Skip to content

Fixing Chunked Prefill Test. #19762

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Jun 19, 2025
Merged

Fixing Chunked Prefill Test. #19762

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
9ac8c15
Fixing Chunked Prefill Test.
Alexei-V-Ivanov-AMD Jun 17, 2025
d48a896
Enabling Chunked Prefill Test in production.
Alexei-V-Ivanov-AMD Jun 17, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion .buildkite/test-pipeline.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -89,7 +89,7 @@ steps:
- VLLM_TEST_ENABLE_ARTIFICIAL_PREEMPT=1 pytest -v -s basic_correctness/test_preemption.py

- label: Chunked Prefill Test
mirror_hardwares: [amdexperimental]
mirror_hardwares: [amdexperimental, amdproduction]
source_file_dependencies:
- vllm/
- tests/basic_correctness/test_chunked_prefill
Expand Down
18 changes: 16 additions & 2 deletions tests/basic_correctness/test_chunked_prefill.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -49,7 +49,13 @@ def use_v0_only(monkeypatch: pytest.MonkeyPatch):
# NOTE: Increasing this in this suite will fail CI because we currently cannot
# reset distributed env properly. Use a value > 1 just when you test.
@pytest.mark.parametrize("tensor_parallel_size", [1])
@pytest.mark.parametrize("attention_backend", ["FLASHINFER", "FLASH_ATTN"])
@pytest.mark.parametrize("attention_backend", [
pytest.param("FLASHINFER",
marks=pytest.mark.skipif(
current_platform.is_rocm(),
reason="FLASHINFER isn't supported on ROCm")),
"FLASH_ATTN"
])
def test_models(
hf_runner: HfRunner,
vllm_runner: VllmRunner,
Expand Down Expand Up @@ -99,7 +105,13 @@ def test_models(
@multi_gpu_test(num_gpus=2)
@pytest.mark.parametrize("distributed_executor_backend", ["ray", "mp"])
@pytest.mark.parametrize("model", MODELS)
@pytest.mark.parametrize("attention_backend", ["FLASHINFER", "FLASH_ATTN"])
@pytest.mark.parametrize("attention_backend", [
pytest.param("FLASHINFER",
marks=pytest.mark.skipif(
current_platform.is_rocm(),
reason="FLASHINFER isn't supported on ROCm")),
"FLASH_ATTN"
])
def test_models_distributed(
hf_runner: HfRunner,
vllm_runner: VllmRunner,
Expand Down Expand Up @@ -172,6 +184,8 @@ def test_models_distributed(
# Due to low-precision numerical divergence, this test is too sensitive to
# the async postprocessor
@pytest.mark.parametrize("disable_async_output_proc", [True])
@pytest.mark.skipif(current_platform.is_rocm(),
reason="machete_prepack_B isn't supported on ROCm")
def test_models_with_fp8_kv_cache(
vllm_runner: VllmRunner,
example_prompts,
Expand Down