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 15, 23:30 UTC
[V1] LM Eval With Streaming Integration Tests (#11590)
Robert Shaw committed
Feb 7, 2025 at 18:07 UTC
932c6b74616d25199e87c96707e8cfea3ab045c0
1 file changed
+3
.buildkite/test-pipeline.yaml
+3
@@ -195,6 +195,9 @@ steps:
195
# TODO: accuracy does not match, whether setting
196
# VLLM_USE_FLASHINFER_SAMPLER or not on H100.
197
- VLLM_USE_V1=1 pytest -v -s v1/e2e
198
+ # Integration test for streaming correctness (requires special branch).
199
+ - pip install -U git+https://github.com/robertgshaw2-neuralmagic/lm-evaluation-harness.git@streaming-api
200
+ - pytest -v -s entrypoints/openai/test_accuracy.py::test_lm_eval_accuracy_v1_engine
201
202
- label: Examples Test # 25min
203
working_dir: "/vllm-workspace/examples"