chore: dependency updates (flash-attn-3, wheels, lockfile) - #2234
Merged
Conversation
samsja
marked this pull request as ready for review
April 9, 2026 07:09
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes and found 1 potential issue.
❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.
Reviewed by Cursor Bugbot for commit 13dbcf0. Configure here.
| flash-attn-4 = { git = "https://github.com/Dao-AILab/flash-attention.git", subdirectory = "flash_attn/cute", rev = "abd9943b" } | ||
| pydantic-config = { git = "https://github.com/samsja/pydantic_config.git", branch = "main" } | ||
| vllm-router = { url = "https://github.com/PrimeIntellect-ai/router/releases/download/v0.1.14/vllm_router-0.1.14-cp38-abi3-linux_x86_64.whl" } | ||
| vllm-router = { url = "https://github.com/PrimeIntellect-ai/router/releases/download/v0.1.18/vllm_router-0.1.16-cp38-abi3-manylinux_2_28_x86_64.whl" } |
There was a problem hiding this comment.
vllm-router release tag/wheel version mismatch
Medium Severity
The vllm-router URL references GitHub release tag v0.1.18 but the wheel filename embeds version 0.1.16. The previous entry had these in sync (v0.1.14 / 0.1.14). This mismatch suggests either the wrong release tag or the wrong wheel version was specified, which could mean the project is pinned to an older wheel than intended.
Additional Locations (1)
Reviewed by Cursor Bugbot for commit 13dbcf0. Configure here.
JannikSt
approved these changes
Apr 9, 2026
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.
Summary
flash_attn_3via PyPI resolution from the PyTorchtest/cu128index (explicitpytorch-cu128-testindex +[tool.uv.sources]) instead of pinning a single GitHub wheel.research-pre-built-wheelsrelease.uv.lock.Made with Cursor
Note
Medium Risk
Moderate risk because it changes how GPU-critical dependencies are sourced (including using the PyTorch test index), which can impact build/install reliability and runtime compatibility.
Overview
Switches the
flash-attn-3extra from a pinned GitHub wheel to resolvingflash_attn_3via a newly added explicitpytorch-cu128-testindex, and updatesuv.lockaccordingly (including aarch64 support entries).Moves
deep-ep,deep-gemm, andnixl-cu12wheel URLs to theresearch-pre-built-wheelsrelease, and bumpsvllm-routerto the0.1.16manylinux 2.28 wheel. The lockfile is refreshed, including updated Torch wheel download host URLs.Reviewed by Cursor Bugbot for commit 13dbcf0. Bugbot is set up for automated code reviews on this repo. Configure here.