Update main.py

main.py

@@ -145,12 +145,16 @@ def main():
         per_device_train_batch_size=1,
         gradient_accumulation_steps=4,
         learning_rate=5e-6,
+        model_init_kwargs={
+            "torch_dtype": "bfloat16",
+            "attn_implementation": "sdpa",
+        },
 
         # Logging/saving
         logging_steps=1,
         save_steps=12,
         report_to="none",
-
+        
         # Precision
         bf16=True,
         hub_model_id="moos124/gold-code-deepspeed-testV2",
@@ -165,7 +169,6 @@ def main():
         "teacher_model": TEACHER_MODEL,
         "args": config,
         "train_dataset": train_dataset,
-        "attn_implementation": "kernels-community/flash-attn2",
     }
 
     signature = inspect.signature(gold.GOLDTrainer)