chore: bump sgl and update dp attn yaml

[ishandhanani]

Have been benchmarking with this commit

Summary by CodeRabbit

• Chores
  • Updated the Dockerfile to use a newer version of the sglang package for improved performance.
• Configuration
  • Changed the default served model in the configuration to "silence09/DeepSeek-R1-Small-2layers".

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[coderabbitai]

Walkthrough

The updates include changing the sglang package commit hash in the Dockerfile to a newer version for improved performance, and updating a model name in a YAML configuration file to reference a different served model. No changes were made to public interfaces or exported entities.

Changes

File(s)	Change Summary
container/Dockerfile.sglang	Updated sglang package install to use a newer commit hash.
examples/sglang/configs/disagg-dp-attention.yaml	Changed the served model name in the configuration file.

Possibly related PRs

• fix: remove sglang hash for pyproject #1281: Updates sglang version references, similar to this PR's Dockerfile change, but in pyproject.toml instead.

Poem

A Dockerfile hops to a newer hash,
Performance tweaks—oh, what a splash!
The YAML points to a model new,
Small and swift, just for you.
With every commit and config delight,
Rabbits code onward, day and night! 🐇✨

---

📜 Recent review details

Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 43991e7 and e347569.

📒 Files selected for processing (2)

• container/Dockerfile.sglang (1 hunks)
• examples/sglang/configs/disagg-dp-attention.yaml (1 hunks)

⏰ Context from checks skipped due to timeout of 90000ms (1)

• GitHub Check: Build and Test - vllm

🔇 Additional comments (2)

examples/sglang/configs/disagg-dp-attention.yaml (1)

17-17:

✅ Verification successful

Verify served model identifier consistency and availability.

All references to the served model have been updated to silence09/DeepSeek-R1-Small-2layers. Please ensure that:

1. This exact identifier matches the published model repository.
2. Downstream deployment manifests and frontend code reference the same string (including underscores vs. dashes).

Also applies to: 22-23, 38-39

---

🏁 Script executed:

#!/bin/bash
# Check that the model repository exists on Hugging Face
curl -s -o /dev/null -w "%{http_code}" https://huggingface.co/silence09/DeepSeek-R1-Small-2layers

Length of output: 103

---

🏁 Script executed:

#!/bin/bash
# Search for all occurrences of the model identifier to verify consistency
rg -n "DeepSeek" .

Length of output: 14461

---

Served model identifier verified and consistent

• The Hugging Face URL for silence09/DeepSeek-R1-Small-2layers returns HTTP 200.
• All 5 occurrences in examples/sglang/configs/disagg-dp-attention.yaml (lines 17, 22–23, 38–39) use the exact same string.
• No other files reference a different identifier or require updating.

container/Dockerfile.sglang (1)

139-141:

❓ Verification inconclusive

Confirm new SGLANG_COMMIT hash.

The SGLANG_COMMIT has been bumped to 6376b632eb4daef306b89ede0eabdcb89ddff728 (PR #6780). Verify that this commit exists upstream and includes the intended DP attention & NIXL performance fixes.

---

🏁 Script executed:

#!/bin/bash
# Verify that the specified commit exists in the sglang repository
git ls-remote https://github.com/sgl-project/sglang.git | grep 6376b632eb4daef306b89ede0eabdcb89ddff728

Length of output: 105

---

❗ Commit SHA Not Found Upstream

The specified SGLANG_COMMIT (6376b632eb4daef306b89ede0eabdcb89ddff728) was not found among the default heads or tags in the sgl-project/sglang repository. Please:

• Confirm that this SHA is correct and has been merged into the upstream repo (or is exposed via a pull-request ref such as refs/pull/6780/head).
• Verify that it contains the intended DP attention and NIXL performance fixes referenced in PR #6780.
• If the commit lives on a fork or feature branch, either merge it upstream or update this Dockerfile to point at a ref that the build can fetch.

---

🪧 Tips

Chat

There are 3 ways to chat with CodeRabbit:

• Review comments: Directly reply to a review comment made by CodeRabbit. Example:
  • I pushed a fix in commit <commit_id>, please review it.
  • Explain this complex logic.
  • Open a follow-up GitHub issue for this discussion.
• Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
  • @coderabbitai explain this code block.
  • @coderabbitai modularize this function.
• PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
  • @coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.
  • @coderabbitai read src/utils.ts and explain its main purpose.
  • @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.
  • @coderabbitai help me debug CodeRabbit configuration file.

Support

Need help? Create a ticket on our support page for assistance with any issues or questions.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (Invoked using PR comments)

• @coderabbitai pause to pause the reviews on a PR.
• @coderabbitai resume to resume the paused reviews.
• @coderabbitai review to trigger an incremental review. This is useful when automatic reviews are disabled for the repository.
• @coderabbitai full review to do a full review from scratch and review all the files again.
• @coderabbitai summary to regenerate the summary of the PR.
• @coderabbitai generate sequence diagram to generate a sequence diagram of the changes in this PR.
• @coderabbitai resolve resolve all the CodeRabbit review comments.
• @coderabbitai configuration to show the current CodeRabbit configuration for the repository.
• @coderabbitai help to get help.

Other keywords and placeholders

• Add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.
• Add @coderabbitai summary to generate the high-level summary at a specific location in the PR description.
• Add @coderabbitai anywhere in the PR title to generate the title automatically.

Documentation and Community

• Visit our Documentation for detailed information on how to use CodeRabbit.
• Join our Discord Community to get help, request features, and share feedback.
• Follow us on X/Twitter for updates and announcements.