我设置了batch_size_per_card: 16运行后报：Traceback (most recent call last): File "E:/PaddleOCR-release-2.7.1/tools/train.py", line 228, in <module> main(config, device, logger, vdl_writer, seed) File "E:/PaddleOCR-release-2.7.1/tools/train.py", line 203, in main amp_custom_white_list, amp_dtype) File "E:\PaddleOCR-release-2.7.1\tools\program.py", line 337, in train post_result = post_process_class(preds, batch[1]) File "E:\PaddleOCR-release-2.7.1\ppocr\postprocess\db_postprocess.py", line 229, in __call__ src_h, src_w, ratio_h, ratio_w = shape_list[batch_index] ValueError: too many values to unpack (expected 4)

[694578908]

Traceback (most recent call last): File "E:/PaddleOCR-release-2.7.1/tools/train.py", line 228, in main(config, device, logger, vdl_writer, seed) File "E:/PaddleOCR-release-2.7.1/tools/train.py", line 203, in main amp_custom_white_list, amp_dtype) File "E:\PaddleOCR-release-2.7.1\tools\program.py", line 337, in train post_result = post_process_class(preds, batch[1]) File "E:\PaddleOCR-release-2.7.1\ppocr\postprocess\db_postprocess.py", line 229, in call src_h, src_w, ratio_h, ratio_w = shape_list[batch_index] ValueError: too many values to unpack (expected 4)

[GreatV]

更新到最新版本试试

[694578908]

已解决了，把cal_metric_during_train: False就可以了