fix: resolve b200 dsv3 mtp issue (#7286)

This commit is contained in:
Yineng Zhang
2025-06-17 16:22:46 -07:00
committed by GitHub
parent 3eb4a800e8
commit 4f204db57c

View File

@@ -1932,6 +1932,8 @@ class DeepseekV2ForCausalLM(nn.Module):
self._weight_requant_ue8m0()
def _weight_requant_ue8m0(self):
if self.config.architectures[0] == "DeepseekV3ForCausalLMNextN":
return
weight_block_size = self.quant_config.weight_block_size
moe_layers = list(