Skip to content

Switching from ds to coqui guide #2031

Answered by reuben
betimd asked this question in Q&A
Discussion options

You must be logged in to vote

You should be able to just switch to the new libs, eg import stt instead of import deepspeech, with no changes. For training, some small changes might be needed, as we switched from absl-py to coqpit for configuration and flag handling. The biggest change is that boolean flags must be specified explicitly, for example --train_cudnn true instead of just --train_cudnn. For TFLite, you can simply re-export old checkpoints to get an updated model in the correct format.

Replies: 1 comment

Comment options

You must be logged in to vote
0 replies
Answer selected by reuben
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Category
Q&A
Labels
None yet
2 participants