Offline recognizer fixes #80
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
I think that sherpa-onnx doesn't support getting timestamps Looks good! |
Taking an owned Vec is unnecessary since the ownership of the samples array doesn't extend past the function.
vlovich
force-pushed
the
vitali/whisper-recognizer-fixes
branch
from
dcfe136 to
3192f63
Compare
|
Merge conflicts resolved. Please feel free to merge. |
|
The tests failed |
Return timestamps and tokens if the underlying FFI result contains it.
vlovich
force-pushed
the
vitali/whisper-recognizer-fixes
branch
from
3192f63 to
46e0a22
Compare
|
Should be fixed now. |
|
Thanks for contribution! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
transcribeinWhisperRecognizerandMoonshineRecognizertake a&[f32]slice since ownership isn't needed. This is a backwards-incompatible change - not sure if a problem or not. Maybe the method could be made generic to somehow accept either Vec or &[f32] (AsRef?)?One problem I've encountered is I'm not sure how to get the whisper recognizer to actually emit the timestamps to check if they're correct - is something missing in the config / stream setup? I've verified that the tokens do get generated and are converted into a Vec correctly.