Jong Wook Kim

Hi, thanks for the PR. I wasn't ready to go fully enterprise open-source with semantic versioning and everything, but date-based versioning seems manageable. The package is now available as [`openai-whisper`...

Thanks, I'll close this for now, since it doesn't quite yet work "out of the box" and relying on nightly versions makes things difficult for me to maintain. I'm hoping...

Thanks, I'll close this for now, since it doesn't quite yet work "out of the box" and relying on nightly versions makes things difficult for me to maintain. I'm hoping...

Thanks for the PR! I've renamed the option to `--output_format` and moved some code around.

I also see the same errors as others mentioned above, on an M1 Mac running arm64 Python.

Thanks for the suggestion; there might be cases where `...` could be useful, e.g. when the speaker is hesitating like "I was... I did something". So I'd keep this enabled...

Hi Sanchit and all, as discussed offline, let me close this in favor of a separate post on the Discussions page.

Thanks for the PR! What I wanted to do instead of this is to mask the timestamp tokens during sampling so that those are conditioned to be monotonically increasing, combined...

Thanks for the comments, all -- this is work in progress and not quite ready for merging. I'm trying to address both hallucination and performance concerns.

Hi @IgnacioSan22, the custom DTW implementation in this PR was for the license issue as noted by others and also for the speed. An alternative is to use the timestamp...