out

app.py

@@ -14,7 +14,13 @@ else:
 
 def generate_response(prompt):
     inputs = tokenizer(prompt, return_tensors="pt")
-    outputs = model.generate(**inputs, max_new_tokens=16, temperature=.7)
+    outputs = model.generate(**inputs, 
+                max_new_tokens=16, 
+                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:]
     new_tokens = tokenizer.decode(new_token_ids, skip_special_tokens=False)