A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 8, 01:15 UTC
[CI] Add GPT-OSS Eval job for H100 (#34359)
Signed-off-by: Michael Goin <mgoin64@gmail.com>
Michael Goin committed
Feb 19, 2026 at 20:14 UTC
16f24e87975ef4cd2c12879425062913ef62f6fd
1 file changed
+13
.buildkite/test_areas/misc.yaml
+13
@@ -147,6 +147,19 @@ steps:
147
- pytest -v -s transformers_utils
148
- pytest -v -s config
149
150
+- label: GPT-OSS Eval (H100)
151
+ timeout_in_minutes: 60
152
+ working_dir: "/vllm-workspace/"
153
+ device: h100
154
+ optional: true
155
+ source_file_dependencies:
156
+ - tests/evals/gpt_oss
157
+ - vllm/model_executor/models/gpt_oss.py
158
+ - vllm/model_executor/layers/quantization/mxfp4.py
159
+ commands:
160
+ - uv pip install --system 'gpt-oss[eval]==0.0.5'
161
+ - pytest -s -v tests/evals/gpt_oss/test_gpqa_correctness.py --model openai/gpt-oss-20b --metric 0.58
162
+
163
- label: GPT-OSS Eval (B200)
164
timeout_in_minutes: 60
165
working_dir: "/vllm-workspace/"