biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[CI] Bump mypy version (#34950)

Browse Source 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2026-03-04 20:55:11 +00:00
committed by GitHub
parent 7eca859110
commit 17dc9c7fc9
13 changed files with 90 additions and 61 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
tests/kernels/quantization/test_fp8_quant.py
View File

@@ -57,11 +57,11 @@ def opcheck_fp8_quant(
@pytest.mark.parametrize("num_tokens", NUM_TOKENS)
@pytest.mark.parametrize("hidden_size", HIDDEN_SIZES)
@pytest.mark.parametrize("dtype", DTYPES)
@pytest.mark.parametrize("scale_ub", SCALE_UBS)
@pytest.mark.parametrize("do_scale_ub", SCALE_UBS)
@pytest.mark.parametrize("seed", SEEDS)
@torch.inference_mode()
def test_dynamic_per_token_fp8_quant(
num_tokens: int, hidden_size: int, dtype: torch.dtype, scale_ub: bool, seed: int
num_tokens: int, hidden_size: int, dtype: torch.dtype, do_scale_ub: bool, seed: int
) -> None:
set_random_seed(seed)
@@ -70,7 +70,7 @@ def test_dynamic_per_token_fp8_quant(
) # avoid nans
scale_ub = (
torch.mean(x).to(dtype=torch.float32, device="cuda") if scale_ub else None
torch.mean(x).to(dtype=torch.float32, device="cuda") if do_scale_ub else None
)
ref_out, ref_scales = ref_dynamic_per_token_quant(x, FP8_DTYPE, scale_ub)
ops_out, ops_scales = ops.scaled_fp8_quant(