Update app.py

app.py

@@ -15,10 +15,13 @@ model_name = "unsloth/Llama-3.2-1B-Instruct"
 device = torch.device('cuda')
 tokenizer = AutoTokenizer.from_pretrained(model_name)
 
-model = AutoModelForCausalLM.from_pretrained(model_name).to("cuda")
+model1 = AutoModelForCausalLM.from_pretrained(model_name).to("cuda")
 
-provider = TransformersProvider(model, tokenizer, device)
-strategy = CreativeWritingStrategy(provider)
+model2 = AutoModelForCausalLM.from_pretrained(model_name)
+device = torch.device('cuda')
+
+strategy = CreativeWritingStrategy()
+provider = TransformersProvider(model2, tokenizer, device)
 creative_sampler = BacktrackSampler(strategy, provider)
 
 def create_chat_template_messages(history, prompt):
@@ -45,7 +48,7 @@ def generate_responses(prompt, history):
         return tokenizer.decode(generated_list, skip_special_tokens=True)
         
     custom_output = asyncio.run(custom_sampler_task())
-    standard_output = model.generate(inputs, max_length=2048, temperature=1)
+    standard_output = model1.generate(inputs, max_length=2048, temperature=1)
     standard_response = tokenizer.decode(standard_output[0][len(inputs[0]):], skip_special_tokens=True)
 
     return standard_response.strip(), custom_output.strip()