Update app.py

app.py

@@ -30,7 +30,7 @@ footer {
 # Initialize the tokenizer
 tokenizer = AutoTokenizer.from_pretrained(model)
 
-llm = LLM(model=model)
+
 
 @spaces.GPU
 def generate(message, history, system, max_tokens, temperature, top_p, top_k, penalty):
@@ -58,6 +58,7 @@ def generate(message, history, system, max_tokens, temperature, top_p, top_k, pe
         eos_token_id=[151645,151643],
     )
     # generate outputs
+    llm = LLM(model=model)
     outputs = llm.generate([text], sampling_params)
     
     # Print the outputs.