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

[Refactor] Remove Duplicate per_block_cast_to_fp8, Remove Dependencies of DeepGEMM (#21787)

Browse Source 
Signed-off-by: yewentao256 <zhyanwentao@126.com>
This commit is contained in:
Wentao Ye
2025-07-31 21:13:27 -04:00
committed by GitHub
parent 0bd409cf01
commit 3700642013
8 changed files with 55 additions and 132 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
tests/kernels/quantization/test_block_fp8.py
View File

@@ -117,7 +117,7 @@ def test_w8a8_block_fp8_deep_gemm_matmul(M, N, K, block_size, out_dtype, seed):
B_fp32 = (torch.rand(N, K, dtype=torch.float32) - 0.5) * 2 * fp8_max
A_fp8, As_fp8 = per_token_group_quant_fp8(A_fp32, block_size[1])
B_fp8, Bs_fp8 = per_block_cast_to_fp8(B_fp32)
B_fp8, Bs_fp8 = per_block_cast_to_fp8(B_fp32, block_size=block_size)
As = As_fp8.to(torch.float32)
Bs = Bs_fp8.to(torch.float32)