| /workspace/bina02-ppocr/.venv/lib/python3.12/site-packages/paddle/utils/cpp_extension/extension_utils.py:712: UserWarning: No ccache found. Please be aware that recompiling all source files may be required. You can download and install ccache from: https://github.com/ccache/ccache/blob/master/doc/INSTALL.md |
| warnings.warn(warning_message) |
| LAUNCH INFO 2026-07-25 18:12:21,479 ----------- Configuration ---------------------- |
| LAUNCH INFO 2026-07-25 18:12:21,479 auto_cluster_config: 0 |
| LAUNCH INFO 2026-07-25 18:12:21,479 auto_parallel_config: None |
| LAUNCH INFO 2026-07-25 18:12:21,479 auto_tuner_json: None |
| LAUNCH INFO 2026-07-25 18:12:21,479 devices: 0,1,2,3,4,5,6,7 |
| LAUNCH INFO 2026-07-25 18:12:21,479 elastic_level: -1 |
| LAUNCH INFO 2026-07-25 18:12:21,479 elastic_timeout: 30 |
| LAUNCH INFO 2026-07-25 18:12:21,479 enable_gpu_log: True |
| LAUNCH INFO 2026-07-25 18:12:21,479 gloo_port: 6767 |
| LAUNCH INFO 2026-07-25 18:12:21,479 host: None |
| LAUNCH INFO 2026-07-25 18:12:21,479 ips: None |
| LAUNCH INFO 2026-07-25 18:12:21,479 job_id: default |
| LAUNCH INFO 2026-07-25 18:12:21,479 legacy: False |
| LAUNCH INFO 2026-07-25 18:12:21,479 log_dir: /workspace/bina02-ppocr/training/stage-005/distributed-logs |
| LAUNCH INFO 2026-07-25 18:12:21,479 log_level: INFO |
| LAUNCH INFO 2026-07-25 18:12:21,479 log_overwrite: False |
| LAUNCH INFO 2026-07-25 18:12:21,479 master: None |
| LAUNCH INFO 2026-07-25 18:12:21,479 max_restart: 3 |
| LAUNCH INFO 2026-07-25 18:12:21,479 nnodes: 1 |
| LAUNCH INFO 2026-07-25 18:12:21,479 nproc_per_node: None |
| LAUNCH INFO 2026-07-25 18:12:21,479 rank: -1 |
| LAUNCH INFO 2026-07-25 18:12:21,479 run_mode: collective |
| LAUNCH INFO 2026-07-25 18:12:21,479 server_num: None |
| LAUNCH INFO 2026-07-25 18:12:21,479 servers: |
| LAUNCH INFO 2026-07-25 18:12:21,479 sort_ip: False |
| LAUNCH INFO 2026-07-25 18:12:21,479 start_port: 6070 |
| LAUNCH INFO 2026-07-25 18:12:21,479 trainer_num: None |
| LAUNCH INFO 2026-07-25 18:12:21,479 trainers: |
| LAUNCH INFO 2026-07-25 18:12:21,479 training_script: /workspace/bina02-ppocr/PaddleOCR/tools/train.py |
| LAUNCH INFO 2026-07-25 18:12:21,479 training_script_args: ['-c', '/workspace/bina02-ppocr/training/stage-005/input_config.yml'] |
| LAUNCH INFO 2026-07-25 18:12:21,479 with_gloo: 1 |
| LAUNCH INFO 2026-07-25 18:12:21,479 -------------------------------------------------- |
| LAUNCH INFO 2026-07-25 18:12:21,480 Job: default, mode collective, replicas 1[1:1], elastic False |
| LAUNCH INFO 2026-07-25 18:12:21,482 Run Pod: qnbrrq, replicas 8, status ready |
| LAUNCH INFO 2026-07-25 18:12:21,735 Watching Pod: qnbrrq, replicas 8, status running |
| /workspace/bina02-ppocr/.venv/lib/python3.12/site-packages/paddle/utils/cpp_extension/extension_utils.py:712: UserWarning: No ccache found. Please be aware that recompiling all source files may be required. You can download and install ccache from: https://github.com/ccache/ccache/blob/master/doc/INSTALL.md |
| warnings.warn(warning_message) |
| Skipping import of the encryption module. |
| [2026/07/25 18:12:29] ppocr INFO: Architecture : |
| [2026/07/25 18:12:29] ppocr INFO: Backbone : |
| [2026/07/25 18:12:29] ppocr INFO: force_pool_width : True |
| [2026/07/25 18:12:29] ppocr INFO: model_size : medium |
| [2026/07/25 18:12:29] ppocr INFO: name : PPLCNetV4 |
| [2026/07/25 18:12:29] ppocr INFO: train_pool_width : 96 |
| [2026/07/25 18:12:29] ppocr INFO: Head : |
| [2026/07/25 18:12:29] ppocr INFO: head_list : |
| [2026/07/25 18:12:29] ppocr INFO: CTCHead : |
| [2026/07/25 18:12:29] ppocr INFO: Head : |
| [2026/07/25 18:12:29] ppocr INFO: fc_decay : 1e-05 |
| [2026/07/25 18:12:29] ppocr INFO: Neck : |
| [2026/07/25 18:12:29] ppocr INFO: depth : 2 |
| [2026/07/25 18:12:29] ppocr INFO: dims : 192 |
| [2026/07/25 18:12:29] ppocr INFO: local_kernel : 7 |
| [2026/07/25 18:12:29] ppocr INFO: mlp_ratio : 4.0 |
| [2026/07/25 18:12:29] ppocr INFO: name : lightsvtr |
| [2026/07/25 18:12:29] ppocr INFO: use_guide : False |
| [2026/07/25 18:12:29] ppocr INFO: NRTRHead : |
| [2026/07/25 18:12:29] ppocr INFO: max_text_length : 128 |
| [2026/07/25 18:12:29] ppocr INFO: nrtr_dim : 512 |
| [2026/07/25 18:12:29] ppocr INFO: name : MultiHead |
| [2026/07/25 18:12:29] ppocr INFO: Transform : None |
| [2026/07/25 18:12:29] ppocr INFO: algorithm : SVTR_LCNet |
| [2026/07/25 18:12:29] ppocr INFO: model_type : rec |
| [2026/07/25 18:12:29] ppocr INFO: Eval : |
| [2026/07/25 18:12:29] ppocr INFO: dataset : |
| [2026/07/25 18:12:29] ppocr INFO: data_dir : / |
| [2026/07/25 18:12:29] ppocr INFO: label_file_list : ['/workspace/bina02-ppocr/stream/datasets/chunk-005/eval_ctc.txt'] |
| [2026/07/25 18:12:29] ppocr INFO: name : SimpleDataSet |
| [2026/07/25 18:12:29] ppocr INFO: transforms : |
| [2026/07/25 18:12:29] ppocr INFO: DecodeImage : |
| [2026/07/25 18:12:29] ppocr INFO: channel_first : False |
| [2026/07/25 18:12:29] ppocr INFO: img_mode : BGR |
| [2026/07/25 18:12:29] ppocr INFO: MultiLabelEncode : |
| [2026/07/25 18:12:29] ppocr INFO: gtc_encode : NRTRLabelEncode |
| [2026/07/25 18:12:29] ppocr INFO: RecResizeImg : |
| [2026/07/25 18:12:29] ppocr INFO: image_shape : [3, 48, 768] |
| [2026/07/25 18:12:29] ppocr INFO: padding : False |
| [2026/07/25 18:12:29] ppocr INFO: KeepKeys : |
| [2026/07/25 18:12:29] ppocr INFO: keep_keys : ['image', 'label_ctc', 'label_gtc', 'length', 'valid_ratio'] |
| [2026/07/25 18:12:29] ppocr INFO: loader : |
| [2026/07/25 18:12:29] ppocr INFO: batch_size_per_card : 32 |
| [2026/07/25 18:12:29] ppocr INFO: drop_last : False |
| [2026/07/25 18:12:29] ppocr INFO: num_workers : 4 |
| [2026/07/25 18:12:29] ppocr INFO: shuffle : False |
| [2026/07/25 18:12:29] ppocr INFO: Global : |
| [2026/07/25 18:12:29] ppocr INFO: cal_metric_during_train : False |
| [2026/07/25 18:12:29] ppocr INFO: character_dict_path : /workspace/bina02-ppocr/stream/datasets/chunk-005/persian_arabic_bina02_dict.txt |
| [2026/07/25 18:12:29] ppocr INFO: checkpoints : None |
| [2026/07/25 18:12:29] ppocr INFO: d2s_train_image_shape : [3, 48, 768] |
| [2026/07/25 18:12:29] ppocr INFO: debug : False |
| [2026/07/25 18:12:29] ppocr INFO: distributed : True |
| [2026/07/25 18:12:29] ppocr INFO: epoch_num : 1 |
| [2026/07/25 18:12:29] ppocr INFO: eval_batch_step : [0, 100] |
| [2026/07/25 18:12:29] ppocr INFO: freeze_batch_norm : True |
| [2026/07/25 18:12:29] ppocr INFO: infer_img : doc/imgs_words/ch/word_1.jpg |
| [2026/07/25 18:12:29] ppocr INFO: infer_mode : False |
| [2026/07/25 18:12:29] ppocr INFO: log_smooth_window : 20 |
| [2026/07/25 18:12:29] ppocr INFO: max_text_length : 128 |
| [2026/07/25 18:12:29] ppocr INFO: model_name : PP-OCRv6_medium_rec |
| [2026/07/25 18:12:29] ppocr INFO: pretrained_model : /workspace/bina02-ppocr/training/stage-004/best_accuracy |
| [2026/07/25 18:12:29] ppocr INFO: print_batch_step : 20 |
| [2026/07/25 18:12:29] ppocr INFO: save_epoch_step : 1 |
| [2026/07/25 18:12:29] ppocr INFO: save_inference_dir : None |
| [2026/07/25 18:12:29] ppocr INFO: save_model_dir : /workspace/bina02-ppocr/training/stage-005 |
| [2026/07/25 18:12:29] ppocr INFO: save_res_path : /workspace/bina02-ppocr/training/stage-005/predicts.txt |
| [2026/07/25 18:12:29] ppocr INFO: trainable_parameter_prefixes : ['head.ctc_head.fc.'] |
| [2026/07/25 18:12:29] ppocr INFO: use_gpu : True |
| [2026/07/25 18:12:29] ppocr INFO: use_space_char : True |
| [2026/07/25 18:12:29] ppocr INFO: use_visualdl : False |
| [2026/07/25 18:12:29] ppocr INFO: Loss : |
| [2026/07/25 18:12:29] ppocr INFO: loss_config_list : |
| [2026/07/25 18:12:29] ppocr INFO: CTCLoss : None |
| [2026/07/25 18:12:29] ppocr INFO: NRTRLoss : None |
| [2026/07/25 18:12:29] ppocr INFO: name : MultiLoss |
| [2026/07/25 18:12:29] ppocr INFO: weight_1 : 1.0 |
| [2026/07/25 18:12:29] ppocr INFO: weight_2 : 0.0 |
| [2026/07/25 18:12:29] ppocr INFO: Metric : |
| [2026/07/25 18:12:29] ppocr INFO: ignore_space : False |
| [2026/07/25 18:12:29] ppocr INFO: main_indicator : norm_edit_dis |
| [2026/07/25 18:12:29] ppocr INFO: name : RecMetric |
| [2026/07/25 18:12:29] ppocr INFO: Optimizer : |
| [2026/07/25 18:12:29] ppocr INFO: beta1 : 0.9 |
| [2026/07/25 18:12:29] ppocr INFO: beta2 : 0.999 |
| [2026/07/25 18:12:29] ppocr INFO: lr : |
| [2026/07/25 18:12:29] ppocr INFO: learning_rate : 1e-05 |
| [2026/07/25 18:12:29] ppocr INFO: name : Cosine |
| [2026/07/25 18:12:29] ppocr INFO: warmup_epoch : 0 |
| [2026/07/25 18:12:29] ppocr INFO: name : Adam |
| [2026/07/25 18:12:29] ppocr INFO: regularizer : |
| [2026/07/25 18:12:29] ppocr INFO: factor : 0.0 |
| [2026/07/25 18:12:29] ppocr INFO: name : L2 |
| [2026/07/25 18:12:29] ppocr INFO: PostProcess : |
| [2026/07/25 18:12:29] ppocr INFO: name : CTCLabelDecode |
| [2026/07/25 18:12:29] ppocr INFO: Train : |
| [2026/07/25 18:12:29] ppocr INFO: dataset : |
| [2026/07/25 18:12:29] ppocr INFO: data_dir : / |
| [2026/07/25 18:12:29] ppocr INFO: ds_width : False |
| [2026/07/25 18:12:29] ppocr INFO: ext_op_transform_idx : 1 |
| [2026/07/25 18:12:29] ppocr INFO: label_file_list : ['/workspace/bina02-ppocr/stream/datasets/chunk-005/train_ctc.txt'] |
| [2026/07/25 18:12:29] ppocr INFO: name : MultiScaleDataSet |
| [2026/07/25 18:12:29] ppocr INFO: padding : False |
| [2026/07/25 18:12:29] ppocr INFO: ratio_list : [1.0] |
| [2026/07/25 18:12:29] ppocr INFO: transforms : |
| [2026/07/25 18:12:29] ppocr INFO: DecodeImage : |
| [2026/07/25 18:12:29] ppocr INFO: channel_first : False |
| [2026/07/25 18:12:29] ppocr INFO: img_mode : BGR |
| [2026/07/25 18:12:29] ppocr INFO: RecConAug : |
| [2026/07/25 18:12:29] ppocr INFO: ext_data_num : 0 |
| [2026/07/25 18:12:29] ppocr INFO: image_shape : [48, 768, 3] |
| [2026/07/25 18:12:29] ppocr INFO: max_text_length : 128 |
| [2026/07/25 18:12:29] ppocr INFO: prob : 0.0 |
| [2026/07/25 18:12:29] ppocr INFO: RecAug : None |
| [2026/07/25 18:12:29] ppocr INFO: MultiLabelEncode : |
| [2026/07/25 18:12:29] ppocr INFO: gtc_encode : NRTRLabelEncode |
| [2026/07/25 18:12:29] ppocr INFO: KeepKeys : |
| [2026/07/25 18:12:29] ppocr INFO: keep_keys : ['image', 'label_ctc', 'label_gtc', 'length', 'valid_ratio'] |
| [2026/07/25 18:12:29] ppocr INFO: loader : |
| [2026/07/25 18:12:29] ppocr INFO: batch_size_per_card : 32 |
| [2026/07/25 18:12:29] ppocr INFO: drop_last : False |
| [2026/07/25 18:12:29] ppocr INFO: num_workers : 4 |
| [2026/07/25 18:12:29] ppocr INFO: shuffle : True |
| [2026/07/25 18:12:29] ppocr INFO: sampler : |
| [2026/07/25 18:12:29] ppocr INFO: divided_factor : [8, 16] |
| [2026/07/25 18:12:29] ppocr INFO: first_bs : 32 |
| [2026/07/25 18:12:29] ppocr INFO: fix_bs : True |
| [2026/07/25 18:12:29] ppocr INFO: is_training : True |
| [2026/07/25 18:12:29] ppocr INFO: name : MultiScaleSampler |
| [2026/07/25 18:12:29] ppocr INFO: scales : [[768, 48]] |
| [2026/07/25 18:12:29] ppocr INFO: profiler_options : None |
| [2026/07/25 18:12:29] ppocr INFO: train with paddle 3.3.0 and device Place(gpu:0) |
| ======================= Modified FLAGS detected ======================= |
| FLAGS(name='FLAGS_curand_dir', current_value='/workspace/bina02-ppocr/.venv/lib/python3.12/site-packages/paddle/../nvidia/curand/lib', default_value='') |
| FLAGS(name='FLAGS_cudnn_deterministic', current_value=True, default_value=False) |
| FLAGS(name='FLAGS_selected_gpus', current_value='0', default_value='') |
| FLAGS(name='FLAGS_cuda_cccl_dir', current_value='/workspace/bina02-ppocr/.venv/lib/python3.12/site-packages/paddle/../nvidia/cuda_cccl/include/', default_value='') |
| FLAGS(name='FLAGS_nccl_dir', current_value='/workspace/bina02-ppocr/.venv/lib/python3.12/site-packages/paddle/../nvidia/nccl/lib', default_value='') |
| FLAGS(name='FLAGS_cudnn_dir', current_value='/workspace/bina02-ppocr/.venv/lib/python3.12/site-packages/paddle/../nvidia/cudnn/lib', default_value='') |
| FLAGS(name='FLAGS_cusparse_dir', current_value='/workspace/bina02-ppocr/.venv/lib/python3.12/site-packages/paddle/../nvidia/cusparse/lib', default_value='') |
| FLAGS(name='FLAGS_cusolver_dir', current_value='/workspace/bina02-ppocr/.venv/lib/python3.12/site-packages/paddle/../nvidia/cusolver/lib', default_value='') |
| FLAGS(name='FLAGS_enable_pir_in_executor', current_value=True, default_value=False) |
| FLAGS(name='FLAGS_nvidia_package_dir', current_value='/workspace/bina02-ppocr/.venv/lib/python3.12/site-packages/paddle/../nvidia', default_value='') |
| FLAGS(name='FLAGS_cupti_dir', current_value='/workspace/bina02-ppocr/.venv/lib/python3.12/site-packages/paddle/../nvidia/cuda_cupti/lib', default_value='') |
| FLAGS(name='FLAGS_cublas_dir', current_value='/workspace/bina02-ppocr/.venv/lib/python3.12/site-packages/paddle/../nvidia/cublas/lib', default_value='') |
| ======================================================================= |
| I0725 18:12:29.193190 115569 tcp_store.cc:336] create libuv server at port: 53903 |
| I0725 18:12:29.193780 115569 tcp_utils.cc:132] Successfully connected to 172.17.0.2:53903 |
| I0725 18:12:30.652879 115569 process_group_nccl.cc:155] ProcessGroupNCCL pg_timeout_ 1800000 |
| I0725 18:12:30.652935 115569 process_group_nccl.cc:156] ProcessGroupNCCL nccl_comm_init_option_ 0 |
| [2026/07/25 18:12:30] ppocr INFO: Initialize indexs of datasets:['/workspace/bina02-ppocr/stream/datasets/chunk-005/train_ctc.txt'] |
| [2026/07/25 18:12:30] ppocr INFO: Initialize indexs of datasets:['/workspace/bina02-ppocr/stream/datasets/chunk-005/eval_ctc.txt'] |
| W0725 18:12:30.851824 115569 gpu_resources.cc:116] Please NOTE: device: 0, GPU Compute Capability: 12.0, Driver API Version: 13.0, Runtime API Version: 13.0 |
| [2026/07/25 18:12:31] ppocr INFO: froze 61 BatchNorm layers with pretrained global statistics |
| [2026/07/25 18:12:31] ppocr INFO: selective training enabled: 2 trainable tensors, 438 frozen tensors; head.ctc_head.fc.weight, head.ctc_head.fc.bias |
| [2026-07-25 18:12:31,243] [ INFO] optimizer.py:254 - If regularizer of a Parameter has been set by 'paddle.ParamAttr' or 'static.WeightNormParamAttr' already. The weight_decay[0.0] in Optimizer will not take effect, and it will only be applied to other Parameters! |
| [2026/07/25 18:12:31] ppocr INFO: train dataloader has 391 iters |
| [2026/07/25 18:12:31] ppocr INFO: valid dataloader has 128 iters |
| [2026/07/25 18:12:31] ppocr INFO: load pretrain successful from /workspace/bina02-ppocr/training/stage-004/best_accuracy |
| [2026/07/25 18:12:32] ppocr INFO: During the training process, after the 0th iteration, an evaluation is run every 100 iterations |
| [2026/07/25 18:12:37] ppocr INFO: epoch: [1/1], global_step: 20, lr: 0.000010, CTCLoss: 57.700485, NRTRLoss: 0.000000, loss: 57.700485, avg_reader_cost: 0.09118 s, avg_batch_cost: 0.20650 s, avg_samples: 32.0, ips: 154.96476 samples/s, eta: 0:01:16, max_mem_reserved: 1045 MB, max_mem_allocated: 656 MB |
| [2026/07/25 18:12:39] ppocr INFO: epoch: [1/1], global_step: 40, lr: 0.000010, CTCLoss: 58.909187, NRTRLoss: 0.000000, loss: 58.909187, avg_reader_cost: 0.00424 s, avg_batch_cost: 0.08007 s, avg_samples: 32.0, ips: 399.66937 samples/s, eta: 0:00:50, max_mem_reserved: 1045 MB, max_mem_allocated: 656 MB |
| [2026/07/25 18:12:42] ppocr INFO: epoch: [1/1], global_step: 60, lr: 0.000010, CTCLoss: 60.158285, NRTRLoss: 0.000000, loss: 60.158285, avg_reader_cost: 0.00172 s, avg_batch_cost: 0.08121 s, avg_samples: 32.0, ips: 394.03815 samples/s, eta: 0:00:40, max_mem_reserved: 1045 MB, max_mem_allocated: 656 MB |
| [2026/07/25 18:12:45] ppocr INFO: epoch: [1/1], global_step: 80, lr: 0.000009, CTCLoss: 61.277010, NRTRLoss: 0.000000, loss: 61.277010, avg_reader_cost: 0.00168 s, avg_batch_cost: 0.08314 s, avg_samples: 32.0, ips: 384.90298 samples/s, eta: 0:00:35, max_mem_reserved: 1045 MB, max_mem_allocated: 656 MB |
| [2026/07/25 18:12:47] ppocr INFO: epoch: [1/1], global_step: 100, lr: 0.000009, CTCLoss: 54.436443, NRTRLoss: 0.000000, loss: 54.436443, avg_reader_cost: 0.00166 s, avg_batch_cost: 0.08089 s, avg_samples: 32.0, ips: 395.58923 samples/s, eta: 0:00:30, max_mem_reserved: 1045 MB, max_mem_allocated: 656 MB |
|
|
| eval model:: 0%| | 0/128 [00:00<?, ?it/s] |
| eval model:: 1%| | 1/128 [00:00<00:52, 2.43it/s] |
| eval model:: 2%|β | 3/128 [00:00<00:21, 5.82it/s] |
| eval model:: 4%|β | 5/128 [00:00<00:14, 8.30it/s] |
| eval model:: 5%|β | 7/128 [00:00<00:12, 9.32it/s] |
| eval model:: 7%|β | 9/128 [00:01<00:11, 10.55it/s] |
| eval model:: 9%|β | 11/128 [00:01<00:10, 11.49it/s] |
| eval model:: 10%|β | 13/128 [00:01<00:09, 12.28it/s] |
| eval model:: 12%|ββ | 15/128 [00:01<00:09, 12.47it/s] |
| eval model:: 13%|ββ | 17/128 [00:01<00:08, 12.77it/s] |
| eval model:: 15%|ββ | 19/128 [00:01<00:08, 12.35it/s] |
| eval model:: 16%|ββ | 21/128 [00:01<00:08, 12.16it/s] |
| eval model:: 18%|ββ | 23/128 [00:02<00:08, 12.61it/s] |
| eval model:: 20%|ββ | 25/128 [00:02<00:07, 13.05it/s] |
| eval model:: 21%|ββ | 27/128 [00:02<00:07, 13.31it/s] |
| eval model:: 23%|βββ | 29/128 [00:02<00:07, 13.55it/s] |
| eval model:: 24%|βββ | 31/128 [00:02<00:07, 13.71it/s] |
| eval model:: 26%|βββ | 33/128 [00:02<00:06, 13.84it/s] |
| eval model:: 27%|βββ | 35/128 [00:03<00:06, 13.30it/s] |
| eval model:: 29%|βββ | 37/128 [00:03<00:07, 12.63it/s] |
| eval model:: 30%|βββ | 39/128 [00:03<00:06, 12.88it/s] |
| eval model:: 32%|ββββ | 41/128 [00:03<00:06, 13.23it/s] |
| eval model:: 34%|ββββ | 43/128 [00:03<00:06, 13.14it/s] |
| eval model:: 35%|ββββ | 45/128 [00:03<00:06, 13.02it/s] |
| eval model:: 37%|ββββ | 47/128 [00:03<00:06, 12.21it/s] |
| eval model:: 38%|ββββ | 49/128 [00:04<00:06, 12.47it/s] |
| eval model:: 40%|ββββ | 51/128 [00:04<00:05, 12.90it/s] |
| eval model:: 41%|βββββ | 53/128 [00:04<00:05, 13.05it/s] |
| eval model:: 43%|βββββ | 55/128 [00:04<00:05, 13.14it/s] |
| eval model:: 45%|βββββ | 57/128 [00:04<00:05, 12.45it/s] |
| eval model:: 46%|βββββ | 59/128 [00:04<00:05, 12.91it/s] |
| eval model:: 48%|βββββ | 61/128 [00:05<00:05, 13.27it/s] |
| eval model:: 49%|βββββ | 63/128 [00:05<00:05, 12.44it/s] |
| eval model:: 51%|βββββ | 65/128 [00:05<00:05, 12.35it/s] |
| eval model:: 52%|ββββββ | 67/128 [00:05<00:04, 12.45it/s] |
| eval model:: 54%|ββββββ | 69/128 [00:05<00:04, 11.81it/s] |
| eval model:: 55%|ββββββ | 71/128 [00:05<00:04, 12.03it/s] |
| eval model:: 57%|ββββββ | 73/128 [00:06<00:04, 12.54it/s] |
| eval model:: 59%|ββββββ | 75/128 [00:06<00:04, 12.89it/s] |
| eval model:: 60%|ββββββ | 77/128 [00:06<00:03, 13.16it/s] |
| eval model:: 62%|βββββββ | 79/128 [00:06<00:03, 13.43it/s] |
| eval model:: 63%|βββββββ | 81/128 [00:06<00:03, 13.53it/s] |
| eval model:: 65%|βββββββ | 83/128 [00:06<00:03, 13.54it/s] |
| eval model:: 66%|βββββββ | 85/128 [00:06<00:03, 13.68it/s] |
| eval model:: 68%|βββββββ | 87/128 [00:07<00:02, 13.68it/s] |
| eval model:: 70%|βββββββ | 89/128 [00:07<00:02, 13.83it/s] |
| eval model:: 71%|βββββββ | 91/128 [00:07<00:02, 13.81it/s] |
| eval model:: 73%|ββββββββ | 93/128 [00:07<00:02, 13.94it/s] |
| eval model:: 74%|ββββββββ | 95/128 [00:07<00:02, 13.99it/s] |
| eval model:: 76%|ββββββββ | 97/128 [00:07<00:02, 14.04it/s] |
| eval model:: 77%|ββββββββ | 99/128 [00:07<00:02, 13.81it/s] |
| eval model:: 79%|ββββββββ | 101/128 [00:08<00:01, 13.90it/s] |
| eval model:: 80%|ββββββββ | 103/128 [00:08<00:01, 14.03it/s] |
| eval model:: 82%|βββββββββ | 105/128 [00:08<00:01, 14.04it/s] |
| eval model:: 84%|βββββββββ | 107/128 [00:08<00:01, 14.07it/s] |
| eval model:: 85%|βββββββββ | 109/128 [00:08<00:01, 13.43it/s] |
| eval model:: 87%|βββββββββ | 111/128 [00:08<00:01, 13.61it/s] |
| eval model:: 88%|βββββββββ | 113/128 [00:08<00:01, 13.50it/s] |
| eval model:: 90%|βββββββββ | 115/128 [00:09<00:01, 12.38it/s] |
| eval model:: 91%|ββββββββββ| 117/128 [00:09<00:00, 11.63it/s] |
| eval model:: 93%|ββββββββββ| 119/128 [00:09<00:00, 11.84it/s] |
| eval model:: 95%|ββββββββββ| 121/128 [00:09<00:00, 11.56it/s] |
| eval model:: 96%|ββββββββββ| 123/128 [00:09<00:00, 11.53it/s] |
| eval model:: 98%|ββββββββββ| 125/128 [00:09<00:00, 12.23it/s] |
| eval model:: 99%|ββββββββββ| 127/128 [00:10<00:00, 12.26it/s] |
| eval model:: 100%|ββββββββββ| 128/128 [00:10<00:00, 12.37it/s] |
| [2026/07/25 18:12:58] ppocr INFO: cur metric, acc: 0.8491210916769505, norm_edit_dis: 0.9879758574858228, fps: 438.04331608250544 |
| [2026/07/25 18:12:58] ppocr INFO: save best model is to /workspace/bina02-ppocr/training/stage-005/best_accuracy |
| [2026/07/25 18:12:58] ppocr INFO: best metric, norm_edit_dis: 0.9879758574858228, is_float16: False, acc: 0.8491210916769505, fps: 438.04331608250544, best_epoch: 1 |
| [2026/07/25 18:13:01] ppocr INFO: epoch: [1/1], global_step: 120, lr: 0.000008, CTCLoss: 56.717579, NRTRLoss: 0.000000, loss: 56.717579, avg_reader_cost: 0.03458 s, avg_batch_cost: 0.11146 s, avg_samples: 32.0, ips: 287.09051 samples/s, eta: 0:00:29, max_mem_reserved: 1045 MB, max_mem_allocated: 672 MB |
| [2026/07/25 18:13:03] ppocr INFO: epoch: [1/1], global_step: 140, lr: 0.000008, CTCLoss: 60.334568, NRTRLoss: 0.000000, loss: 60.334568, avg_reader_cost: 0.04406 s, avg_batch_cost: 0.12459 s, avg_samples: 32.0, ips: 256.83833 samples/s, eta: 0:00:27, max_mem_reserved: 1045 MB, max_mem_allocated: 672 MB |
| [2026/07/25 18:13:06] ppocr INFO: epoch: [1/1], global_step: 160, lr: 0.000007, CTCLoss: 57.639336, NRTRLoss: 0.000000, loss: 57.639336, avg_reader_cost: 0.00626 s, avg_batch_cost: 0.08457 s, avg_samples: 32.0, ips: 378.36305 samples/s, eta: 0:00:24, max_mem_reserved: 1045 MB, max_mem_allocated: 672 MB |
| [2026/07/25 18:13:08] ppocr INFO: epoch: [1/1], global_step: 180, lr: 0.000006, CTCLoss: 55.047016, NRTRLoss: 0.000000, loss: 55.047016, avg_reader_cost: 0.00168 s, avg_batch_cost: 0.08111 s, avg_samples: 32.0, ips: 394.54932 samples/s, eta: 0:00:21, max_mem_reserved: 1045 MB, max_mem_allocated: 672 MB |
| [2026/07/25 18:13:11] ppocr INFO: epoch: [1/1], global_step: 200, lr: 0.000005, CTCLoss: 59.132174, NRTRLoss: 0.000000, loss: 59.132174, avg_reader_cost: 0.00171 s, avg_batch_cost: 0.08207 s, avg_samples: 32.0, ips: 389.91562 samples/s, eta: 0:00:19, max_mem_reserved: 1045 MB, max_mem_allocated: 672 MB |
|
|
| eval model:: 0%| | 0/128 [00:00<?, ?it/s] |
| eval model:: 1%| | 1/128 [00:00<00:55, 2.30it/s] |
| eval model:: 2%|β | 2/128 [00:00<00:30, 4.19it/s] |
| eval model:: 3%|β | 4/128 [00:00<00:17, 7.17it/s] |
| eval model:: 5%|β | 6/128 [00:00<00:13, 9.27it/s] |
| eval model:: 6%|β | 8/128 [00:01<00:12, 9.99it/s] |
| eval model:: 8%|β | 10/128 [00:01<00:11, 10.48it/s] |
| eval model:: 9%|β | 12/128 [00:01<00:10, 11.26it/s] |
| eval model:: 11%|β | 14/128 [00:01<00:09, 11.80it/s] |
| eval model:: 12%|ββ | 16/128 [00:01<00:09, 12.44it/s] |
| eval model:: 14%|ββ | 18/128 [00:01<00:08, 12.29it/s] |
| eval model:: 16%|ββ | 20/128 [00:01<00:08, 12.09it/s] |
| eval model:: 17%|ββ | 22/128 [00:02<00:08, 11.83it/s] |
| eval model:: 19%|ββ | 24/128 [00:02<00:08, 12.25it/s] |
| eval model:: 20%|ββ | 26/128 [00:02<00:08, 12.59it/s] |
| eval model:: 22%|βββ | 28/128 [00:02<00:07, 13.03it/s] |
| eval model:: 23%|βββ | 30/128 [00:02<00:07, 12.65it/s] |
| eval model:: 25%|βββ | 32/128 [00:02<00:07, 12.84it/s] |
| eval model:: 27%|βββ | 34/128 [00:03<00:07, 12.82it/s] |
| eval model:: 28%|βββ | 36/128 [00:03<00:07, 12.55it/s] |
| eval model:: 30%|βββ | 38/128 [00:03<00:06, 13.00it/s] |
| eval model:: 31%|ββββ | 40/128 [00:03<00:06, 13.29it/s] |
| eval model:: 33%|ββββ | 42/128 [00:03<00:06, 13.44it/s] |
| eval model:: 34%|ββββ | 44/128 [00:03<00:06, 13.06it/s] |
| eval model:: 36%|ββββ | 46/128 [00:03<00:06, 13.24it/s] |
| eval model:: 38%|ββββ | 48/128 [00:04<00:05, 13.49it/s] |
| eval model:: 39%|ββββ | 50/128 [00:04<00:05, 13.60it/s] |
| eval model:: 41%|ββββ | 52/128 [00:04<00:05, 13.68it/s] |
| eval model:: 42%|βββββ | 54/128 [00:04<00:05, 13.68it/s] |
| eval model:: 44%|βββββ | 56/128 [00:04<00:05, 13.78it/s] |
| eval model:: 45%|βββββ | 58/128 [00:04<00:05, 13.77it/s] |
| eval model:: 47%|βββββ | 60/128 [00:04<00:04, 13.93it/s] |
| eval model:: 48%|βββββ | 62/128 [00:05<00:04, 13.90it/s] |
| eval model:: 50%|βββββ | 64/128 [00:05<00:04, 13.94it/s] |
| eval model:: 52%|ββββββ | 66/128 [00:05<00:04, 13.86it/s] |
| eval model:: 53%|ββββββ | 68/128 [00:05<00:04, 13.97it/s] |
| eval model:: 55%|ββββββ | 70/128 [00:05<00:04, 13.95it/s] |
| eval model:: 56%|ββββββ | 72/128 [00:05<00:04, 13.93it/s] |
| eval model:: 58%|ββββββ | 74/128 [00:05<00:03, 13.85it/s] |
| eval model:: 59%|ββββββ | 76/128 [00:06<00:03, 13.91it/s] |
| eval model:: 61%|ββββββ | 78/128 [00:06<00:03, 13.65it/s] |
| eval model:: 62%|βββββββ | 80/128 [00:06<00:03, 13.50it/s] |
| eval model:: 64%|βββββββ | 82/128 [00:06<00:03, 13.18it/s] |
| eval model:: 66%|βββββββ | 84/128 [00:06<00:03, 13.41it/s] |
| eval model:: 67%|βββββββ | 86/128 [00:06<00:03, 12.53it/s] |
| eval model:: 69%|βββββββ | 88/128 [00:07<00:03, 12.70it/s] |
| eval model:: 70%|βββββββ | 90/128 [00:07<00:03, 11.93it/s] |
| eval model:: 72%|ββββββββ | 92/128 [00:07<00:02, 12.21it/s] |
| eval model:: 73%|ββββββββ | 94/128 [00:07<00:02, 12.42it/s] |
| eval model:: 75%|ββββββββ | 96/128 [00:07<00:02, 12.86it/s] |
| eval model:: 77%|ββββββββ | 98/128 [00:07<00:02, 12.95it/s] |
| eval model:: 78%|ββββββββ | 100/128 [00:08<00:02, 12.89it/s] |
| eval model:: 80%|ββββββββ | 102/128 [00:08<00:02, 12.38it/s] |
| eval model:: 81%|βββββββββ | 104/128 [00:08<00:01, 12.68it/s] |
| eval model:: 83%|βββββββββ | 106/128 [00:08<00:01, 13.07it/s] |
| eval model:: 84%|βββββββββ | 108/128 [00:08<00:01, 12.72it/s] |
| eval model:: 86%|βββββββββ | 110/128 [00:08<00:01, 12.47it/s] |
| eval model:: 88%|βββββββββ | 112/128 [00:08<00:01, 12.71it/s] |
| eval model:: 89%|βββββββββ | 114/128 [00:09<00:01, 12.45it/s] |
| eval model:: 91%|βββββββββ | 116/128 [00:09<00:00, 12.95it/s] |
| eval model:: 92%|ββββββββββ| 118/128 [00:09<00:00, 12.79it/s] |
| eval model:: 94%|ββββββββββ| 120/128 [00:09<00:00, 13.14it/s] |
| eval model:: 95%|ββββββββββ| 122/128 [00:09<00:00, 13.44it/s] |
| eval model:: 97%|ββββββββββ| 124/128 [00:09<00:00, 13.69it/s] |
| eval model:: 98%|ββββββββββ| 126/128 [00:10<00:00, 13.81it/s] |
| eval model:: 100%|ββββββββββ| 128/128 [00:10<00:00, 13.93it/s] |
| eval model:: 100%|ββββββββββ| 128/128 [00:10<00:00, 12.39it/s] |
| [2026/07/25 18:13:21] ppocr INFO: cur metric, acc: 0.8491210916769505, norm_edit_dis: 0.9880078785722896, fps: 443.62973027675116 |
| [2026/07/25 18:13:22] ppocr INFO: save best model is to /workspace/bina02-ppocr/training/stage-005/best_accuracy |
| [2026/07/25 18:13:22] ppocr INFO: best metric, norm_edit_dis: 0.9880078785722896, is_float16: False, acc: 0.8491210916769505, fps: 443.62973027675116, best_epoch: 1 |
| [2026/07/25 18:13:24] ppocr INFO: epoch: [1/1], global_step: 220, lr: 0.000004, CTCLoss: 59.328920, NRTRLoss: 0.000000, loss: 59.328920, avg_reader_cost: 0.01462 s, avg_batch_cost: 0.09255 s, avg_samples: 32.0, ips: 345.74657 samples/s, eta: 0:00:17, max_mem_reserved: 1045 MB, max_mem_allocated: 672 MB |
| [2026/07/25 18:13:27] ppocr INFO: epoch: [1/1], global_step: 240, lr: 0.000004, CTCLoss: 59.362158, NRTRLoss: 0.000000, loss: 59.362158, avg_reader_cost: 0.05295 s, avg_batch_cost: 0.13442 s, avg_samples: 32.0, ips: 238.05987 samples/s, eta: 0:00:15, max_mem_reserved: 1045 MB, max_mem_allocated: 672 MB |
| [2026/07/25 18:13:30] ppocr INFO: epoch: [1/1], global_step: 260, lr: 0.000003, CTCLoss: 57.535942, NRTRLoss: 0.000000, loss: 57.535942, avg_reader_cost: 0.00180 s, avg_batch_cost: 0.07875 s, avg_samples: 32.0, ips: 406.32988 samples/s, eta: 0:00:13, max_mem_reserved: 1045 MB, max_mem_allocated: 672 MB |
| [2026/07/25 18:13:32] ppocr INFO: epoch: [1/1], global_step: 280, lr: 0.000002, CTCLoss: 65.895004, NRTRLoss: 0.000000, loss: 65.895004, avg_reader_cost: 0.00201 s, avg_batch_cost: 0.08196 s, avg_samples: 32.0, ips: 390.45759 samples/s, eta: 0:00:11, max_mem_reserved: 1045 MB, max_mem_allocated: 672 MB |
| [2026/07/25 18:13:35] ppocr INFO: epoch: [1/1], global_step: 300, lr: 0.000002, CTCLoss: 57.611902, NRTRLoss: 0.000000, loss: 57.611902, avg_reader_cost: 0.00674 s, avg_batch_cost: 0.09191 s, avg_samples: 32.0, ips: 348.14865 samples/s, eta: 0:00:09, max_mem_reserved: 1045 MB, max_mem_allocated: 672 MB |
|
|
| eval model:: 0%| | 0/128 [00:00<?, ?it/s] |
| eval model:: 1%| | 1/128 [00:00<01:02, 2.04it/s] |
| eval model:: 2%|β | 2/128 [00:00<00:32, 3.83it/s] |
| eval model:: 3%|β | 4/128 [00:00<00:18, 6.70it/s] |
| eval model:: 5%|β | 6/128 [00:00<00:13, 8.89it/s] |
| eval model:: 6%|β | 8/128 [00:01<00:12, 9.99it/s] |
| eval model:: 8%|β | 10/128 [00:01<00:11, 10.61it/s] |
| eval model:: 9%|β | 12/128 [00:01<00:10, 11.42it/s] |
| eval model:: 11%|β | 14/128 [00:01<00:09, 11.98it/s] |
| eval model:: 12%|ββ | 16/128 [00:01<00:08, 12.58it/s] |
| eval model:: 14%|ββ | 18/128 [00:01<00:09, 11.65it/s] |
| eval model:: 16%|ββ | 20/128 [00:02<00:09, 11.76it/s] |
| eval model:: 17%|ββ | 22/128 [00:02<00:08, 12.39it/s] |
| eval model:: 19%|ββ | 24/128 [00:02<00:08, 12.67it/s] |
| eval model:: 20%|ββ | 26/128 [00:02<00:07, 13.06it/s] |
| eval model:: 22%|βββ | 28/128 [00:02<00:07, 13.15it/s] |
| eval model:: 23%|βββ | 30/128 [00:02<00:07, 13.37it/s] |
| eval model:: 25%|βββ | 32/128 [00:02<00:07, 13.34it/s] |
| eval model:: 27%|βββ | 34/128 [00:03<00:06, 13.54it/s] |
| eval model:: 28%|βββ | 36/128 [00:03<00:06, 13.47it/s] |
| eval model:: 30%|βββ | 38/128 [00:03<00:06, 13.62it/s] |
| eval model:: 31%|ββββ | 40/128 [00:03<00:06, 13.67it/s] |
| eval model:: 33%|ββββ | 42/128 [00:03<00:06, 13.79it/s] |
| eval model:: 34%|ββββ | 44/128 [00:03<00:06, 13.85it/s] |
| eval model:: 36%|ββββ | 46/128 [00:03<00:05, 13.82it/s] |
| eval model:: 38%|ββββ | 48/128 [00:04<00:05, 13.85it/s] |
| eval model:: 39%|ββββ | 50/128 [00:04<00:05, 13.95it/s] |
| eval model:: 41%|ββββ | 52/128 [00:04<00:05, 13.96it/s] |
| eval model:: 42%|βββββ | 54/128 [00:04<00:05, 12.89it/s] |
| eval model:: 44%|βββββ | 56/128 [00:04<00:05, 12.11it/s] |
| eval model:: 45%|βββββ | 58/128 [00:04<00:05, 12.17it/s] |
| eval model:: 47%|βββββ | 60/128 [00:05<00:05, 12.40it/s] |
| eval model:: 48%|βββββ | 62/128 [00:05<00:05, 12.84it/s] |
| eval model:: 50%|βββββ | 64/128 [00:05<00:04, 12.92it/s] |
| eval model:: 52%|ββββββ | 66/128 [00:05<00:04, 12.91it/s] |
| eval model:: 53%|ββββββ | 68/128 [00:05<00:05, 11.82it/s] |
| eval model:: 55%|ββββββ | 70/128 [00:05<00:04, 11.73it/s] |
| eval model:: 56%|ββββββ | 72/128 [00:06<00:04, 12.15it/s] |
| eval model:: 58%|ββββββ | 74/128 [00:06<00:04, 12.48it/s] |
| eval model:: 59%|ββββββ | 76/128 [00:06<00:04, 12.29it/s] |
| eval model:: 61%|ββββββ | 78/128 [00:06<00:04, 12.07it/s] |
| eval model:: 62%|βββββββ | 80/128 [00:06<00:04, 11.76it/s] |
| eval model:: 64%|βββββββ | 82/128 [00:06<00:03, 12.12it/s] |
| eval model:: 66%|βββββββ | 84/128 [00:06<00:03, 12.65it/s] |
| eval model:: 67%|βββββββ | 86/128 [00:07<00:03, 12.73it/s] |
| eval model:: 69%|βββββββ | 88/128 [00:07<00:03, 13.00it/s] |
| eval model:: 70%|βββββββ | 90/128 [00:07<00:02, 13.26it/s] |
| eval model:: 72%|ββββββββ | 92/128 [00:07<00:02, 13.43it/s] |
| eval model:: 73%|ββββββββ | 94/128 [00:07<00:02, 13.56it/s] |
| eval model:: 75%|ββββββββ | 96/128 [00:07<00:02, 13.60it/s] |
| eval model:: 77%|ββββββββ | 98/128 [00:08<00:02, 13.48it/s] |
| eval model:: 78%|ββββββββ | 100/128 [00:08<00:02, 13.51it/s] |
| eval model:: 80%|ββββββββ | 102/128 [00:08<00:01, 13.67it/s] |
| eval model:: 81%|βββββββββ | 104/128 [00:08<00:01, 13.73it/s] |
| eval model:: 83%|βββββββββ | 106/128 [00:08<00:01, 13.72it/s] |
| eval model:: 84%|βββββββββ | 108/128 [00:08<00:01, 13.84it/s] |
| eval model:: 86%|βββββββββ | 110/128 [00:08<00:01, 13.84it/s] |
| eval model:: 88%|βββββββββ | 112/128 [00:09<00:01, 13.74it/s] |
| eval model:: 89%|βββββββββ | 114/128 [00:09<00:01, 13.78it/s] |
| eval model:: 91%|βββββββββ | 116/128 [00:09<00:00, 13.84it/s] |
| eval model:: 92%|ββββββββββ| 118/128 [00:09<00:00, 13.76it/s] |
| eval model:: 94%|ββββββββββ| 120/128 [00:09<00:00, 13.52it/s] |
| eval model:: 95%|ββββββββββ| 122/128 [00:09<00:00, 13.32it/s] |
| eval model:: 97%|ββββββββββ| 124/128 [00:09<00:00, 13.54it/s] |
| eval model:: 98%|ββββββββββ| 126/128 [00:10<00:00, 12.49it/s] |
| eval model:: 100%|ββββββββββ| 128/128 [00:10<00:00, 12.62it/s] |
| eval model:: 100%|ββββββββββ| 128/128 [00:10<00:00, 12.29it/s] |
| [2026/07/25 18:13:45] ppocr INFO: cur metric, acc: 0.8491210916769505, norm_edit_dis: 0.9879924388315864, fps: 440.39978412555865 |
| [2026/07/25 18:13:45] ppocr INFO: best metric, norm_edit_dis: 0.9880078785722896, is_float16: False, acc: 0.8491210916769505, fps: 443.62973027675116, best_epoch: 1 |
| LAUNCH INFO 2026-07-25 18:13:45,824 Terminating with signal 2 |
| LAUNCH INFO 2026-07-25 18:13:45,918 Force quit in 10 seconds... |
|
|