search

ZhengkunTian / rnn-transducer

A Pytorch Implementation of Transducer Model for End-to-End Speech Recognition
231 stars 57 forks source link

train.py return these errors #11

Closed oshindow closed 4 years ago

oshindow commented 4 years ago

But I can't figure out which one is the 'NoneType' variables.

image

oshindow commented 4 years ago

Traceback (most recent call last): File "/home/wxt/rnn-transducer-master/train.py", line 218, in main() File "/home/wxt/rnn-transducer-master/train.py", line 193, in main for epoch in range(start_epoch, config.training.epochs): TypeError: 'NoneType' object cannot be interpreted as an integer

oshindow commented 4 years ago

hi, the problem is just 'epochs' in code, but 'epoches' in timit.yaml

BuaaAlban commented 4 years ago

hi, the problem is just 'epochs' in code, but 'epoches' in timit.yaml hi, can you run it now? Have you generated the timit feature files using Kaldi example scripts and put them into the author's generated egs/timit/data directory?

oshindow commented 4 years ago

hi, the problem is just 'epochs' in code, but 'epoches' in timit.yaml hi, can you run it now? Have you generated the timit feature files using Kaldi example scripts and put them into the author's generated egs/timit/data directory?

Yes, I run it successfully now. The only problem is the wrong spelling and the feature files are ok.

ZhengkunTian commented 4 years ago

Sorry about that!