Yuecong Min

welcome follow my fork, I will keep updated~

I haven't met this problem before. It seems like it is about switches between native and cudnn in earlier discussion about pytorch [discussion](https://discuss.pytorch.org/t/bug-ctc-loss-makes-me-confuse/51038) The recent version of pytorch adopt two...

1.10.0 and 1.10.1 works for me.

The version has been presented in Readme, it seems like you did not install ctcdecoder successfully.

@wljcode Thanks for your attention to our work. It seems like you used batch size=1, which may affect the robustness of the model. Besides, the learning rate does not decay...

@wljcode Have you successfully reimplemented the experimental results? I checked the relevant logs and found that you adopt the load_weights to continue training rather than load_checkpoints, the former only load...

Hi, @sunke123, thanks for your attention to our work. We will explain this performance gap in our next update, perhaps in two weeks, which can achieve better performance (about 20%...

Hi, @sunke123, the code has been updated~

@herochen7372 You can first check whether the evaluation script runs as expected with the provided pretrained model, and than check whether the loss decreases as the iteration progresses.

Similar to issue #17, you can post issues if you meet any problems during implementation.