enhanced_ctc_loss

[Delay0123456789]

你好，在我想用enhanced_ctc_loss时，怎样在ch_PP-OCRv2_rec_enhanced_ctc_loss.yml中设置，其中 Focal-CTC Loss，A-CTC Loss， 怎么设置用哪一个，非常感谢 Loss: name: CombinedLoss loss_config_list:

• CTCLoss: use_focal_loss: false weight: 1.0
• CenterLoss: weight: 0.05 num_classes: 6625 feat_dim: 96 center_file_path:

you can also try to add ace loss on your own dataset

- ACELoss:

weight: 0.1

[an1018]

现在ch_PP-OCRv2_rec_enhanced_ctc_loss.yml使用的就是enhanced_ctc_loss，可以参考enhanced_ctc_loss文档，其中C-CTC是CTC Loss + Center Loss，效果最好，所以采用C-CTC

[Delay0123456789]

非常感谢

[nissansz]

设置成功了吗？ 我用上面的设置，放到resnet34中的loss部分，没有训练进度，读完数据半天后报以下错误？

Traceback (most recent call last): File "./tools/train.py", line 191, in main(config, device, logger, vdl_writer) File "./tools/train.py", line 164, in main program.train(config, train_dataloader, valid_dataloader, device, model, File "C:\F\pycharm2020.2\PaddleOCR-release-2.5\tools\program.py", line 268, in train loss = loss_class(preds, batch) File "C:\Program Files\Python38\lib\site-packages\paddle\fluid\dygraph\layers.py", line 930, in call return self._dygraph_call_func(*inputs, **kwargs) File "C:\Program Files\Python38\lib\site-packages\paddle\fluid\dygraph\layers.py", line 915, in _dygraph_call_func outputs = self.forward(*inputs, **kwargs) File "C:\F\pycharm2020.2\PaddleOCR-release-2.5\ppocr\losses\combined_loss.py", line 55, in forward loss = loss_func(input, batch, **kargs) File "C:\F\pycharm2020.2\PaddleOCR-release-2.5\ppocr\losses\center_loss.py", line 50, in call assert isinstance(predicts, (list, tuple)) AssertionError