jonathanjordan21
/

mos-mamba-6x130m-hf

Text Generation

Model card Files Files and versions Community

jonathanjordan21 commited on Jul 16

Commit

9edc8f9

•

1 Parent(s): 2a5c0b5

Update modeling_mos_mamba.py

Files changed (1) hide show

modeling_mos_mamba.py +2 -2

modeling_mos_mamba.py CHANGED Viewed

@@ -703,8 +703,8 @@ class MoSMambaPreTrainedModel(PreTrainedModel):
                 module.dt_proj.bias.copy_(inv_dt)
             module.dt_proj.bias._no_reinit = True
-        if isinstance(module, nn.ModuleList):
-            nn.init.uniform_(module.weight, -0.0001, 0.0001)
         if isinstance(module, nn.Linear):
             if module.bias is not None:

                 module.dt_proj.bias.copy_(inv_dt)
             module.dt_proj.bias._no_reinit = True
+        # if isinstance(module, nn.ModuleList):
+        #     nn.init.uniform_(module.weight, -0.0001, 0.0001)
         if isinstance(module, nn.Linear):
             if module.bias is not None: