Closed
Conversation
Contributor
Author
|
cc @chunfangamd |
Comment on lines
+36
to
+54
| export AMDGCN_USE_BUFFER_OPS=0 | ||
| export VLLM_ROCM_USE_AITER=1 | ||
| export VLLM_ROCM_USE_AITER_UNIFIED_ATTENTION=1 | ||
| export VLLM_ROCM_USE_AITER_MHA=0 | ||
| export VLLM_ROCM_USE_AITER_TRITON_ROPE=1 | ||
| export VLLM_ROCM_QUICK_REDUCE_QUANTIZATION=INT4 | ||
| ATTN_BACKEND="--attention-backend ROCM_AITER_UNIFIED_ATTN" | ||
| FUSE_ROPE_KVCACHE="-cc.pass_config.fuse_rope_kvcache=True -cc.use_inductor_graph_partition=True" | ||
|
|
||
| SERVER_LOG=/workspace/server.log | ||
| PORT=${PORT:-8888} | ||
|
|
||
| set -x | ||
| vllm serve $MODEL --port $PORT \ | ||
| --tensor-parallel-size=$TP \ | ||
| --gpu-memory-utilization 0.95 \ | ||
| --max-model-len $MAX_MODEL_LEN \ | ||
| --compilation-config '{"cudagraph_mode": "FULL_AND_PIECEWISE"}' \ | ||
| --block-size=64 \ | ||
| --no-enable-prefix-caching \ | ||
| --disable-log-requests > $SERVER_LOG 2>&1 & | ||
| $ATTN_BACKEND $FUSE_ROPE_KVCACHE \ | ||
| --tensor-parallel-size=$TP \ | ||
| --gpu-memory-utilization 0.95 \ | ||
| --max-model-len $MAX_MODEL_LEN \ | ||
| --block-size=64 \ | ||
| --no-enable-prefix-caching \ | ||
| --disable-log-requests > $SERVER_LOG 2>&1 & |
Contributor
There was a problem hiding this comment.
thanks for this PR, can u open an PR in vllm recipes repo to update with these new flags. Lets ensure that the documentation is first class such that the entire ml community can benefit from your hard work!
https://github.com/vllm-project/recipes/blob/main/OpenAI/GPT-OSS.md#mi355xgfx950
Collaborator
|
Wow, awesome that y'all have this queued up and ready to go! We will test and ship as soon as vLLM 0.17.0 is out. |
cquil11
approved these changes
Mar 7, 2026
Collaborator
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
3 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.
Opening as draft until vllm v0.17.0 is released. This release will include multiple optimizations for gpt-oss:
Some other updates:
--attention-backend: [Deprecation] Remove deprecated environment variables vllm-project/vllm#32812