.

app.py

@@ -84,17 +84,19 @@ if st.button("Generate"):
 
 
     # Generate text
+    chat = [{"role": "user", "content": prompt}]
+    formatted_prompt = tokenizer.apply_chat_template(chat, tokenize=False, add_generation_prompt=True)
 
-    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
+    inputs = tokenizer(formatted_prompt, return_tensors="pt").to(model.device)
     with torch.no_grad():
         outputs = model.generate(   **inputs,
                                     # max_new_tokens=100, 
-                                    max_new_tokens=512, 
+                                    max_new_tokens=256, 
                                     do_sample=True,
                                     temperature=0.6,
                                     top_p=0.95,
                                     top_k=50, 
-                                    num_return_sequences=1, 
+                                    # num_return_sequences=1, 
                                     eos_token_id=tokenizer.eos_token_id
                                     )