[ROCm]Enable AITER MoE backend for MiniMax-M3-MXFP4 (#46419)
Signed-off-by: Qiang Li <qiang.li2@amd.com> Co-authored-by: TJian <tunjian.tan@embeddedllm.com>
This commit is contained in:
@@ -309,6 +309,7 @@ class MiniMaxM3MoE(nn.Module):
|
||||
top_k=config.num_experts_per_tok,
|
||||
hidden_size=config.hidden_size,
|
||||
intermediate_size=config.intermediate_size,
|
||||
intermediate_pad=0,
|
||||
scoring_func=config.scoring_func,
|
||||
e_score_correction_bias=self.e_score_correction_bias,
|
||||
renormalize=True,
|
||||
|
||||
Reference in New Issue
Block a user