Fix GLM-5 fused shared expert (#18804)
Co-authored-by: FrankMinions <liuchen@shinemo.com> Co-authored-by: Brayden Zhong <b8zhong@uwaterloo.ca> Co-authored-by: Kangyan-Zhou <zky314343421@gmail.com>
This commit is contained in:
co-authored by
FrankMinions
Brayden Zhong
Kangyan-Zhou
parent
0ff24159a5
commit
1b659bcb08
@@ -1281,7 +1281,8 @@ class Glm4MoeForCausalLM(nn.Module):
|
|||||||
|
|
||||||
|
|
||||||
class GlmMoeDsaForCausalLM(DeepseekV2ForCausalLM):
|
class GlmMoeDsaForCausalLM(DeepseekV2ForCausalLM):
|
||||||
pass
|
def determine_num_fused_shared_experts(self):
|
||||||
|
super().determine_num_fused_shared_experts("GlmMoeDsaForCausalLM")
|
||||||
|
|
||||||
|
|
||||||
EntryClass = [Glm4MoeForCausalLM, GlmMoeDsaForCausalLM]
|
EntryClass = [Glm4MoeForCausalLM, GlmMoeDsaForCausalLM]
|
||||||
|
|||||||
Reference in New Issue
Block a user