Skip to content

GitLab

Commit cfd6a543 authored by SAC_fanth's avatar SAC_fanth
Browse files

修复channel-int8 的config读取bug

parent db85ab07
Show whitespace changes
Inline Side-by-side
Showing with 3 additions and 3 deletions
vllm/model_executor/layers/fused_moe/config.py
View file @ cfd6a543
...@@ -246,8 +246,8 @@ class FusedMoEQuantConfig: ...@@ -246,8 +246,8 @@ class FusedMoEQuantConfig:
@property @property
def block_shape(self) -> list[int] | None: def block_shape(self) -> list[int] | None:
if self.use_int8_w8a8: # if self.use_int8_w8a8:
return [256, 256] # return [256, 256]
if ( if (
self._a1.shape is not None self._a1.shape is not None
...@@ -572,7 +572,7 @@ def int8_w8a8_moe_quant_config( ...@@ -572,7 +572,7 @@ def int8_w8a8_moe_quant_config(
a2_scale=a2_scale, a2_scale=a2_scale,
per_act_token_quant=per_act_token_quant, per_act_token_quant=per_act_token_quant,
per_out_ch_quant=False, per_out_ch_quant=False,
block_shape=[256, 256], block_shape=None,
) )
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment