Skip to content

Upgrade MLX audio STT models#15

Merged
hehehai merged 3 commits intomainfrom
upgrade-mlx-audio-stt-models
Apr 6, 2026
Merged

Upgrade MLX audio STT models#15
hehehai merged 3 commits intomainfrom
upgrade-mlx-audio-stt-models

Conversation

@hehehai
Copy link
Copy Markdown
Owner

@hehehai hehehai commented Mar 15, 2026

Summary

  • bump mlx-audio-swift to a newer revision with the expanded model support used by this branch
  • add more MLX speech model variants across Qwen3, Voxtral, Parakeet, and Granite, plus a legacy Voxtral repo mapping
  • simplify model description rendering in settings by resolving descriptions from the canonical available model list

Test plan

  • Not run locally

@hehehai hehehai self-assigned this Mar 15, 2026
@hehehai hehehai added the Enhancement New feature or request label Mar 15, 2026
hehehai added 2 commits April 6, 2026 10:31
- constrain model and provider tables to fixed internal scroll areas\n- add MainActor isolation to MLX model manager tests\n\nGenerated with Codex
@hehehai hehehai force-pushed the upgrade-mlx-audio-stt-models branch from ab6b589 to a2bb657 Compare April 6, 2026 05:00
- document the current local MLX Audio STT families in English and Chinese\n- clarify the mirror fork tag policy and current package pin\n\nGenerated with Codex
@hehehai hehehai merged commit db85ca3 into main Apr 6, 2026
1 check passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Enhancement New feature or request

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant