[CI] Add DGX Spark GPQA smoke test (#39541)

Signed-off-by: mgoin <mgoin64@gmail.com>
Signed-off-by: Michael Goin <mgoin64@gmail.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
Michael Goin
2026-06-22 14:52:38 -07:00
committed by GitHub
co-authored by Claude Opus 4.8
parent c0b2d8f471
commit 6cc2c9ba3a
4 changed files with 24 additions and 2 deletions
+3 -2
View File
@@ -21,12 +21,13 @@ else
exit 0
fi
# build (Grace/GH200 is the arm64 GPU target; sm_90)
# build for arm64 GPU targets: Grace/GH200 (sm_90) and DGX Spark/GB10
# (sm_121, family-covered by 12.0 under CUDA 13)
docker build --file docker/Dockerfile \
--platform linux/arm64 \
--build-arg max_jobs=16 \
--build-arg nvcc_threads=4 \
--build-arg torch_cuda_arch_list="9.0" \
--build-arg torch_cuda_arch_list="9.0 12.0" \
--build-arg USE_SCCACHE=1 \
--build-arg buildkite_commit="$BUILDKITE_COMMIT" \
--tag "$REGISTRY"/"$REPO":"$BUILDKITE_COMMIT"-arm64 \
+14
View File
@@ -220,6 +220,20 @@ steps:
- uv pip install --system 'gpt-oss[eval]==0.0.5'
- pytest -s -v evals/gpt_oss/test_gpqa_correctness.py --config-list-file=configs/models-b200.txt
- label: GPQA Eval (GPT-OSS) (DGX Spark)
key: gpqa-eval-gpt-oss-spark
timeout_in_minutes: 120
device: dgx-spark
optional: true
num_devices: 1
source_file_dependencies:
- csrc/
- vllm/model_executor/layers/quantization
- tests/evals/gpt_oss/
commands:
- uv pip install --system 'gpt-oss[eval]==0.0.5'
- pytest -s -v evals/gpt_oss/test_gpqa_correctness.py --config-list-file=configs/models-spark.txt
- label: MRCR Eval Small Models
device: h200_35gb
timeout_in_minutes: 30
@@ -0,0 +1,5 @@
# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
model_name: "openai/gpt-oss-20b"
metric_threshold: 0.568
reasoning_effort: "low"
@@ -0,0 +1,2 @@
# DGX Spark model configurations for GPQA evaluation
gpt-oss-20b-sm120.yaml