perf(asr): cache MLX model across sessions to avoid repeated loads#30
Merged
Conversation
The MLX model was reloaded from disk on every session connect, adding significant latency for Qwen3-ASR models. Track the loaded model path and skip loading when the same path is requested again.
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
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
The MLX model is reloaded from disk on every session's
connect()call. For Qwen3-ASR models (0.6B–1.7B parameters), this adds significant latency that dominates the "press and speak" interaction.Solution
Track the loaded model path in
MLXAsrManager. WhenloadModelis called with the same path that's already loaded, return immediately. The cached path is cleared onunloadModel.Files changed
Packages/KoeMLX/Sources/KoeMLX/MLXAsrManager.swiftTest plan
make buildpasses