#%%



from langchain_core.prompts import PromptTemplate



# ctransformers is no longer used
""" from langchain_community.llms import CTransformers

# Initialize the language model
llm = CTransformers(model='Model/llama-2-7b-chat.ggmlv3.q2_K.bin',  # 2 bit quantized model
                    model_type='llama',
                    config={'max_new_tokens': 256,  # max tokens in reply
                            'temperature': 0.01, }  # randomness of the reply
                    )
 """

# Initialize the language model with LlamaCpp
#llm = LlamaCpp(model_path="Model/llama-2-7b-chat.Q4_K_M.gguf",  #  token streaming to terminal
               #device="cuda",n_gpu_layers=-1,verbose = True, max_tokens = 4096,  #offloads ALL layers to GPU, uses around 6 GB of Vram
               #config={  # max tokens in reply
                #       'temperature': 0.75}  # randomness of the reply
               #)





template = """
You are an ai that answers medical related queries in a helpful,clear and concise manner.
Avoid giving misinformation, and provide details about your results. Dont answer any questions outside medical domain.

Question: {query}

Answer:

"""

prompt_template = PromptTemplate(
    input_variables=["query"],
    template=template

)

print(
    prompt_template.format(
        query="how can I change a tyre on my car"
    )
)

# %%