Update app.py

app.py

@@ -23,16 +23,23 @@ tokenizer, model = load_model()
 
 @spaces.GPU
 def generate(prompt, history):
-    conversation = [
+    
+    if len(history) > 0:
+        messages = history + [
         {"role": "user", "content": prompt},
     ]
+    else:
+        messages = [
+            {"role": "user", "content": prompt},
+        ]
 
+    print(history)
     inputs = tokenizer.apply_chat_template(
-        conversation,
+        messages,
         add_generation_prompt=True,
         tokenize=True,
         return_tensors="pt",
-        return_dict=True
+        return_dict=True,
     ).to(device)
     
     with torch.no_grad():
@@ -42,7 +49,9 @@ def generate(prompt, history):
         )
     
     generated_tokens = outputs[0, inputs.input_ids.shape[1]:]
-    return tokenizer.decode(generated_tokens, skip_special_tokens=True)
+    output = tokenizer.decode(generated_tokens, skip_special_tokens=True)
+
+    return output
 
 
 demo = gr.ChatInterface(fn=generate, type="messages", examples=["hello", "hola", "merhaba"], title="NanoChat")