chore(deps): make vLLM an optional extra so base install works on macOS#4
Open
Khurdhula-Harshavardhan wants to merge 3 commits intomainfrom
Open
chore(deps): make vLLM an optional extra so base install works on macOS#4Khurdhula-Harshavardhan wants to merge 3 commits intomainfrom
Khurdhula-Harshavardhan wants to merge 3 commits intomainfrom
Conversation
…rature Opus 4.7 returns a 400 when temperature/top_p/top_k are set; the safest migration per Anthropic's docs is to omit them entirely. Guard on the model id so existing 4.6 / Sonnet 4.6 runs continue to send temperature=0.0 unchanged. Ref: https://platform.claude.com/docs/en/about-claude/models/whats-new-claude-4-7 Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
vllm pulls CUDA-only transitive deps (nvidia-cudnn-frontend etc.) that have no macOS / arm64 wheels, breaking `uv sync` for anyone running hosted-API providers on a Mac. Move vllm into [project.optional-dependencies] so the base resolve is cross-platform; CUDA users opt in with `uv sync --extra vllm`. Verified base `uv sync` succeeds on darwin-arm64 and that `uv pip compile --extra vllm --python-platform x86_64-unknown-linux-gnu` still resolves vllm==0.17.0 with its CUDA deps for Linux users. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
nvidia-cudnn-frontend, etc.) with no macOS / arm64 wheels —uv syncfails immediately on a Mac.vllm==0.17.0into[project.optional-dependencies]so the base resolve is cross-platform. CUDA users opt in withuv sync --extra vllm.Test plan
uv syncsucceeds on darwin-arm64 (was failing onnvidia-cudnn-frontend).uv run python -c "import datasets, anthropic, openai, google.genai, sentence_transformers"works on Mac.uv pip compile pyproject.toml --extra vllm --python-platform x86_64-unknown-linux-gnustill resolvesvllm==0.17.0and its CUDA deps for Linux users.uv sync --extra vllmfollowed by a vLLM-provider inference run.🤖 Generated with Claude Code