from threading import Thread
from typing import Iterator

import torch


model_id = 'xuqinyang/baichuan-13b-chat-ggml-int4'

from huggingface_hub import snapshot_download

snapshot_download(model_id, local_dir=".\")
from llama_cpp import Llama
self.llm = Llama(model_path="./ggml-model-q4_0.bin", n_ctx=4096,seed=-1,n_threads=4)

def run(message: str,
        chat_history: list[tuple[str, str]],
        system_prompt: str,
        max_new_tokens: int = 1024,
        temperature: float = 0.8,
        top_p: float = 0.95,
        top_k: int = 50) -> Iterator[str]:
    history = []
    for i in chat_history:
        history.append({"role": "user", "content": i[0]})
        history.append({"role": "assistant", "content": i[1]})
    for response in self.llm.create_chat_completion(messages,stream=True,max_tokens=-1,temperature=0.3,top_k=5,top_p=0.85,repeat_penalty=1.1):
        if "content" in response["choices"][0]["delta"]:
            yield response["choices"][0]["delta"]["content"]