Skip to content
This repository was archived by the owner on Oct 13, 2022. It is now read-only.

Conversation

@pkufool
Copy link
Contributor

@pkufool pkufool commented Jul 21, 2021

FYI @glynpu

My current result are as follows (without transformer decoder, say att_rate = 0.0):

decoding model WER % on test-clean WER % on test-other
avg epoch 18 - 28 11.94 26.78
avg epoch 9 - 13 31.65 58.88

@pkufool
Copy link
Contributor Author

pkufool commented Jul 21, 2021

There was a bpe_run.sh in master branch, posted by Liyong in #217, it only did decoding things, so I rename it to bpe_decode.sh. All the training and decoding pipeline are in bpe_run.sh script in this PR.

@pkufool
Copy link
Contributor Author

pkufool commented Jul 25, 2021

We will discuss BPE training with k2 ctc loss in this PR #219

@pkufool pkufool closed this Jul 25, 2021
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant