Open
Conversation
msluszniak
reviewed
Feb 17, 2026
.../react-native-executorch/common/rnexecutorch/models/speech_to_text/common/schema/OnlineASR.h
Outdated
Show resolved
Hide resolved
packages/react-native-executorch/common/rnexecutorch/models/speech_to_text/whisper/ASR.h
Outdated
Show resolved
Hide resolved
IgorSwat
force-pushed
the
@is/speech-to-text
branch
from
3ca7f15 to
ea943e4
Compare
msluszniak
reviewed
Feb 18, 2026
packages/react-native-executorch/common/rnexecutorch/models/speech_to_text/SpeechToText.h
Outdated
Show resolved
Hide resolved
IgorSwat
force-pushed
the
@is/speech-to-text
branch
from
7b1e6ff to
2ee6d1d
Compare
msluszniak
reviewed
Mar 2, 2026
Member
msluszniak
left a comment
msluszniak
left a comment
There was a problem hiding this comment.
Some comments are not needed imo
packages/react-native-executorch/common/rnexecutorch/models/speech_to_text/SpeechToText.h
Outdated
Show resolved
Hide resolved
packages/react-native-executorch/common/rnexecutorch/models/speech_to_text/whisper/Utils.h
Outdated
Show resolved
Hide resolved
...ages/react-native-executorch/common/rnexecutorch/models/speech_to_text/whisper/OnlineASR.cpp
Outdated
Show resolved
Hide resolved
...ages/react-native-executorch/common/rnexecutorch/models/speech_to_text/whisper/OnlineASR.cpp
Show resolved
Hide resolved
...act-native-executorch/common/rnexecutorch/models/speech_to_text/whisper/HypothesisBuffer.cpp
Outdated
Show resolved
Hide resolved
msluszniak
reviewed
Mar 3, 2026
chmjkb
requested changes
Mar 4, 2026
Collaborator
chmjkb
left a comment
chmjkb
left a comment
There was a problem hiding this comment.
Overall solid work, thanks 👏🏻
Left a couple of nits
packages/react-native-executorch/common/rnexecutorch/models/speech_to_text/common/schema/ASR.h
Outdated
Show resolved
Hide resolved
...act-native-executorch/common/rnexecutorch/models/speech_to_text/common/types/ProcessResult.h
Show resolved
Hide resolved
...-native-executorch/common/rnexecutorch/models/speech_to_text/common/types/GenerationResult.h
Show resolved
Hide resolved
packages/react-native-executorch/common/rnexecutorch/models/speech_to_text/whisper/Constants.h
Outdated
Show resolved
Hide resolved
packages/react-native-executorch/src/modules/natural_language_processing/SpeechToTextModule.ts
Outdated
Show resolved
Hide resolved
| this->decoder->unload(); | ||
| : callInvoker_(std::move(callInvoker)) { | ||
| // Switch between the ASR implementations based on model name | ||
| if (modelName == "whisper") { |
Collaborator
There was a problem hiding this comment.
food for thought: as we discussed a few days back, think about how we can make it work so that the native side doesn't need the model name, but accepts a bunch of configurable pipeline steps. no need to do this now IMO, but just a note.
Maybe we can have different ASR implementations based on whether the model does support timestamps or not?
| std::shared_ptr<OwningArrayBuffer> | ||
| SpeechToText::encode(std::span<float> waveform) const { | ||
| std::vector<float> encoderOutput = this->asr->encode(waveform); | ||
| std::vector<float> encoderOutput = transcriber_->encode(waveform); |
Collaborator
There was a problem hiding this comment.
I'm thinking whether we need to return std::vector from the encoder? Maybe we would just return a span. We wrap this in OwningArrayBuffer, which copies the data.
chmjkb
requested changes
Mar 5, 2026
Collaborator
chmjkb
left a comment
chmjkb
left a comment
There was a problem hiding this comment.
Two more things:
- I wasn't able to compile the app for Android (due to Norbert bumping minSdkVersion in RNET). You have to bump the minSdkVersion in the example app.
- Once compiled, it doesn't ask for mic permissions (im using a Pixel 10) and silently fails.
IgorSwat
force-pushed
the
@is/speech-to-text
branch
from
ef854bc to
d253381
Compare
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
Description
Various improvements & adjustments in Speech-to-Text module. The list of changes includes:
Introduces a breaking change?
Type of change
Tested on
Testing instructions
You can run the tests defined for Speech-to-Text module, as well as test it manually with the 'speech' demo app (SpeechToText screen).
Screenshots
Related issues
Checklist
Additional notes
IMPORTANT:
This PR is not yet ready to be merged - I am still working on some concrete aspects of the streaming algorithm. However, you are welcome to evaluate and review the architectural design of the code - especially the proposed solution to handle multiple different implementations of STT module.