fix(mobile): disable reasoning during voice cleanup - #30
Merged
Conversation
github-actions
Bot
added
size:M
vouch:trusted
Thread transfer impact✅ Thread transfer remains within every enforced ceiling.
Baseline: Scenario and decoded snapshot size10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.
Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed. |
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The shipped 2B cleanup model spent its entire 1,024-token budget on hidden reasoning, leaving no visible rewrite. Use the model template's non-thinking generation prefix so cleanup produces the answer directly. Report unfinished generation before empty output, and log native termination reasons without transcript content.
Verification: reproduced the failure with the user's exact sentence using the pinned llama.cpp revision and shipped 2B weights. With the fix, all three shipped models completed five inputs each on macOS. 75 shared tests, 8 Swift tests, focused lint, and iOS native typechecking pass. Physical iPad verification was attempted but did not complete; the user requested TestFlight delivery to test there.
Model: GPT-6. Harness: Codex.