[Feature] Support XiaoMi MIMO Flash V2 (#62)
* [Feature] Support MIMO Flash V2
This commit is contained in:
@@ -57,6 +57,8 @@ def vllm_kunlun_forward_cuda(
|
||||
)
|
||||
return out
|
||||
|
||||
RMSNorm.forward_cuda = vllm_kunlun_forward_cuda
|
||||
RMSNorm.forward = vllm_kunlun_forward_cuda
|
||||
|
||||
class KunlunGemmaRMSNorm(OriGemmaRMSNorm):
|
||||
@staticmethod
|
||||
|
||||
Reference in New Issue
Block a user