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

Convert formatting to use ruff instead of yapf + isort (#26247)

Browse Source 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-10-05 15:06:22 +01:00
committed by GitHub
parent 17edd8a807
commit d6953beb91
1508 changed files with 115244 additions and 94146 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
tests/kernels/quantization/test_per_token_group_quant.py
View File

@@ -13,15 +13,15 @@ from vllm.model_executor.layers.quantization.utils import fp8_utils, int8_utils
@pytest.mark.parametrize("scale_ue8m0", [False, True])
@pytest.mark.parametrize("group_size", [64, 128])
@pytest.mark.skipif(not torch.cuda.is_available(), reason="CUDA not available")
def test_per_token_group_quant_fp8(shape, column_major: bool,
scale_ue8m0: bool, group_size: int):
def test_per_token_group_quant_fp8(
shape, column_major: bool, scale_ue8m0: bool, group_size: int
):
device = "cuda"
torch.manual_seed(42)
num_tokens, hidden_dim = shape
x = (torch.randn(
(num_tokens, hidden_dim), device=device, dtype=torch.bfloat16) * 8)
x = torch.randn((num_tokens, hidden_dim), device=device, dtype=torch.bfloat16) * 8
# cuda path
out_q, scale = fp8_utils.per_token_group_quant_fp8(
@@ -53,8 +53,7 @@ def test_per_token_group_quant_int8(shape, group_size: int):
torch.manual_seed(42)
num_tokens, hidden_dim = shape
x = (torch.randn(
(num_tokens, hidden_dim), device=device, dtype=torch.bfloat16) * 8)
x = torch.randn((num_tokens, hidden_dim), device=device, dtype=torch.bfloat16) * 8
# cuda path
out_q, scale = int8_utils.per_token_group_quant_int8(