import os
import platform
import signal
import torch
from transformers import AutoConfig, AutoModel, AutoTokenizer
import readline

modelPath = "chatglm-6b"
# 可以正常输出，但是回答问题不对应
# checkpointPath = "ptuning/output/adgen-chatglm-6b-pt-128-2e-2-1/checkpoint-1000"
# 输出中含有<UNK>
# checkpointPath = "ptuning/output/adgen-chatglm-6b-pt-128-2e-2-1/checkpoint-2000"
# 输出中含有<UNK>
# checkpointPath = "ptuning/output/adgen-chatglm-6b-pt-128-2e-2-1/checkpoint-3000"
# 重复回答一个答案
# checkpointPath = "ptuning/output/adgen-chatglm-6b-pt-128-2e-1-weitiao1/checkpoint-1000"
checkpointPath = "ptuning/output/adgen-chatglm-6b-pt-128-2e-2-微调3/checkpoint-3000"
preSeqLen = 128

# Load model and tokenizer of ChatGLM-6B
config = AutoConfig.from_pretrained(modelPath, trust_remote_code=True, pre_seq_len=preSeqLen)
tokenizer = AutoTokenizer.from_pretrained(modelPath, trust_remote_code=True)
model = AutoModel.from_pretrained(modelPath, config=config, trust_remote_code=True)

# Load PrefixEncoder
prefix_state_dict = torch.load(os.path.join(checkpointPath, "pytorch_model.bin"))
new_prefix_state_dict = {}
for k, v in prefix_state_dict.items():
    new_prefix_state_dict[k[len("transformer.prefix_encoder."):]] = v
model.transformer.prefix_encoder.load_state_dict(new_prefix_state_dict)

print(f"Quantized to 4 bit")
model = model.quantize(4)
model = model.half().cuda()
model.transformer.prefix_encoder.float()
model = model.eval()

os_name = platform.system()
clear_command = 'cls' if os_name == 'Windows' else 'clear'
stop_stream = False


def build_prompt(history):
    prompt = "欢迎使用 ChatGLM-6B 模型，输入内容即可进行对话，clear 清空对话历史，stop 终止程序"
    for query, response in history:
        prompt += f"\n\n用户：{query}"
        prompt += f"\n\nChatGLM-6B：{response}"
    return prompt


def signal_handler(signal, frame):
    global stop_stream
    stop_stream = True


def main():
    history = []
    global stop_stream
    print("欢迎使用 ChatGLM-6B 模型，输入内容即可进行对话，clear 清空对话历史，stop 终止程序")
    while True:
        query = input("\n用户：")
        if query.strip() == "stop":
            break
        if query.strip() == "clear":
            history = []
            os.system(clear_command)
            print("欢迎使用 ChatGLM-6B 模型，输入内容即可进行对话，clear 清空对话历史，stop 终止程序")
            continue
        count = 0
        for response, history in model.stream_chat(tokenizer, query, history=history):
            if stop_stream:
                stop_stream = False
                break
            else:
                count += 1
                if count % 8 == 0:
                    os.system(clear_command)
                    print(build_prompt(history), flush=True)
                    signal.signal(signal.SIGINT, signal_handler)
        os.system(clear_command)
        print(build_prompt(history), flush=True)


if __name__ == "__main__":
    main()