tidely
tidely
Clicking on the the three dots ... will let you install an older version. It seems the newest versions aren't compatible with your Zed installation.
Additionally the condition for checking the stop criteria is: ```python if len(stop_sequence_buffer) > max_stop_id_sequence_len: ``` It is not inclusive, meaning there will always be an extra token appended. Meaning no...
I'll try to sum it up max_stop_id_sequence_len is the length of the longest stop id sequence. Now let's assume the buffer is the same size as the longest stop sequence....
Yeah, I think it was intended that it runs on every token generated, and we can throw away the buffer entirely. This would address most issues. It still needs a...
Yeah, I'll have a look into writing a unittest
> Confidence values are provided on a per-token basis from whisper.cpp. I hadn't planned on an API to interface on the token level, but it may be worth looking in...
Got unlucky with a flaky test :) I don't think I have permissions to rerun them @mgsloan
This has been sitting around for a while, I feel like this would simplify the installation process. Any updates? @kaixxx
Still a worthwhile change but it's been over a year without comment and the PR is out of date now