Update app.py

app.py

@@ -12,10 +12,10 @@ model = AutoModelForCausalLM.from_pretrained(
 )
 tokenizer = AutoTokenizer.from_pretrained(model_name)
 
-@spaces.GPU
+@spaces.GPU(duration=100)
 def generate(prompt, history):
     messages = [
-        {"role": "system", "content": "Je bent een vriendelijke, behulpzame assistent."},
+        {"role": "system", "content": "あなたはフレンドリーなチャットボットです。"},
         {"role": "user", "content": prompt}
     ]
     text = tokenizer.apply_chat_template(
@@ -27,7 +27,7 @@ def generate(prompt, history):
     
     generated_ids = model.generate(
         **model_inputs,
-        max_new_tokens=512
+        max_new_tokens=864
     )
     generated_ids = [
         output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)