rebase: align vllm-omni with vLLM ac9126e58aa7

Signed-off-by: tzhouam <tzhouam@connect.ust.hk>
This commit is contained in:
tzhouam committed 2026-10-02 11:06:23 +08:00
1 parent 2721f16ca8
commit 0c84edee02
4 files changed
+55 -7

No files matched your search

+9 -1
View File
@@ -172,7 +172,11 @@ def test_mirror_hardwares_l4_1_expands_to_agents_and_plugins(monkeypatch: pytest
env_names = {item["name"] for item in container["env"]}
assert "VLLM_CI_HF_TOKEN" in env_names
assert "HF_TOKEN" not in env_names
assert step["commands"] == [CUDA_HF_TOKEN_EXPORT, "pytest -sv tests/example"]
assert step["commands"] == [
CUDA_HF_TOKEN_EXPORT,
"python3 .buildkite/cuda/scripts/check_vllm_runtime.py",
"pytest -sv tests/example",
]
def test_mirror_hardwares_l4_preserves_explicit_retry() -> None:
@@ -239,6 +243,10 @@ def test_all_cuda_mirror_hardwares_restore_hf_token_at_runtime() -> None:
assert "VLLM_CI_HF_TOKEN" in env_names
assert "HF_TOKEN" not in env_names
assert step["commands"][0] == CUDA_HF_TOKEN_EXPORT
assert step["commands"][1:] == [
"python3 .buildkite/cuda/scripts/check_vllm_runtime.py",
"pytest -sv tests/example",
]
def _gpu_limit(step: dict) -> int: