if teacher_force is false, raw_pred should be topk(1)?

Alexander-H-Liu / End-to-end-ASR-Pytorch

This is an open source project (formerly named Listen, Attend and Spell - PyTorch Implementation) for end-to-end ASR implemented with Pytorch, the well known deep learning toolkit.

MIT License

1.18k stars 317 forks source link

if teacher_force is false, raw_pred should be topk(1)? #4

Closed lezasantaizi closed 6 years ago

lezasantaizi commented 6 years ago

Alexander-H-Liu commented 6 years ago

Thanks for your suggestion, I've added a parameter decode_mode in config. Output format (distribution/ topk(1)/ sampled onehot) of Speller feeding foward to next time step can be selelcted by this flag.

ghost commented 5 years ago

@lezasantaizi Hello, did you get the project working for limit and librispeech? Could you take a look at https://github.com/Alexander-H-Liu/Listen-Attend-and-Spell-Pytorch/issues/12. thank you!