Skip to content

Commit f83d4df

Browse files
Doug Lehrdllehr-amd
authored andcommitted
Fix Qwen accuracy fix by not sending quant_config to MOE self.gate RLU
1 parent 176244a commit f83d4df

File tree

1 file changed

+1
-1
lines changed

1 file changed

+1
-1
lines changed

vllm/model_executor/models/qwen3_moe.py

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -149,7 +149,7 @@ def __init__(
149149
self.gate = ReplicatedLinear(config.hidden_size,
150150
config.num_experts,
151151
bias=False,
152-
quant_config=quant_config,
152+
quant_config=None,
153153
prefix=f"{prefix}.gate")
154154

155155
def forward(self, hidden_states: torch.Tensor) -> torch.Tensor:

0 commit comments

Comments
 (0)