Update app.py

app.py

@@ -27,15 +27,17 @@ def generate_code(prompt):
         add_generation_prompt=True
     )
     model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
-    with torch.no_grad():
-        generated_ids = model.generate(
-            **model_inputs,
-            max_new_tokens=1024
-        )
-    generated_ids = [
-        output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
-    ]
-    response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
+    #with torch.no_grad():
+    generated_ids = model.generate(
+        **model_inputs,
+        max_length = 1024,
+        min_new_tokens = 256,
+        #low_memory = True,
+        token_healing = True,
+        guidance_scale = 3.8,
+    )
+    generated_ids = [output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)]
+        response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
     return response
 
 with gr.Blocks(title="Qwen 14b") as demo:  # Updated title