biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[ROCm][CI] Add LM Eval Qwen3.5 Models test for MI355 (#38155)

Browse Source 
Signed-off-by: Andreas Karatzas <akaratza@amd.com>
This commit is contained in:
Andreas Karatzas
2026-03-26 11:51:18 -05:00
committed by GitHub
parent a8e48a7b85
commit 9c3ae04bfe
2 changed files with 25 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
.buildkite/test-amd.yaml
View File

@@ -3749,3 +3749,27 @@ steps:
- vllm/platforms/rocm.py
commands:
- python3 benchmarks/attention_benchmarks/benchmark.py --backends ROCM_ATTN ROCM_AITER_FA ROCM_AITER_UNIFIED_ATTN --batch-specs "8q1s1k" --repeats 1 --warmup-iters 1
- label: LM Eval Qwen3-5 Models (B200-MI355) # TBD
timeout_in_minutes: 120
mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355]
agent_pool: mi355_2
num_gpus: 2
optional: true
working_dir: "/vllm-workspace/tests"
source_file_dependencies:
- vllm/model_executor/models/qwen3_5.py
- vllm/model_executor/models/qwen3_5_mtp.py
- vllm/transformers_utils/configs/qwen3_5.py
- vllm/transformers_utils/configs/qwen3_5_moe.py
- vllm/model_executor/models/qwen.py
- vllm/model_executor/models/qwen2.py
- vllm/model_executor/models/qwen3.py
- vllm/model_executor/models/qwen3_next.py
- vllm/model_executor/models/qwen3_next_mtp.py
- vllm/model_executor/layers/fla/ops/
- vllm/_aiter_ops.py
- vllm/platforms/rocm.py
commands:
- pytest -s -v evals/gsm8k/test_gsm8k_correctness.py --config-list-file=configs/models-qwen35-mi355.txt

1
tests/evals/gsm8k/configs/models-qwen35-mi355.txt Normal file
View File

@@ -0,0 +1 @@
Qwen3.5-35B-A3B-DEP2.yaml