdar15_train.yml b/PaddleOCR2Pytorch/configs/rec/rec_icdar15_train.yml new file mode 100644 index 0000000000000000000000000000000000000000..8a743b501618dc6dabf9b1d419431a148236c690 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_icdar15_train.yml @@ -0,0 +1,100 @@ +Global: + use_gpu: true + epoch_num: 72 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/ic15/ + save_epoch_step: 3 + # evaluation is run every 2000 iterations + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: ppocr/utils/ic15_dict.txt + character_type: ch + max_text_length: 25 + infer_mode: False + use_space_char: False + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + learning_rate: 0.0005 + regularizer: + name: 'L2' + factor: 0 + +Architecture: + model_type: rec + algorithm: CRNN + Transform: + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: large + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 96 + Head: + name: CTCHead + fc_decay: 0 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/train_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + use_shared_memory: False + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ + label_file_list: ["./train_data/train_list.txt"] + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 4 + use_shared_memory: False diff --git a/PaddleOCR2Pytorch/configs/rec/rec_mtb_nrtr.yml b/PaddleOCR2Pytorch/configs/rec/rec_mtb_nrtr.yml new file mode 100644 index 0000000000000000000000000000000000000000..765e28275a85f3d92abbece45d487a28a873f4e6 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_mtb_nrtr.yml @@ -0,0 +1,101 @@ +Global: + use_gpu: True + epoch_num: 21 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/nrtr/ + save_epoch_step: 1 + # evaluation is run every 2000 iterations + eval_batch_step: [0, 2000] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: pytorchocr/utils/EN_symbol_dict.txt + max_text_length: 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_nrtr.txt + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.99 + clip_norm: 5.0 + lr: + name: Cosine + learning_rate: 0.0005 + warmup_epoch: 2 + regularizer: + name: 'L2' + factor: 0. + +Architecture: + model_type: rec + algorithm: NRTR + in_channels: 1 + Transform: + Backbone: + name: MTB + cnn_num: 2 + Head: + name: Transformer + d_model: 512 + num_encoder_layers: 6 + beam_size: -1 # When Beam size is greater than 0, it means to use beam search when evaluation. + + +Loss: + name: NRTRLoss + smoothing: True + +PostProcess: + name: NRTRLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - NRTRLabelEncode: # Class handling label + - NRTRRecResizeImg: + image_shape: [100, 32] + resize_type: PIL # PIL or OpenCV + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 512 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/evaluation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - NRTRLabelEncode: # Class handling label + - NRTRRecResizeImg: + image_shape: [100, 32] + resize_type: PIL # PIL or OpenCV + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 1 + use_shared_memory: False diff --git a/PaddleOCR2Pytorch/configs/rec/rec_mv3_none_bilstm_ctc.yml b/PaddleOCR2Pytorch/configs/rec/rec_mv3_none_bilstm_ctc.yml new file mode 100644 index 0000000000000000000000000000000000000000..00c1db885e000d80ed3c3f42c2afbaa11c452ab5 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_mv3_none_bilstm_ctc.yml @@ -0,0 +1,96 @@ +Global: + use_gpu: True + epoch_num: 72 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/mv3_none_bilstm_ctc/ + save_epoch_step: 3 + # evaluation is run every 2000 iterations + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + infer_mode: False + use_space_char: False + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + learning_rate: 0.0005 + regularizer: + name: 'L2' + factor: 0 + +Architecture: + model_type: rec + algorithm: CRNN + Transform: + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: large + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 96 + Head: + name: CTCHead + fc_decay: 0 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/validation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 4 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_mv3_none_none_ctc.yml b/PaddleOCR2Pytorch/configs/rec/rec_mv3_none_none_ctc.yml new file mode 100644 index 0000000000000000000000000000000000000000..6711b1d23f843551d72e1dffc003637734727754 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_mv3_none_none_ctc.yml @@ -0,0 +1,95 @@ +Global: + use_gpu: True + epoch_num: 72 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/mv3_none_none_ctc/ + save_epoch_step: 3 + # evaluation is run every 2000 iterations + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + infer_mode: False + use_space_char: False + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + learning_rate: 0.0005 + regularizer: + name: 'L2' + factor: 0 + +Architecture: + model_type: rec + algorithm: Rosetta + Transform: + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: large + Neck: + name: SequenceEncoder + encoder_type: reshape + Head: + name: CTCHead + fc_decay: 0.0004 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/validation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_mv3_tps_bilstm_att.yml b/PaddleOCR2Pytorch/configs/rec/rec_mv3_tps_bilstm_att.yml new file mode 100644 index 0000000000000000000000000000000000000000..3cf1f7abfe9b29f010976d10b130a5199b2f2c0b --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_mv3_tps_bilstm_att.yml @@ -0,0 +1,102 @@ +Global: + use_gpu: True + epoch_num: 72 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/rec_mv3_tps_bilstm_att/ + save_epoch_step: 3 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words/ch/word_1.jpg + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + infer_mode: False + use_space_char: False + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + learning_rate: 0.0005 + regularizer: + name: 'L2' + factor: 0.00001 + +Architecture: + model_type: rec + algorithm: RARE + Transform: + name: TPS + num_fiducial: 20 + loc_lr: 0.1 + model_name: small + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: large + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 96 + Head: + name: AttentionHead + hidden_size: 96 + + +Loss: + name: AttentionLoss + +PostProcess: + name: AttnLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - AttnLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/validation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - AttnLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 1 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_mv3_tps_bilstm_ctc.yml b/PaddleOCR2Pytorch/configs/rec/rec_mv3_tps_bilstm_ctc.yml new file mode 100644 index 0000000000000000000000000000000000000000..4e86709942bcde410dc22df439fdd40e9a94fdef --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_mv3_tps_bilstm_ctc.yml @@ -0,0 +1,100 @@ +Global: + use_gpu: True + epoch_num: 72 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/mv3_tps_bilstm_ctc/ + save_epoch_step: 3 + # evaluation is run every 2000 iterations + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + infer_mode: False + use_space_char: False + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + learning_rate: 0.0005 + regularizer: + name: 'L2' + factor: 0 + +Architecture: + model_type: rec + algorithm: STARNet + Transform: + name: TPS + num_fiducial: 20 + loc_lr: 0.1 + model_name: small + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: large + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 96 + Head: + name: CTCHead + fc_decay: 0.0004 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/validation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 4 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_r31_sar.yml b/PaddleOCR2Pytorch/configs/rec/rec_r31_sar.yml new file mode 100644 index 0000000000000000000000000000000000000000..65e7877b28da80e0730f551b07d60b8a8c0ac48e --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_r31_sar.yml @@ -0,0 +1,98 @@ +Global: + use_gpu: true + epoch_num: 5 + log_smooth_window: 20 + print_batch_step: 20 + save_model_dir: ./sar_rec + save_epoch_step: 1 + # evaluation is run every 2000 iterations + eval_batch_step: [0, 2000] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: + # for data or label process + character_dict_path: ppocr/utils/dict90.txt + max_text_length: 30 + infer_mode: False + use_space_char: False + rm_symbol: True + save_res_path: ./output/rec/predicts_sar.txt + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + name: Piecewise + decay_epochs: [3, 4] + values: [0.001, 0.0001, 0.00001] + regularizer: + name: 'L2' + factor: 0 + +Architecture: + model_type: rec + algorithm: SAR + Transform: + Backbone: + name: ResNet31 + Head: + name: SARHead + +Loss: + name: SARLoss + +PostProcess: + name: SARLabelDecode + +Metric: + name: RecMetric + + +Train: + dataset: + name: SimpleDataSet + label_file_list: ['./train_data/train_list.txt'] + data_dir: ./train_data/ + ratio_list: 1.0 + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - SARLabelEncode: # Class handling label + - SARRecResizeImg: + image_shape: [3, 48, 48, 160] # h:48 w:[48,160] + width_downsample_ratio: 0.25 + - KeepKeys: + keep_keys: ['image', 'label', 'valid_ratio'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 64 + drop_last: True + num_workers: 8 + use_shared_memory: False + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/evaluation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - SARLabelEncode: # Class handling label + - SARRecResizeImg: + image_shape: [3, 48, 48, 160] + width_downsample_ratio: 0.25 + - KeepKeys: + keep_keys: ['image', 'label', 'valid_ratio'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 64 + num_workers: 4 + use_shared_memory: False + diff --git a/PaddleOCR2Pytorch/configs/rec/rec_r34_vd_none_bilstm_ctc.yml b/PaddleOCR2Pytorch/configs/rec/rec_r34_vd_none_bilstm_ctc.yml new file mode 100644 index 0000000000000000000000000000000000000000..e4d301a6a173ea772898c0528c4b3082670870ff --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_r34_vd_none_bilstm_ctc.yml @@ -0,0 +1,95 @@ +Global: + use_gpu: true + epoch_num: 72 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/r34_vd_none_bilstm_ctc/ + save_epoch_step: 3 + # evaluation is run every 2000 iterations + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + infer_mode: False + use_space_char: False + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + learning_rate: 0.0005 + regularizer: + name: 'L2' + factor: 0 + +Architecture: + model_type: rec + algorithm: CRNN + Transform: + Backbone: + name: ResNet + layers: 34 + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 256 + Head: + name: CTCHead + fc_decay: 0 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/validation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 4 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_r34_vd_none_none_ctc.yml b/PaddleOCR2Pytorch/configs/rec/rec_r34_vd_none_none_ctc.yml new file mode 100644 index 0000000000000000000000000000000000000000..4a17a004228185db7e52dd71aadcff36d407d2cf --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_r34_vd_none_none_ctc.yml @@ -0,0 +1,93 @@ +Global: + use_gpu: true + epoch_num: 72 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/r34_vd_none_none_ctc/ + save_epoch_step: 3 + # evaluation is run every 2000 iterations + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + infer_mode: False + use_space_char: False + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + learning_rate: 0.0005 + regularizer: + name: 'L2' + factor: 0 + +Architecture: + model_type: rec + algorithm: Rosetta + Backbone: + name: ResNet + layers: 34 + Neck: + name: SequenceEncoder + encoder_type: reshape + Head: + name: CTCHead + fc_decay: 0.0004 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/validation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 4 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_r34_vd_tps_bilstm_att.yml b/PaddleOCR2Pytorch/configs/rec/rec_r34_vd_tps_bilstm_att.yml new file mode 100644 index 0000000000000000000000000000000000000000..659a1723da04db51901775949cfcad250cea2dce --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_r34_vd_tps_bilstm_att.yml @@ -0,0 +1,101 @@ +Global: + use_gpu: True + epoch_num: 400 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/b3_rare_r34_none_gru/ + save_epoch_step: 3 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words/ch/word_1.jpg + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + infer_mode: False + use_space_char: False + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + learning_rate: 0.0005 + regularizer: + name: 'L2' + factor: 0.00000 + +Architecture: + model_type: rec + algorithm: RARE + Transform: + name: TPS + num_fiducial: 20 + loc_lr: 0.1 + model_name: large + Backbone: + name: ResNet + layers: 34 + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 256 #96 + Head: + name: AttentionHead # AttentionHead + hidden_size: 256 # + l2_decay: 0.00001 + +Loss: + name: AttentionLoss + +PostProcess: + name: AttnLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - AttnLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/validation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - AttnLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 8 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_r34_vd_tps_bilstm_ctc.yml b/PaddleOCR2Pytorch/configs/rec/rec_r34_vd_tps_bilstm_ctc.yml new file mode 100644 index 0000000000000000000000000000000000000000..7e9d13aa55d755e645b15a3801887095806af517 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_r34_vd_tps_bilstm_ctc.yml @@ -0,0 +1,99 @@ +Global: + use_gpu: true + epoch_num: 72 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/r34_vd_tps_bilstm_ctc/ + save_epoch_step: 3 + # evaluation is run every 2000 iterations + eval_batch_step: [0, 2000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + infer_mode: False + use_space_char: False + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + lr: + learning_rate: 0.0005 + regularizer: + name: 'L2' + factor: 0 + +Architecture: + model_type: rec + algorithm: STARNet + Transform: + name: TPS + num_fiducial: 20 + loc_lr: 0.1 + model_name: large + Backbone: + name: ResNet + layers: 34 + Neck: + name: SequenceEncoder + encoder_type: reshape + hidden_size: 256 + Head: + name: CTCHead + fc_decay: 0 + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/validation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 32, 100] + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 4 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_r50_fpn_srn.yml b/PaddleOCR2Pytorch/configs/rec/rec_r50_fpn_srn.yml new file mode 100644 index 0000000000000000000000000000000000000000..6b38616c7eb17a3d4578222fc36efd971e9428c6 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_r50_fpn_srn.yml @@ -0,0 +1,107 @@ +Global: + use_gpu: True + epoch_num: 72 + log_smooth_window: 20 + print_batch_step: 5 + save_model_dir: ./output/rec/srn_new + save_epoch_step: 3 + # evaluation is run every 5000 iterations after the 4000th iteration + eval_batch_step: [0, 5000] + # if pretrained_model is saved in static mode, load_static_weights must set to True + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words/ch/word_1.jpg + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + num_heads: 8 + infer_mode: False + use_space_char: False + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + clip_norm: 10.0 + lr: + learning_rate: 0.0001 + +Architecture: + model_type: rec + algorithm: SRN + in_channels: 1 + Transform: + Backbone: + name: ResNetFPN + Head: + name: SRNHead + max_text_length: 25 + num_heads: 8 + num_encoder_TUs: 2 + num_decoder_TUs: 4 + hidden_dims: 512 + +Loss: + name: SRNLoss + +PostProcess: + name: SRNLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - SRNLabelEncode: # Class handling label + - SRNRecResizeImg: + image_shape: [1, 64, 256] + - KeepKeys: + keep_keys: ['image', + 'label', + 'length', + 'encoder_word_pos', + 'gsrm_word_pos', + 'gsrm_slf_attn_bias1', + 'gsrm_slf_attn_bias2'] # dataloader will return list in this order + loader: + shuffle: False + batch_size_per_card: 64 + drop_last: False + num_workers: 4 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/validation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - SRNLabelEncode: # Class handling label + - SRNRecResizeImg: + image_shape: [1, 64, 256] + - KeepKeys: + keep_keys: ['image', + 'label', + 'length', + 'encoder_word_pos', + 'gsrm_word_pos', + 'gsrm_slf_attn_bias1', + 'gsrm_slf_attn_bias2'] + loader: + shuffle: False + drop_last: False + batch_size_per_card: 32 + num_workers: 4 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_base_8local_10global_stn_en.yml b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_base_8local_10global_stn_en.yml new file mode 100644 index 0000000000000000000000000000000000000000..9127838e5e6e595c0abacc4c8c0129f0f6921946 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_base_8local_10global_stn_en.yml @@ -0,0 +1,117 @@ +Global: + use_gpu: True + epoch_num: 20 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/rec_svtr_base_stn_en/ + save_epoch_step: 1 + # evaluation is run every 2000 iterations after the 0th iteration + eval_batch_step: [0, 2000] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_svtr_base.txt + + +Optimizer: + name: AdamW + beta1: 0.9 + beta2: 0.99 + epsilon: 0.00000008 + weight_decay: 0.05 + no_weight_decay_name: norm pos_embed + one_dim_param_no_weight_decay: true + lr: + name: Cosine + learning_rate: 0.00025 + warmup_epoch: 2 + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + name: STN_ON + tps_inputsize: [32, 64] + tps_outputsize: [48, 160] + num_control_points: 20 + tps_margins: [0.05,0.05] + stn_activation: none + Backbone: + name: SVTRNet + img_size: [48, 160] + out_char_num: 40 # output char patch + out_channels: 256 # char patch dim + patch_merging: 'Conv' + embed_dim: [128, 256, 384] + depth: [3, 6, 9] + num_heads: [4, 8, 12] + mixer: ['Local','Local','Local','Local','Local','Local','Local','Local','Global','Global','Global','Global','Global','Global','Global','Global','Global','Global'] + local_mixer: [[7, 11], [7, 11], [7, 11]] + last_stage: True + prenorm: False + Neck: + name: SequenceEncoder + encoder_type: reshape + Head: + name: CTCHead + +Loss: + name: CTCLoss + +PostProcess: + name: SVTRLabelDecode # SVTRLabelDecode is used for eval after train, please change to CTCLabelDecode when training + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + character_dict_path: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 256 + drop_last: True + num_workers: 4 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/evaluation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - SVTRRecResizeImg: # SVTRRecResizeImg is used for eval after train, please change to RecResizeImg when training + character_dict_path: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 128 + num_workers: 2 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_large_10local_11global_stn_ch.yml b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_large_10local_11global_stn_ch.yml new file mode 100644 index 0000000000000000000000000000000000000000..ed393ad018491c5442c0978fb187171b19dd6877 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_large_10local_11global_stn_ch.yml @@ -0,0 +1,113 @@ +Global: + use_gpu: True + epoch_num: 100 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/rec_svtr_large_ch/ + save_epoch_step: 10 + # evaluation is run every 2000 iterations after the 0th iteration + eval_batch_step: [0, 2000] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words/ch/word_1.jpg + # for data or label process + character_dict_path: ppocr/utils/ppocr_keys_v1.txt + max_text_length: 40 + infer_mode: False + use_space_char: True + save_res_path: ./output/rec/predicts_svtr_large_ch.txt + + +Optimizer: + name: AdamW + beta1: 0.9 + beta2: 0.99 + epsilon: 0.00000008 + weight_decay: 0.05 + no_weight_decay_name: norm pos_embed + one_dim_param_no_weight_decay: true + lr: + name: Cosine + learning_rate: 0.0003 + warmup_epoch: 5 + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + name: STN_ON + tps_inputsize: [32, 64] + tps_outputsize: [32, 320] + num_control_points: 20 + tps_margins: [0.05,0.05] + stn_activation: none + Backbone: + name: SVTRNet + img_size: [32, 320] + out_char_num: 40 + out_channels: 384 + patch_merging: 'Conv' + embed_dim: [192, 256, 512] + depth: [3, 9, 9] + num_heads: [6, 8, 16] + mixer: ['Local','Local','Local','Local','Local','Local','Local','Local','Local','Local','Global','Global','Global','Global','Global','Global','Global','Global','Global','Global','Global'] + local_mixer: [[7, 11], [7, 11], [7, 11]] + prenorm: False + Neck: + name: SequenceEncoder + encoder_type: reshape + Head: + name: CTCHead + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/scene_ch/ch_scene + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 128 + drop_last: True + num_workers: 4 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/scene_ch/scene_test + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 2 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_large_10local_11global_stn_en.yml b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_large_10local_11global_stn_en.yml new file mode 100644 index 0000000000000000000000000000000000000000..aac75169b35040ffc18c2e89a1566a24688df3a8 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_large_10local_11global_stn_en.yml @@ -0,0 +1,117 @@ +Global: + use_gpu: True + epoch_num: 20 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/rec_svtr_large_en/ + save_epoch_step: 1 + # evaluation is run every 2000 iterations after the 0th iteration + eval_batch_step: [0, 2000] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_svtr_large.txt + + +Optimizer: + name: AdamW + beta1: 0.9 + beta2: 0.99 + epsilon: 0.00000008 + weight_decay: 0.05 + no_weight_decay_name: norm pos_embed + one_dim_param_no_weight_decay: true + lr: + name: Cosine + learning_rate: 0.000125 + warmup_epoch: 2 + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + name: STN_ON + tps_inputsize: [32, 64] + tps_outputsize: [48, 160] + num_control_points: 20 + tps_margins: [0.05,0.05] + stn_activation: none + Backbone: + name: SVTRNet + img_size: [48, 160] + out_char_num: 40 + out_channels: 384 + patch_merging: 'Conv' + embed_dim: [192, 256, 512] + depth: [3, 9, 9] + num_heads: [6, 8, 16] + mixer: ['Local','Local','Local','Local','Local','Local','Local','Local','Local','Local','Global','Global','Global','Global','Global','Global','Global','Global','Global','Global','Global'] + local_mixer: [[7, 11], [7, 11], [7, 11]] + prenorm: false + Neck: + name: SequenceEncoder + encoder_type: reshape + Head: + name: CTCHead + +Loss: + name: CTCLoss + +PostProcess: + name: SVTRLabelDecode # SVTRLabelDecode is used for eval after train, please change to CTCLabelDecode when training + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - RecAug: + - CTCLabelEncode: # Class handling label + - RecResizeImg: + character_dict_path: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 128 + drop_last: True + num_workers: 4 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/evaluation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - SVTRRecResizeImg: # SVTRRecResizeImg is used for eval after train, please change to RecResizeImg when training + character_dict_path: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 128 + num_workers: 2 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_small_8local_7global_stn_ch.yml b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_small_8local_7global_stn_ch.yml new file mode 100644 index 0000000000000000000000000000000000000000..985f0d387a689902a6cf9539b7bf561cccee2255 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_small_8local_7global_stn_ch.yml @@ -0,0 +1,114 @@ +Global: + use_gpu: True + epoch_num: 100 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/rec_svtr_small_ch/ + save_epoch_step: 10 + # evaluation is run every 2000 iterations after the 0th iteration + eval_batch_step: [0, 2000] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words/ch/word_1.jpg + # for data or label process + character_dict_path: ppocr/utils/ppocr_keys_v1.txt + max_text_length: 40 + infer_mode: False + use_space_char: True + save_res_path: ./output/rec/predicts_svtr_small_ch.txt + + +Optimizer: + name: AdamW + beta1: 0.9 + beta2: 0.99 + epsilon: 0.00000008 + weight_decay: 0.05 + no_weight_decay_name: norm pos_embed + one_dim_param_no_weight_decay: true + lr: + name: Cosine + learning_rate: 0.0003 + warmup_epoch: 5 + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + name: STN_ON + tps_inputsize: [32, 64] + tps_outputsize: [32, 320] + num_control_points: 20 + tps_margins: [0.05,0.05] + stn_activation: none + Backbone: + name: SVTRNet + img_size: [32, 320] + out_char_num: 40 + out_channels: 192 + patch_merging: 'Conv' + embed_dim: [96, 192, 256] + depth: [3, 6, 6] + num_heads: [3, 6, 8] + mixer: ['Local','Local','Local','Local','Local','Local','Local','Local','Global','Global','Global','Global','Global','Global','Global'] + local_mixer: [[7, 11], [7, 11], [7, 11]] + last_stage: True + prenorm: False + Neck: + name: SequenceEncoder + encoder_type: reshape + Head: + name: CTCHead + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/scene_ch/ch_scene + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 128 + drop_last: True + num_workers: 4 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/scene_ch/scene_test + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 2 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_small_8local_7global_stn_en.yml b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_small_8local_7global_stn_en.yml new file mode 100644 index 0000000000000000000000000000000000000000..2405d3efc7005cda1ea7fd9c01dd4c3d6eb4ca80 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_small_8local_7global_stn_en.yml @@ -0,0 +1,117 @@ +Global: + use_gpu: True + epoch_num: 20 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/rec_svtr_small_stn_en/ + save_epoch_step: 1 + # evaluation is run every 2000 iterations after the 0th iteration + eval_batch_step: [0, 2000] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_svtr_small.txt + + +Optimizer: + name: AdamW + beta1: 0.9 + beta2: 0.99 + epsilon: 0.00000008 + weight_decay: 0.05 + no_weight_decay_name: norm pos_embed + one_dim_param_no_weight_decay: true + lr: + name: Cosine + learning_rate: 0.0005 + warmup_epoch: 2 + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + name: STN_ON + tps_inputsize: [32, 64] + tps_outputsize: [32, 100] + num_control_points: 20 + tps_margins: [0.05,0.05] + stn_activation: none + Backbone: + name: SVTRNet + img_size: [32, 100] + out_char_num: 25 + out_channels: 192 + patch_merging: 'Conv' + embed_dim: [96, 192, 256] + depth: [3, 6, 6] + num_heads: [3, 6, 8] + mixer: ['Local','Local','Local','Local','Local','Local','Local','Local','Global','Global','Global','Global','Global','Global','Global'] + local_mixer: [[7, 11], [7, 11], [7, 11]] + last_stage: True + prenorm: False + Neck: + name: SequenceEncoder + encoder_type: reshape + Head: + name: CTCHead + +Loss: + name: CTCLoss + +PostProcess: + name: SVTRLabelDecode # SVTRLabelDecode is used for eval after train, please change to CTCLabelDecode when training + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + character_dict_path: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 512 + drop_last: True + num_workers: 4 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/evaluation + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - SVTRRecResizeImg: # SVTRRecResizeImg is used for eval after train, please change to RecResizeImg when training + character_dict_path: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 2 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_tiny_6local_6global_stn_ch.yml b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_tiny_6local_6global_stn_ch.yml new file mode 100644 index 0000000000000000000000000000000000000000..42fe93bf0f4a5918b925a25109c1e00a1222b901 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_tiny_6local_6global_stn_ch.yml @@ -0,0 +1,114 @@ +Global: + use_gpu: True + epoch_num: 100 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/rec_svtr_tiny_ch/ + save_epoch_step: 10 + # evaluation is run every 2000 iterations after the 0th iteration + eval_batch_step: [0, 2000] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words/ch/word_1.jpg + # for data or label process + character_dict_path: ppocr/utils/ppocr_keys_v1.txt + max_text_length: 40 + infer_mode: False + use_space_char: True + save_res_path: ./output/rec/predicts_svtr_tiny_ch.txt + + +Optimizer: + name: AdamW + beta1: 0.9 + beta2: 0.99 + epsilon: 0.00000008 + weight_decay: 0.05 + no_weight_decay_name: norm pos_embed + one_dim_param_no_weight_decay: true + lr: + name: Cosine + learning_rate: 0.0003 + warmup_epoch: 5 + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + name: STN_ON + tps_inputsize: [32, 64] + tps_outputsize: [32, 320] + num_control_points: 20 + tps_margins: [0.05,0.05] + stn_activation: none + Backbone: + name: SVTRNet + img_size: [32, 320] + out_char_num: 40 + out_channels: 192 + patch_merging: 'Conv' + embed_dim: [64, 128, 256] + depth: [3, 6, 3] + num_heads: [2, 4, 8] + mixer: ['Local','Local','Local','Local','Local','Local','Global','Global','Global','Global','Global','Global'] + local_mixer: [[7, 11], [7, 11], [7, 11]] + last_stage: True + prenorm: false + Neck: + name: SequenceEncoder + encoder_type: reshape + Head: + name: CTCHead + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/scene_ch/ch_scene + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 128 + drop_last: True + num_workers: 4 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/scene_ch/scene_test + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 2 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_tiny_6local_6global_stn_en.yml b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_tiny_6local_6global_stn_en.yml new file mode 100644 index 0000000000000000000000000000000000000000..2a413909267dd748b528aa325f7f9159d6997431 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_svtr/rec_svtr_tiny_6local_6global_stn_en.yml @@ -0,0 +1,117 @@ +Global: + use_gpu: True + epoch_num: 20 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/rec_svtr_tiny_en/ + save_epoch_step: 1 + # evaluation is run every 2000 iterations after the 0th iteration + eval_batch_step: [0, 2000] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_svtr_tiny.txt + + +Optimizer: + name: AdamW + beta1: 0.9 + beta2: 0.99 + epsilon: 0.00000008 + weight_decay: 0.05 + no_weight_decay_name: norm pos_embed + one_dim_param_no_weight_decay: true + lr: + name: Cosine + learning_rate: 0.0005 + warmup_epoch: 2 + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + name: STN_ON + tps_inputsize: [32, 64] + tps_outputsize: [32, 100] + num_control_points: 20 + tps_margins: [0.05,0.05] + stn_activation: none + Backbone: + name: SVTRNet + img_size: [32, 100] + out_char_num: 25 + out_channels: 192 + patch_merging: 'Conv' + embed_dim: [64, 128, 256] + depth: [3, 6, 3] + num_heads: [2, 4, 8] + mixer: ['Local','Local','Local','Local','Local','Local','Global','Global','Global','Global','Global','Global'] + local_mixer: [[7, 11], [7, 11], [7, 11]] + last_stage: True + prenorm: false + Neck: + name: SequenceEncoder + encoder_type: reshape + Head: + name: CTCHead + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + character_dict_path: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 512 + drop_last: True + num_workers: 4 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/evaluation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + character_dict_path: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 2 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_svtrnet.yml b/PaddleOCR2Pytorch/configs/rec/rec_svtrnet.yml new file mode 100644 index 0000000000000000000000000000000000000000..233d5e276577cad0144456ef7df1e20de99891f9 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_svtrnet.yml @@ -0,0 +1,117 @@ +Global: + use_gpu: True + epoch_num: 20 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/svtr/ + save_epoch_step: 1 + # evaluation is run every 2000 iterations after the 0th iteration + eval_batch_step: [0, 2000] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: + character_type: en + max_text_length: 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_svtr_tiny.txt + + +Optimizer: + name: AdamW + beta1: 0.9 + beta2: 0.99 + epsilon: 0.00000008 + weight_decay: 0.05 + no_weight_decay_name: norm pos_embed + one_dim_param_no_weight_decay: true + lr: + name: Cosine + learning_rate: 0.0005 + warmup_epoch: 2 + +Architecture: + model_type: rec + algorithm: SVTR + Transform: + name: STN_ON + tps_inputsize: [32, 64] + tps_outputsize: [32, 100] + num_control_points: 20 + tps_margins: [0.05,0.05] + stn_activation: none + Backbone: + name: SVTRNet + img_size: [32, 100] + out_char_num: 25 + out_channels: 192 + patch_merging: 'Conv' + embed_dim: [64, 128, 256] + depth: [3, 6, 3] + num_heads: [2, 4, 8] + mixer: ['Local','Local','Local','Local','Local','Local','Global','Global','Global','Global','Global','Global'] + local_mixer: [[7, 11], [7, 11], [7, 11]] + last_stage: True + prenorm: false + Neck: + name: SequenceEncoder + encoder_type: reshape + Head: + name: CTCHead + +Loss: + name: CTCLoss + +PostProcess: + name: CTCLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + character_dict_path: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 512 + drop_last: True + num_workers: 4 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/validation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - CTCLabelEncode: # Class handling label + - RecResizeImg: + character_dict_path: + image_shape: [3, 64, 256] + padding: False + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 2 diff --git a/PaddleOCR2Pytorch/configs/rec/rec_vitstr_none_ce.yml b/PaddleOCR2Pytorch/configs/rec/rec_vitstr_none_ce.yml new file mode 100644 index 0000000000000000000000000000000000000000..ebe304fa4bad3a33b4a226e761ccabbfda41e202 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/rec/rec_vitstr_none_ce.yml @@ -0,0 +1,102 @@ +Global: + use_gpu: True + epoch_num: 20 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/rec/vitstr_none_ce/ + save_epoch_step: 1 + # evaluation is run every 2000 iterations after the 0th iteration# + eval_batch_step: [0, 2000] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words_en/word_10.png + # for data or label process + character_dict_path: ppocr/utils/EN_symbol_dict.txt + max_text_length: 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_vitstr.txt + + +Optimizer: + name: Adadelta + epsilon: 1.e-8 + rho: 0.95 + clip_norm: 5.0 + lr: + learning_rate: 1.0 + +Architecture: + model_type: rec + algorithm: ViTSTR + in_channels: 1 + Transform: + Backbone: + name: ViTSTR + scale: tiny + Neck: + name: SequenceEncoder + encoder_type: reshape + Head: + name: CTCHead + +Loss: + name: CELoss + with_all: True + ignore_index: &ignore_index 0 # Must be zero or greater than the number of character classes + +PostProcess: + name: ViTSTRLabelDecode + +Metric: + name: RecMetric + main_indicator: acc + +Train: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/training/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - ViTSTRLabelEncode: # Class handling label + ignore_index: *ignore_index + - GrayRecResizeImg: + image_shape: [224, 224] # W H + resize_type: PIL # PIL or OpenCV + inter_type: 'Image.BICUBIC' + scale: false + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: True + batch_size_per_card: 48 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: LMDBDataSet + data_dir: ./train_data/data_lmdb_release/evaluation/ + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - ViTSTRLabelEncode: # Class handling label + ignore_index: *ignore_index + - GrayRecResizeImg: + image_shape: [224, 224] # W H + resize_type: PIL # PIL or OpenCV + inter_type: 'Image.BICUBIC' + scale: false + - KeepKeys: + keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 256 + num_workers: 2 diff --git a/PaddleOCR2Pytorch/configs/sr/sr_telescope.yml b/PaddleOCR2Pytorch/configs/sr/sr_telescope.yml new file mode 100644 index 0000000000000000000000000000000000000000..1152596083bd336cebe67df4684adecc3b373a88 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/sr/sr_telescope.yml @@ -0,0 +1,84 @@ +Global: + use_gpu: true + epoch_num: 100 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/sr/sr_telescope/ + save_epoch_step: 3 + # evaluation is run every 2000 iterations + eval_batch_step: [0, 1000] + cal_metric_during_train: False + pretrained_model: + checkpoints: + save_inference_dir: ./output/sr/sr_telescope/infer + use_visualdl: False + infer_img: doc/imgs_words_en/word_52.png + # for data or label process + character_dict_path: + max_text_length: 100 + infer_mode: False + use_space_char: False + save_res_path: ./output/sr/predicts_telescope.txt + +Optimizer: + name: Adam + beta1: 0.5 + beta2: 0.999 + clip_norm: 0.25 + lr: + learning_rate: 0.0001 + +Architecture: + model_type: sr + algorithm: Telescope + Transform: + name: TBSRN + STN: True + infer_mode: True + +Loss: + name: TelescopeLoss + confuse_dict_path: ./ppocr/utils/dict/confuse.pkl + + +PostProcess: + name: None + +Metric: + name: SRMetric + main_indicator: all + +Train: + dataset: + name: LMDBDataSetSR + data_dir: ./train_data/TextZoom/train + transforms: + - SRResize: + imgH: 32 + imgW: 128 + down_sample_scale: 2 + - KeepKeys: + keep_keys: ['img_lr', 'img_hr', 'label'] # dataloader will return list in this order + loader: + shuffle: False + batch_size_per_card: 16 + drop_last: True + num_workers: 4 + +Eval: + dataset: + name: LMDBDataSetSR + data_dir: ./train_data/TextZoom/test + transforms: + - SRResize: + imgH: 32 + imgW: 128 + down_sample_scale: 2 + - KeepKeys: + keep_keys: ['img_lr', 'img_hr', 'label'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 16 + num_workers: 4 + diff --git a/PaddleOCR2Pytorch/configs/sr/sr_tsrn_transformer_strock.yml b/PaddleOCR2Pytorch/configs/sr/sr_tsrn_transformer_strock.yml new file mode 100644 index 0000000000000000000000000000000000000000..2e5094863ae0244302fd7ed192e1f45bf8865b1f --- /dev/null +++ b/PaddleOCR2Pytorch/configs/sr/sr_tsrn_transformer_strock.yml @@ -0,0 +1,85 @@ +Global: + use_gpu: true + epoch_num: 500 + log_smooth_window: 20 + print_batch_step: 10 + save_model_dir: ./output/sr/sr_tsrn_transformer_strock/ + save_epoch_step: 3 + # evaluation is run every 2000 iterations + eval_batch_step: [0, 1000] + cal_metric_during_train: False + pretrained_model: + checkpoints: + save_inference_dir: sr_output + use_visualdl: False + infer_img: doc/imgs_words_en/word_52.png + # for data or label process + character_dict_path: ./train_data/srdata/english_decomposition.txt + max_text_length: 100 + infer_mode: False + use_space_char: False + save_res_path: ./output/sr/predicts_gestalt.txt + +Optimizer: + name: Adam + beta1: 0.5 + beta2: 0.999 + clip_norm: 0.25 + lr: + learning_rate: 0.0001 + +Architecture: + model_type: sr + algorithm: Gestalt + Transform: + name: TSRN + STN: True + infer_mode: True + +Loss: + name: StrokeFocusLoss + character_dict_path: ./train_data/srdata/english_decomposition.txt + +PostProcess: + name: None + +Metric: + name: SRMetric + main_indicator: all + +Train: + dataset: + name: LMDBDataSetSR + data_dir: ./train_data/srdata/train + transforms: + - SRResize: + imgH: 32 + imgW: 128 + down_sample_scale: 2 + - SRLabelEncode: # Class handling label + - KeepKeys: + keep_keys: ['img_lr', 'img_hr', 'length', 'input_tensor', 'label'] # dataloader will return list in this order + loader: + shuffle: False + batch_size_per_card: 16 + drop_last: True + num_workers: 4 + +Eval: + dataset: + name: LMDBDataSetSR + data_dir: ./train_data/srdata/test + transforms: + - SRResize: + imgH: 32 + imgW: 128 + down_sample_scale: 2 + - SRLabelEncode: # Class handling label + - KeepKeys: + keep_keys: ['img_lr', 'img_hr','length', 'input_tensor', 'label'] # dataloader will return list in this order + loader: + shuffle: False + drop_last: False + batch_size_per_card: 16 + num_workers: 4 + diff --git a/PaddleOCR2Pytorch/configs/table/table_mv3.yml b/PaddleOCR2Pytorch/configs/table/table_mv3.yml new file mode 100644 index 0000000000000000000000000000000000000000..a74e18d318699685400cc48430c04db3fef70b60 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/table/table_mv3.yml @@ -0,0 +1,116 @@ +Global: + use_gpu: true + epoch_num: 50 + log_smooth_window: 20 + print_batch_step: 5 + save_model_dir: ./output/table_mv3/ + save_epoch_step: 5 + # evaluation is run every 400 iterations after the 0th iteration + eval_batch_step: [0, 400] + cal_metric_during_train: True + pretrained_model: + checkpoints: + save_inference_dir: + use_visualdl: False + infer_img: doc/imgs_words/ch/word_1.jpg + # for data or label process + character_dict_path: ppocr/utils/dict/table_structure_dict.txt + character_type: en + max_text_length: 100 + max_elem_length: 500 + max_cell_num: 500 + infer_mode: False + process_total_num: 0 + process_cut_num: 0 + + +Optimizer: + name: Adam + beta1: 0.9 + beta2: 0.999 + clip_norm: 5.0 + lr: + learning_rate: 0.001 + regularizer: + name: 'L2' + factor: 0.00000 + +Architecture: + model_type: table + algorithm: TableAttn + Backbone: + name: MobileNetV3 + scale: 1.0 + model_name: small + disable_se: True + Head: + name: TableAttentionHead + hidden_size: 256 + l2_decay: 0.00001 + loc_type: 2 + +Loss: + name: TableAttentionLoss + structure_weight: 100.0 + loc_weight: 10000.0 + +PostProcess: + name: TableLabelDecode + +Metric: + name: TableMetric + main_indicator: acc + +Train: + dataset: + name: PubTabDataSet + data_dir: train_data/table/pubtabnet/train/ + label_file_path: train_data/table/pubtabnet/PubTabNet_2.0.0_train.jsonl + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - ResizeTableImage: + max_len: 488 + - TableLabelEncode: + - NormalizeImage: + scale: 1./255. + mean: [0.485, 0.456, 0.406] + std: [0.229, 0.224, 0.225] + order: 'hwc' + - PaddingTableImage: + - ToCHWImage: + - KeepKeys: + keep_keys: ['image', 'structure', 'bbox_list', 'sp_tokens', 'bbox_list_mask'] + loader: + shuffle: True + batch_size_per_card: 32 + drop_last: True + num_workers: 1 + +Eval: + dataset: + name: PubTabDataSet + data_dir: train_data/table/pubtabnet/val/ + label_file_path: train_data/table/pubtabnet/PubTabNet_2.0.0_val.jsonl + transforms: + - DecodeImage: # load image + img_mode: BGR + channel_first: False + - ResizeTableImage: + max_len: 488 + - TableLabelEncode: + - NormalizeImage: + scale: 1./255. + mean: [0.485, 0.456, 0.406] + std: [0.229, 0.224, 0.225] + order: 'hwc' + - PaddingTableImage: + - ToCHWImage: + - KeepKeys: + keep_keys: ['image', 'structure', 'bbox_list', 'sp_tokens', 'bbox_list_mask'] + loader: + shuffle: False + drop_last: False + batch_size_per_card: 16 + num_workers: 1 diff --git a/PaddleOCR2Pytorch/configs/table/table_mv3_det.yml b/PaddleOCR2Pytorch/configs/table/table_mv3_det.yml new file mode 100644 index 0000000000000000000000000000000000000000..f9bb4ba72f54538b6b873a791f7292bece86d6ab --- /dev/null +++ b/PaddleOCR2Pytorch/configs/table/table_mv3_det.yml @@ -0,0 +1,15 @@ +Architecture: + model_type: det + algorithm: DB + Transform: + Backbone: + name: MobileNetV3 + scale: 0.5 + model_name: large + disable_se: true + Neck: + name: DBFPN + out_channels: 96 + Head: + name: DBHead + k: 50 \ No newline at end of file diff --git a/PaddleOCR2Pytorch/configs/table/table_mv3_rec.yml b/PaddleOCR2Pytorch/configs/table/table_mv3_rec.yml new file mode 100644 index 0000000000000000000000000000000000000000..8510ca14119f4a4155ae4b6d2c28ceeb04061613 --- /dev/null +++ b/PaddleOCR2Pytorch/configs/table/table_mv3_rec.yml @@ -0,0 +1,14 @@ +Architecture: + model_type: rec + algorithm: CRNN + Transform: + Backbone: + name: MobileNetV3 + model_name: large + Neck: + name: SequenceEncoder + encoder_type: rnn + hidden_size: 96 + Head: + name: CTCHead + fc_decay: 0 diff --git a/PaddleOCR2Pytorch/configs/table/table_mv3_table_structure.yml b/PaddleOCR2Pytorch/configs/table/table_mv3_table_structure.yml new file mode 100644 index 0000000000000000000000000000000000000000..6398e253935d1182a3786c1c64435a86fcd46b2d --- /dev/null +++ b/PaddleOCR2Pytorch/configs/table/table_mv3_table_structure.yml @@ -0,0 +1,15 @@ +Architecture: + model_type: table + algorithm: TableAttn + Backbone: + name: MobileNetV3 + scale: 1.0 + model_name: large + disable_se: True + Head: + name: TableAttentionHead + hidden_size: 256 + l2_decay: 0.00001 + loc_type: 2 + max_elem_length: 800 + in_max_len: 512 diff --git a/PaddleOCR2Pytorch/en_ptocr_v4_rec_infer.pth b/PaddleOCR2Pytorch/en_ptocr_v4_rec_infer.pth new file mode 100644 index 0000000000000000000000000000000000000000..fefa1adf38f398353cbe24c534aa6a2637b04f36 --- /dev/null +++ b/PaddleOCR2Pytorch/en_ptocr_v4_rec_infer.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f9723c94847de59df9b059faaed041fbb35f014ea973da40290b1e97004e1d4e +size 23761948