Update app.py

app.py

@@ -7,7 +7,7 @@ client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
 tokenizer = AutoTokenizer.from_pretrained("BeastGokul/Bio-Mistral-7B-finetuned")
 base_model = AutoModelForCausalLM.from_pretrained("BioMistral/BioMistral-7B")
 base_model.resize_token_embeddings(len(tokenizer))
-model = PeftModel.from_pretrained(base_model, "BeastGokul/Bio-Mistral-7B-finetuned")
+model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3")
 
 def generate_response(user_query):
     inputs = tokenizer(user_query, return_tensors="pt")