Update main.py

main.py

@@ -5,9 +5,7 @@ device = "cuda" # the device to load the model onto
 
 from ctransformers import AutoModelForCausalLM
 
-llm = AutoModelForCausalLM.from_pretrained("TheBloke/Llama-2-7b-Chat-GGUF", model_file="llama-2-7b-chat.q4_K_M.gguf", model_type="llama", gpu_layers=0)
-
-
+llm = AutoModelForCausalLM.from_pretrained("TheBloke/Mistral-7B-v0.1-GGUF", model_file="mistral-7b-v0.1.Q4_K_M.gguf", model_type="mistral", gpu_layers=0)
 
 
 @app.route('/recommend', methods=['POST'])