# define inputs
model_path = "mlx-community/Mistral-7B-Instruct-v0.2-4bit"
prompt = prompt_builder("Great content, thank you!")
max_tokens = 140

# load model
model, tokenizer = load(model_path)

# generate response
response = generate(model, tokenizer, prompt=prompt, 
                                      max_tokens = max_tokens, 
                                      verbose=True)