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

[Docs] Replace rst style double-backtick with md single-backtick (#27091)

Browse Source 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-10-17 10:47:34 +01:00
committed by GitHub
parent 483ea64611
commit 6c9fdbf725
31 changed files with 98 additions and 98 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
vllm/model_executor/layers/quantization/utils/fp8_utils.py
View File

@@ -887,11 +887,11 @@ def requant_weight_ue8m0_inplace(
UE8M0 (power-of-two) format expected by the new DeepGEMM kernels inplace.
Args:
weight: Block-quantised weight tensor stored in ``torch.float8_e4m3fn``.
Expected shape ``(..., M, K)``.
weight_scale: Corresponding per-block scale tensor (``torch.float32``)
with shape ``(..., M // block_size[0], K // block_size[1])``.
block_size: 2-element iterable ``[block_m, block_k]`` describing the
weight: Block-quantised weight tensor stored in `torch.float8_e4m3fn`.
Expected shape `(..., M, K)`.
weight_scale: Corresponding per-block scale tensor (`torch.float32`)
with shape `(..., M // block_size[0], K // block_size[1])`.
block_size: 2-element iterable `[block_m, block_k]` describing the
block quantisation granularity.
"""
if weight.numel() == 0: