Repository navigation
qwen3_5: the --drafter openers move to their own file, so cuda_native.zig has room again - #1797
Merged
Merged
Conversation
…_native.zig has room again cuda_native.zig sat at exactly 600 lines after the shared-round pricing landed, so any change to it failed the 600-line rule (#1750). The DFlash2 and MTP openers, their reserve/close callbacks and the DFlash2 calibration move unchanged; Opened and openDrafter are re-exported, so nothing else changes.
This was referenced Nov 1, 2026
Closed
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
zig/src/families/qwen3_5/cuda_native.zighas sat at exactly 600 lines since the shared-round pricing landed, so any contributor change to it (#1750, #1678) fails the 600-line rule. This moves the DFlash2 and MTP drafter openers unchanged tocuda_drafter_open.zig:Opened,openDrafter,openDFlash,openMtp, their reserve and close callbacks, and the DFlash2 calibration.OpenedandopenDrafterare re-exported, andLanes,cacheBytesandtimeWindowsbecomepubfor the new file. The file is now 522 lines. CUDA host tests and the Linux CUDA cross-compile pass, and lean adds nothing.🤖 Generated with Claude Code