From d623bf9b31f82b80cc093486bd701969bb860993 Mon Sep 17 00:00:00 2001 From: tink2123 Date: Wed, 24 Nov 2021 03:51:58 +0000 Subject: [PATCH] add rec ctc model for tipc --- .../rec_icdar15_train.yml | 95 ++++++++++++++++ .../train_infer_python.txt | 51 +++++++++ .../rec_icdar15_train.yml | 96 +++++++++++++++++ .../train_infer_python.txt | 51 +++++++++ .../rec_icdar15_train.yml | 101 ++++++++++++++++++ .../train_infer_python.txt | 51 +++++++++ .../rec_icdar15_train.yml | 96 +++++++++++++++++ .../train_infer_python.txt | 51 +++++++++ .../rec_icdar15_train.yml | 94 ++++++++++++++++ .../train_infer_python.txt | 51 +++++++++ .../rec_icdar15_train.yml | 100 +++++++++++++++++ .../train_infer_python.txt | 51 +++++++++ test_tipc/test_train_inference_python.sh | 6 +- 13 files changed, 891 insertions(+), 3 deletions(-) create mode 100644 test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/rec_icdar15_train.yml create mode 100644 test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/train_infer_python.txt create mode 100644 test_tipc/configs/rec_mv3_none_none_ctc_v2.0/rec_icdar15_train.yml create mode 100644 test_tipc/configs/rec_mv3_none_none_ctc_v2.0/train_infer_python.txt create mode 100644 test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml create mode 100644 test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/train_infer_python.txt create mode 100644 test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/rec_icdar15_train.yml create mode 100644 test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/train_infer_python.txt create mode 100644 test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/rec_icdar15_train.yml create mode 100644 test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/train_infer_python.txt create mode 100644 test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml create mode 100644 test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/train_infer_python.txt diff --git a/test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/rec_icdar15_train.yml b/test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/rec_icdar15_train.yml new file mode 100644 index 00000000..9a950923 --- /dev/null +++ b/test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/rec_icdar15_train.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_bilstm_ctc/ + save_epoch_step: 3 + # 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: + max_text_length: 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_mv3_none_bilstm_ctc.txt + +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/test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/train_infer_python.txt b/test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/train_infer_python.txt new file mode 100644 index 00000000..79725e7b --- /dev/null +++ b/test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/train_infer_python.txt @@ -0,0 +1,51 @@ +===========================train_params=========================== +model_name:ocr_rec +python:python3.7 +gpu_list:0|0,1 +Global.use_gpu:True|True +Global.auto_cast:null +Global.epoch_num:lite_train_infer=2|whole_train_infer=300 +Global.save_model_dir:./output/ +Train.loader.batch_size_per_card:lite_train_infer=128|whole_train_infer=128 +Global.pretrained_model:null +train_model_name:latest +train_infer_img_dir:./inference/rec_inference +null:null +## +trainer:norm_train +norm_train:tools/train.py -c test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +pact_train:null +fpgm_train:null +distill_train:null +null:null +null:null +## +===========================eval_params=========================== +eval:tools/eval.py -c test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +null:null +## +===========================infer_params=========================== +Global.save_inference_dir:./output/ +Global.pretrained_model: +norm_export:tools/export_model.py -c test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +quant_export:null +fpgm_export:null +distill_export:null +export1:null +export2:null +## +infer_model:null +infer_export:tools/export_model.py -c test_tipc/configs/rec_mv3_none_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +infer_quant:False +inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100" +--use_gpu:True|False +--enable_mkldnn:True|False +--cpu_threads:1|6 +--rec_batch_num:1|6 +--use_tensorrt:True|False +--precision:fp32|fp16|int8 +--rec_model_dir: +--image_dir:./inference/rec_inference +--save_log_path:./test/output/ +--benchmark:True +null:null diff --git a/test_tipc/configs/rec_mv3_none_none_ctc_v2.0/rec_icdar15_train.yml b/test_tipc/configs/rec_mv3_none_none_ctc_v2.0/rec_icdar15_train.yml new file mode 100644 index 00000000..06a2b71e --- /dev/null +++ b/test_tipc/configs/rec_mv3_none_none_ctc_v2.0/rec_icdar15_train.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_none_ctc/ + save_epoch_step: 3 + # 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: + max_text_length: 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_mv3_none_none_ctc.txt + +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: SimpleDataSet + data_dir: ./train_data/ic15_data/ + label_file_list: ["./train_data/ic15_data/rec_gt_train.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 + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ic15_data + label_file_list: ["./train_data/ic15_data/rec_gt_test.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: 8 diff --git a/test_tipc/configs/rec_mv3_none_none_ctc_v2.0/train_infer_python.txt b/test_tipc/configs/rec_mv3_none_none_ctc_v2.0/train_infer_python.txt new file mode 100644 index 00000000..4b4137a8 --- /dev/null +++ b/test_tipc/configs/rec_mv3_none_none_ctc_v2.0/train_infer_python.txt @@ -0,0 +1,51 @@ +===========================train_params=========================== +model_name:ocr_rec +python:python3.7 +gpu_list:0|0,1 +Global.use_gpu:True|True +Global.auto_cast:null +Global.epoch_num:lite_train_infer=2|whole_train_infer=300 +Global.save_model_dir:./output/ +Train.loader.batch_size_per_card:lite_train_infer=128|whole_train_infer=128 +Global.pretrained_model:null +train_model_name:latest +train_infer_img_dir:./inference/rec_inference +null:null +## +trainer:norm_train +norm_train:tools/train.py -c test_tipc/configs/rec_mv3_none_none_ctc_v2.0/rec_icdar15_train.yml -o +pact_train:null +fpgm_train:null +distill_train:null +null:null +null:null +## +===========================eval_params=========================== +eval:tools/eval.py -c test_tipc/configs/rec_mv3_none_none_ctc_v2.0/rec_icdar15_train.yml -o +null:null +## +===========================infer_params=========================== +Global.save_inference_dir:./output/ +Global.pretrained_model: +norm_export:tools/export_model.py -c test_tipc/configs/rec_mv3_none_none_ctc_v2.0/rec_icdar15_train.yml -o +quant_export:null +fpgm_export:null +distill_export:null +export1:null +export2:null +## +infer_model:null +infer_export:tools/export_model.py -c test_tipc/configs/rec_mv3_none_none_ctc_v2.0/rec_icdar15_train.yml -o +infer_quant:False +inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100" +--use_gpu:True|False +--enable_mkldnn:True|False +--cpu_threads:1|6 +--rec_batch_num:1|6 +--use_tensorrt:True|False +--precision:fp32|fp16|int8 +--rec_model_dir: +--image_dir:./inference/rec_inference +--save_log_path:./test/output/ +--benchmark:True +null:null diff --git a/test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml b/test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml new file mode 100644 index 00000000..e20ea00f --- /dev/null +++ b/test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml @@ -0,0 +1,101 @@ +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] + 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: + max_text_length: 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_mv3_tps_bilstm_ctc.txt + +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: SimpleDataSet + data_dir: ./train_data/ic15_data/ + label_file_list: ["./train_data/ic15_data/rec_gt_train.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 + batch_size_per_card: 256 + drop_last: True + num_workers: 8 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ic15_data + label_file_list: ["./train_data/ic15_data/rec_gt_test.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 diff --git a/test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/train_infer_python.txt b/test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/train_infer_python.txt new file mode 100644 index 00000000..7ab8ceab --- /dev/null +++ b/test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/train_infer_python.txt @@ -0,0 +1,51 @@ +===========================train_params=========================== +model_name:ocr_rec +python:python3.7 +gpu_list:0|0,1 +Global.use_gpu:True|True +Global.auto_cast:null +Global.epoch_num:lite_train_infer=2|whole_train_infer=300 +Global.save_model_dir:./output/ +Train.loader.batch_size_per_card:lite_train_infer=128|whole_train_infer=128 +Global.pretrained_model:null +train_model_name:latest +train_infer_img_dir:./inference/rec_inference +null:null +## +trainer:norm_train +norm_train:tools/train.py -c test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +pact_train:null +fpgm_train:null +distill_train:null +null:null +null:null +## +===========================eval_params=========================== +eval:tools/eval.py -c test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +null:null +## +===========================infer_params=========================== +Global.save_inference_dir:./output/ +Global.pretrained_model: +norm_export:tools/export_model.py -c test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +quant_export:null +fpgm_export:null +distill_export:null +export1:null +export2:null +## +infer_model:null +infer_export:tools/export_model.py -c test_tipc/configs/rec_mv3_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +infer_quant:False +inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100" +--use_gpu:True|False +--enable_mkldnn:True|False +--cpu_threads:1|6 +--rec_batch_num:1|6 +--use_tensorrt:True|False +--precision:fp32|fp16|int8 +--rec_model_dir: +--image_dir:./inference/rec_inference +--save_log_path:./test/output/ +--benchmark:True +null:null diff --git a/test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/rec_icdar15_train.yml b/test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/rec_icdar15_train.yml new file mode 100644 index 00000000..3e9ea486 --- /dev/null +++ b/test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/rec_icdar15_train.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/r34_vd_none_bilstm_ctc/ + save_epoch_step: 3 + # 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: + max_text_length: 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_r34_vd_none_bilstm_ctc.txt + +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: SimpleDataSet + data_dir: ./train_data/ic15_data/ + label_file_list: ["./train_data/ic15_data/rec_gt_train.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 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ic15_data + label_file_list: ["./train_data/ic15_data/rec_gt_test.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 diff --git a/test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/train_infer_python.txt b/test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/train_infer_python.txt new file mode 100644 index 00000000..0cbeca8a --- /dev/null +++ b/test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/train_infer_python.txt @@ -0,0 +1,51 @@ +===========================train_params=========================== +model_name:ocr_rec +python:python3.7 +gpu_list:0|0,1 +Global.use_gpu:True|True +Global.auto_cast:null +Global.epoch_num:lite_train_infer=2|whole_train_infer=300 +Global.save_model_dir:./output/ +Train.loader.batch_size_per_card:lite_train_infer=128|whole_train_infer=128 +Global.pretrained_model:null +train_model_name:latest +train_infer_img_dir:./inference/rec_inference +null:null +## +trainer:norm_train +norm_train:tools/train.py -c test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +pact_train:null +fpgm_train:null +distill_train:null +null:null +null:null +## +===========================eval_params=========================== +eval:tools/eval.py -c test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +null:null +## +===========================infer_params=========================== +Global.save_inference_dir:./output/ +Global.pretrained_model: +norm_export:tools/export_model.py -c test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +quant_export:null +fpgm_export:null +distill_export:null +export1:null +export2:null +## +infer_model:null +infer_export:tools/export_model.py -c test_tipc/configs/rec_r34_vd_none_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +infer_quant:False +inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100" +--use_gpu:True|False +--enable_mkldnn:True|False +--cpu_threads:1|6 +--rec_batch_num:1|6 +--use_tensorrt:True|False +--precision:fp32|fp16|int8 +--rec_model_dir: +--image_dir:./inference/rec_inference +--save_log_path:./test/output/ +--benchmark:True +null:null diff --git a/test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/rec_icdar15_train.yml b/test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/rec_icdar15_train.yml new file mode 100644 index 00000000..ce15e277 --- /dev/null +++ b/test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/rec_icdar15_train.yml @@ -0,0 +1,94 @@ +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] + 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: + max_text_length: 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_r34_vd_none_none_ctc.txt + +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: SimpleDataSet + data_dir: ./train_data/ic15_data/ + label_file_list: ["./train_data/ic15_data/rec_gt_train.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 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ic15_data + label_file_list: ["./train_data/ic15_data/rec_gt_test.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 diff --git a/test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/train_infer_python.txt b/test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/train_infer_python.txt new file mode 100644 index 00000000..a90cf105 --- /dev/null +++ b/test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/train_infer_python.txt @@ -0,0 +1,51 @@ +===========================train_params=========================== +model_name:ocr_rec +python:python3.7 +gpu_list:0|0,1 +Global.use_gpu:True|True +Global.auto_cast:null +Global.epoch_num:lite_train_infer=2|whole_train_infer=300 +Global.save_model_dir:./output/ +Train.loader.batch_size_per_card:lite_train_infer=128|whole_train_infer=128 +Global.pretrained_model:null +train_model_name:latest +train_infer_img_dir:./inference/rec_inference +null:null +## +trainer:norm_train +norm_train:tools/train.py -c test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/rec_icdar15_train.yml -o +pact_train:null +fpgm_train:null +distill_train:null +null:null +null:null +## +===========================eval_params=========================== +eval:tools/eval.py -c test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/rec_icdar15_train.yml -o +null:null +## +===========================infer_params=========================== +Global.save_inference_dir:./output/ +Global.pretrained_model: +norm_export:tools/export_model.py -c test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/rec_icdar15_train.yml -o +quant_export:null +fpgm_export:null +distill_export:null +export1:null +export2:null +## +infer_model:null +infer_export:tools/export_model.py -c test_tipc/configs/rec_r34_vd_none_none_ctc_v2.0/rec_icdar15_train.yml -o +infer_quant:False +inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100" +--use_gpu:True|False +--enable_mkldnn:True|False +--cpu_threads:1|6 +--rec_batch_num:1|6 +--use_tensorrt:True|False +--precision:fp32|fp16|int8 +--rec_model_dir: +--image_dir:./inference/rec_inference +--save_log_path:./test/output/ +--benchmark:True +null:null diff --git a/test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml b/test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml new file mode 100644 index 00000000..2dcba81e --- /dev/null +++ b/test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/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/r34_vd_tps_bilstm_ctc/ + save_epoch_step: 3 + # 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: + max_text_length: 25 + infer_mode: False + use_space_char: False + save_res_path: ./output/rec/predicts_r34_vd_tps_bilstm_ctc.txt + +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: 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: SimpleDataSet + data_dir: ./train_data/ic15_data/ + label_file_list: ["./train_data/ic15_data/rec_gt_train.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 + +Eval: + dataset: + name: SimpleDataSet + data_dir: ./train_data/ic15_data + label_file_list: ["./train_data/ic15_data/rec_gt_test.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 diff --git a/test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/train_infer_python.txt b/test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/train_infer_python.txt new file mode 100644 index 00000000..4a04f5e6 --- /dev/null +++ b/test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/train_infer_python.txt @@ -0,0 +1,51 @@ +===========================train_params=========================== +model_name:ocr_rec +python:python3.7 +gpu_list:0|0,1 +Global.use_gpu:True|True +Global.auto_cast:null +Global.epoch_num:lite_train_infer=2|whole_train_infer=300 +Global.save_model_dir:./output/ +Train.loader.batch_size_per_card:lite_train_infer=128|whole_train_infer=128 +Global.pretrained_model:null +train_model_name:latest +train_infer_img_dir:./inference/rec_inference +null:null +## +trainer:norm_train +norm_train:tools/train.py -c test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +pact_train:null +fpgm_train:null +distill_train:null +null:null +null:null +## +===========================eval_params=========================== +eval:tools/eval.py -c test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +null:null +## +===========================infer_params=========================== +Global.save_inference_dir:./output/ +Global.pretrained_model: +norm_export:tools/export_model.py -c test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +quant_export:null +fpgm_export:null +distill_export:null +export1:null +export2:null +## +infer_model:null +infer_export:tools/export_model.py -c test_tipc/configs/rec_r34_vd_tps_bilstm_ctc_v2.0/rec_icdar15_train.yml -o +infer_quant:False +inference:tools/infer/predict_rec.py --rec_char_dict_path=./ppocr/utils/ic15_dict.txt --rec_image_shape="3,32,100" +--use_gpu:True|False +--enable_mkldnn:True|False +--cpu_threads:1|6 +--rec_batch_num:1|6 +--use_tensorrt:True|False +--precision:fp32|fp16|int8 +--rec_model_dir: +--image_dir:./inference/rec_inference +--save_log_path:./test/output/ +--benchmark:True +null:null diff --git a/test_tipc/test_train_inference_python.sh b/test_tipc/test_train_inference_python.sh index 321eeb55..2b696d87 100644 --- a/test_tipc/test_train_inference_python.sh +++ b/test_tipc/test_train_inference_python.sh @@ -226,7 +226,7 @@ if [ ${MODE} = "whole_infer" ] || [ ${MODE} = "klquant_whole_infer" ]; then set_save_infer_key=$(func_set_params "${save_infer_key}" "${save_infer_dir}") export_cmd="${python} ${infer_run_exports[Count]} ${set_export_weight} ${set_save_infer_key}" echo ${infer_run_exports[Count]} - echo $export_cmd + echo $export_cmd eval $export_cmd status_export=$? status_check $status_export "${export_cmd}" "${status_log}" @@ -336,7 +336,7 @@ else set_eval_pretrain=$(func_set_params "${pretrain_model_key}" "${save_log}/${train_model_name}") # save norm trained models to set pretrain for pact training and fpgm training - if [ ${trainer} = ${trainer_norm} ] && [ ${nodes} -le 1]; then + if [ ${trainer} = ${trainer_norm} ] && [ ${nodes} -le 1 ]; then load_norm_train_model=${set_eval_pretrain} fi # run eval @@ -359,7 +359,7 @@ else #run inference eval $env save_infer_path="${save_log}" - if [ ${inference_dir} != "null" ] && [ ${inference_dir} != '##' ]; then + if [[ ${inference_dir} != "null" ]] && [[ ${inference_dir} != '##' ]]; then infer_model_dir="${save_infer_path}/${inference_dir}" else infer_model_dir=${save_infer_path} -- GitLab