yo

mosaicml · Jul 28, 2024 · 7a66beb · 7a66beb
1 parent 5e07a05
commit 7a66beb
Showing 1 changed file with 1 addition and 0 deletions.
diff --git a/llmfoundry/callbacks/hf_checkpointer.py b/llmfoundry/callbacks/hf_checkpointer.py
@@ -517,6 +517,7 @@ def tensor_hook(
                     new_model_instance.generation_config.update(
                         **original_model.generation_config.to_dict(),
                     )
+                new_model_instance.name_or_path = original_model.name_or_path
 
             # Then load the state dict in with "assign" so that the state dict
             # is loaded properly even though the model is initially on meta device.