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

[Refactor] Make FP8 Linear Ops use kernel abstraction (#27814)

Browse Source 
Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com>
This commit is contained in:
vllmellm
2026-01-20 14:48:20 +08:00
committed by GitHub
parent e9c83cdc51
commit 148117ea2e
30 changed files with 1467 additions and 1038 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
vllm/model_executor/layers/quantization/utils/quant_utils.py
View File

@@ -123,6 +123,9 @@ kFp8StaticTensorSym = QuantKey(FP8_DTYPE, kStaticTensorScale, symmetric=True)
kDynamicTensorScale = ScaleDesc(torch.float32, False, GroupShape.PER_TENSOR)
kFp8DynamicTensorSym = QuantKey(FP8_DTYPE, kDynamicTensorScale, symmetric=True)
kStaticTokenScale = ScaleDesc(torch.float32, True, GroupShape.PER_TOKEN)
kFp8StaticTokenSym = QuantKey(FP8_DTYPE, kStaticTokenScale, symmetric=True)
kDynamicTokenScale = ScaleDesc(torch.float32, False, GroupShape.PER_TOKEN)
kFp8DynamicTokenSym = QuantKey(FP8_DTYPE, kDynamicTokenScale, symmetric=True)