[Bugfix] [AITER] [ROCm] Fix Quark MoE Quant Config and AITER Fused MoE quant type logic (#27029)
Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com>
This commit is contained in:
@@ -342,7 +342,8 @@ class QuarkW8A8Fp8MoEMethod(QuarkMoEMethod):
|
||||
w2_scale=layer.w2_weight_scale,
|
||||
a1_scale=layer.w13_input_scale,
|
||||
a2_scale=layer.w2_input_scale,
|
||||
per_act_token_quant=self.weight_qscheme == "per_channel",
|
||||
per_act_token_quant=self.input_qscheme == "per_channel",
|
||||
per_out_ch_quant=self.weight_qscheme == "per_channel",
|
||||
)
|
||||
|
||||
def apply(
|
||||
|
||||
Reference in New Issue
Block a user