kuleshov-group
/

bd3lm-owt-block_size8

Text Generation

language-modeling

Model card Files Files and versions

marriola commited on Mar 11

Commit

abd95cf

·

verified ·

1 Parent(s): e0c72eb

Upload BD3LM

Files changed (1) hide show

modeling_bd3lm.py +1 -1

modeling_bd3lm.py CHANGED Viewed

@@ -565,7 +565,7 @@ class BD3LM(transformers.PreTrainedModel):
   def reset_kv_cache(self):
     for block in self.backbone.blocks:
-      block.kv_cache = None
   def forward(
       self,

   def reset_kv_cache(self):
     for block in self.backbone.blocks:
+      block.kv_cache = None
   def forward(
       self,