attn

app.py

@@ -19,7 +19,6 @@ def generate_response(prompt):
                 do_sample=True,
                 temperature=1,
                 pad_token_id=tokenizer.eos_token_id,
-                attention_mask=torch.ones_like(inputs)
             )
     input_length = inputs['input_ids'].shape[1]
     new_token_ids = outputs[0][input_length:]