fix!: speech to text live transcription#816
Conversation
3ca7f15 to
ea943e4
Compare
7b1e6ff to
2ee6d1d
Compare
msluszniak
left a comment
There was a problem hiding this comment.
Some comments are not needed imo
chmjkb
left a comment
There was a problem hiding this comment.
Overall solid work, thanks 👏🏻
Left a couple of nits
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.
816c75a to
ae017ef
Compare
chmjkb
left a comment
There was a problem hiding this comment.
I think you should change the TS side as you're returnign a different thing from C++, for example:
public async encode(waveform: Float32Array): Promise<Float32Array> {
return new Float32Array(await this.nativeModule.encode(waveform));
}
Also, why did you switch back to type in SpeechToTextModelConfig?
a2e04e0 to
081aea0
Compare
|
@IgorSwat could you fix the lint and reference API, then I will run test and speech demo app. |
chmjkb
left a comment
There was a problem hiding this comment.
Please update the docs (not the API reference) as they're outdated now.
|
Hot take: can we just remove the API reference from the repo, please? |
No, why would we do that? |
acd7682 to
04a71b4
Compare
Okay, nevermind, It passed after just 2 rebases. Could be worse :) |
It will always pass after rebase, I can guarantee you. So really it's not that bad, you can handle it once at the very end of the review process and it's just one command, but big benefit of having complete API. First try didn't work because signature of updated package Zod changed, and then you changed code so lines had mismatches ;) |
1e9d17e to
3874dae
Compare
3874dae to
b75f178
Compare
msluszniak
left a comment
There was a problem hiding this comment.
Brilliant work on this one 🚀
chmjkb
left a comment
There was a problem hiding this comment.
One minor comment, overall great work 👏🏻
|
Also please get rid of the API reference files |
Done. |
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