[AMD] ENV Flags tuning and cleanup#19176
Conversation
HaiShaw
requested review from
Fridge003,
ishandhanani,
ispobock and
yctseng0211
as code owners
Summary of ChangesHello @HaiShaw, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed! This pull request focuses on refining the environment variable configurations within the ROCm Dockerfiles. It cleans up the build environment by removing outdated Highlights
🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console. Changelog
Activity
Using Gemini Code AssistThe full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips. Invoking Gemini You can request assistance from Gemini at any point by creating a comment using either
Customization To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a Limitations & Feedback Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here. You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension. Footnotes
|
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request cleans up several unused VLLM_* environment variables from the ROCm Dockerfiles and enables ROCM_QUICK_REDUCE_QUANTIZATION=INT8 by default. While the cleanup is good, enabling a quantization feature by default without documentation can be surprising for users. My feedback suggests adding comments to the Dockerfiles to explain this new setting and its potential impact.
| ENV VLLM_FP8_ACT_PADDING=1 | ||
| ENV VLLM_FP8_WEIGHT_PADDING=1 | ||
| ENV VLLM_FP8_REDUCE_CONV=1 | ||
| ENV ROCM_QUICK_REDUCE_QUANTIZATION=INT8 |
There was a problem hiding this comment.
Enabling ROCM_QUICK_REDUCE_QUANTIZATION=INT8 by default could have performance and accuracy implications that might be unexpected for users. It's good practice to document such impactful settings. Please consider adding a comment to explain this environment variable.
ENV ROCM_QUICK_REDUCE_QUANTIZATION=INT8 # Enable INT8 quantization for quick all-reduce on ROCm. Other values: FP, INT6, INT4, NONE.
| ENV VLLM_FP8_ACT_PADDING=1 | ||
| ENV VLLM_FP8_WEIGHT_PADDING=1 | ||
| ENV VLLM_FP8_REDUCE_CONV=1 | ||
| ENV ROCM_QUICK_REDUCE_QUANTIZATION=INT8 |
There was a problem hiding this comment.
Enabling ROCM_QUICK_REDUCE_QUANTIZATION=INT8 by default could have performance and accuracy implications that might be unexpected for users. It's good practice to document such impactful settings. Please consider adding a comment to explain this environment variable.
ENV ROCM_QUICK_REDUCE_QUANTIZATION=INT8 # Enable INT8 quantization for quick all-reduce on ROCm. Other values: FP, INT6, INT4, NONE.
…o xverse_moe * 'xverse_moe' of https://github.com/xiaobaicxy/sglang: (275 commits) fix: add missing blank line after docstring in serving_transcription.py (sgl-project#19206) Whisper model support & `/v1/audio/transcriptions` endpoint & benchmark (sgl-project#16983) fix: patch docker image fixes (sgl-project#19100) [PD-Disagg] Unify prefill info data transition flow, all with `PrefillServerInfo` (sgl-project#19195) [CI] Tiny enhance the dp attention load blance benchmark (sgl-project#19194) add new ci user (sgl-project#19133) [CI] fix the teardown output of disaggregation test (sgl-project#19193) [PD-Disagg] Support query dp rank from bootstrap server. (sgl-project#19168) [Kernel Slimming] Migrate AWQ marlin repack kernel to JIT (sgl-project#18949) [Diffusion] Match rotary_embedding module name style (sgl-project#19179) [Refactor] Split rotary_embedding.py into a modular package (sgl-project#19144) [NPU] bump sgl-kernel-npu to 2026.02.01.post2 (sgl-project#19178) Use single mma warp group for short q_len in FA to optimize decoding performance (sgl-project#18985) Reorganize topk logic to clean up code and expose logical experts (sgl-project#16945) [ROCm] Use unreg path for custom all-reduce during CUDA graph capture (sgl-project#19162) [diffusion] feat: detect Flux2 custom VAE path from component_paths (sgl-project#19170) [AMD] ENV flags tuning and cleanup (sgl-project#19176) Fix bench_one_batch_server by moving the print statements (sgl-project#19175) Update rocm7.2 Dockerfile to install amdsmi for QuickReduce Initialization (sgl-project#19091) Revert "Refactor graph input buffers (sgl-project#18991)" (sgl-project#19173) ...
1 participant
Motivation
Modifications
Accuracy Tests
Benchmarking and Profiling
Checklist
Review Process
/tag-run-ci-label,/rerun-failed-ci,/tag-and-rerun-ci