deepseekmirror/DeepSeek-V3
1
0
Fork 0
mirror of https://github.com/deepseek-ai/DeepSeek-V3.git synced 2025-05-22 02:06:45 -04:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

Update kernel.py

Browse Source
This commit is contained in:
Arupm 2025-05-07 19:26:51 -04:00 committed by GitHub
parent 4cc6253d5c
commit 37f0b41ac6
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
inference/kernel.py
View File

@ -32,7 +32,7 @@ def act_quant_kernel(x_ptr, y_ptr, s_ptr, BLOCK_SIZE: tl.constexpr):
def act_quant(x: torch.Tensor, block_size: int = 128) -> Tuple[torch.Tensor, torch.Tensor]:
"""
Quantizes the input tensor `x` using block-wise quantization.
Quantizes the input tensor `x` using block-wise quantization as implemented in the given function.
Args:
x (torch.Tensor): The input tensor to be quantized. Must be contiguous and its last dimension size must be divisible by `block_size`.