Manish

Results 6 comments of Manish

implement the ASR to TensorFlow warp binding

@zhoudoufu Have a look at this [issue](https://github.com/HawkAaron/E2E-ASR/issues/14). I think you can help.

A problem of distributed training

@jonathanasdf What is the meaning of worker_gpus and worker_replicas here? Can you share some references to understand this better?

About preprocessing data in ASR

@alessiaatunimi I faced the same issue. It is fixed by using the solution you suggested. Thanks

ASR model with unidirectional lstm and Monotonic attention not converging

@jonathanasdf Your comments can be helpful! Should I decrease the learning rate?

ASR model with unidirectional lstm and Monotonic attention not converging

@jonathanasdf After some more training this unidirectional model: Around 20K steps: wer ~ 0.40 while after noise addition, model seems to be not converging. At 40K steps: wer ~ 0.97...

ASR model with unidirectional lstm and Monotonic attention not converging

@jonathanasdf @drpngx I changed the noise adding steps from 20K to 30K, thinking that the model might be underfit before noise addition, but the model doesn't seems to be converging....