Skip to content

Add MI325X recipe for Laguna S 2.1 - #665

Open
akii96 wants to merge 2 commits into
vllm-project:mainfrom
akii96:add-laguna-s-mi325x
Open

Add MI325X recipe for Laguna S 2.1#665
akii96 wants to merge 2 commits into
vllm-project:mainfrom
akii96:add-laguna-s-mi325x

Conversation

@akii96

@akii96 akii96 commented Jul 22, 2026

Copy link
Copy Markdown

Validation

  • Hardware: 8× AMD MI325X (gfx942)
  • ROCm: 7.2
  • vLLM image: vllm/vllm-openai-rocm:nightly-387189c42997b27e2c04b5d97ef8190ffa2bf909
  • Chat-completion smoke test passed.
  • node scripts/build-recipes-api.mjs passed with 149 models.
@vercel

vercel Bot commented Jul 22, 2026

Copy link
Copy Markdown
Contributor

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated (UTC)
vllm-recipes Ready Ready Preview, Comment Jul 22, 2026 3:29pm

Request Review

@akii96
akii96 force-pushed the add-laguna-s-mi325x branch from 28f930e to 21dee91 Compare July 22, 2026 13:28

@gemini-code-assist gemini-code-assist Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Review

This pull request adds support and documentation for running the Laguna-S-2.1 model on AMD ROCm hardware (specifically 8× MI325X GPUs in BF16 precision), including hardware overrides and a launch guide. The review feedback points out that the new AMD ROCm launch command is missing the --max-model-len 262144 parameter, which is inconsistent with other multi-GPU commands and could lead to out-of-memory errors.

Comment thread models/poolside/Laguna-S-2.1.yaml
Signed-off-by: Aakif Nawaz <aakif.nawaz@amd.com>
Signed-off-by: Aakif Nawaz <aakif.nawaz@amd.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

1 participant