model_chatglm3.py

"""
Inference code for ChatGLM.
Adapted from https://huggingface.co/THUDM/chatglm3-6b/blob/main/modeling_chatglm.py.
"""
import re

import torch
from transformers.generation.logits_process import LogitsProcessor


class InvalidScoreLogitsProcessor(LogitsProcessor):
    def __call__(
        self, input_ids: torch.LongTensor, scores: torch.FloatTensor
    ) -> torch.FloatTensor:
        if torch.isnan(scores).any() or torch.isinf(scores).any():
            scores.zero_()
            scores[..., 5] = 5e4
        return scores


invalid_score_processor = InvalidScoreLogitsProcessor()


def process_response(output):
    content = ""
    for response in output.split("<|assistant|>"):
        metadata, content = response.split("\n", maxsplit=1)
        if not metadata.strip():
            content = content.strip()
            content = content.replace("[[训练时间]]", "2023年")
        else:
            try:
                parameters = eval(content)
                content = {"name": metadata.strip(), "parameters": parameters}
            except Exception as e:
                print(e)
                content = {"name": metadata.strip(), "content": content}
    return content


@torch.inference_mode()
def generate_stream_chatglm3(
    model,
    tokenizer,
    params,
    device,
    context_len=2048,
    stream_interval=2,
    judge_sent_end=False,
):
    prompt = params["prompt"]
    temperature = float(params.get("temperature", 1.0))
    repetition_penalty = float(params.get("repetition_penalty", 1.0))
    top_p = float(params.get("top_p", 1.0))
    max_new_tokens = int(params.get("max_new_tokens", 256))
    echo = params.get("echo", True)

    inputs = tokenizer.build_chat_input(prompt,role="user").to(model.device)
    input_echo_len = len(inputs["input_ids"][0])

    gen_kwargs = {
        "max_length": max_new_tokens + input_echo_len,
        "do_sample": True if temperature > 1e-5 else False,
        "top_p": top_p,
        "repetition_penalty": repetition_penalty,
        "logits_processor": [invalid_score_processor],
    }
    if temperature > 1e-5:
        gen_kwargs["temperature"] = temperature

    total_len = 0
    for total_ids in model.stream_generate(**inputs, **gen_kwargs):
        total_ids = total_ids.tolist()[0]
        total_len = len(total_ids)
        if echo:
            output_ids = total_ids
        else:
            output_ids = total_ids[input_echo_len:]
        response = tokenizer.decode(output_ids)
        response = process_response(response)

        yield {
            "text": response,
            "usage": {
                "prompt_tokens": input_echo_len,
                "completion_tokens": total_len - input_echo_len,
                "total_tokens": total_len,
            },
            "finish_reason": None,
        }

    # TODO: ChatGLM stop when it reach max length
    # Only last stream result contains finish_reason, we set finish_reason as stop
    ret = {
        "text": response,
        "usage": {
            "prompt_tokens": input_echo_len,
            "completion_tokens": total_len - input_echo_len,
            "total_tokens": total_len,
        },
        "finish_reason": "stop",
    }
    yield ret