Update app.py

app.py

@@ -17,7 +17,7 @@ MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
 
 #Transformers Code
 if torch.cuda.is_available():
-    model_id = "BenBranyon/zephyr-sumbot-all-songs"
+    model_id = "BenBranyon/zephyr-sumbot-all-songs-split"
     model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")
     tokenizer = AutoTokenizer.from_pretrained(model_id)
     tokenizer.use_default_system_prompt = False
@@ -71,7 +71,7 @@ def generate(
         conversation.append({"role": "system", "content": system_prompt})
     for user, assistant in chat_history:
         conversation.extend([{"role": "user", "content": user}, {"role": "assistant", "content": assistant}])
-    conversation.append({"role": "user", "content": "Write a positive rap about " + message})
+    conversation.append({"role": "user", "content": "Generate rap lyrics in the style of the artist Sumkilla about " + message})
 
     input_ids = tokenizer.apply_chat_template(conversation, return_tensors="pt")
     if input_ids.shape[1] > MAX_INPUT_TOKEN_LENGTH: