A way to support stream_chat() when using tool_call()

[alin995]

A way to support stream_chat() when using tool_call()

this is my code

# Split ChatGLMForConditionalGeneration.process_response into 2 methods
def inject_process_response(output, history):
    content = ""
    history = deepcopy(history)
    for response in output.split("<|assistant|>"):
        if "\n" in response:
            metadata, content = response.split("\n", maxsplit=1)
        else:
            metadata, content = "", response
        if not metadata.strip():
            content = content.strip()
            history.append({"role": "assistant", "metadata": metadata, "content": content})
            content = content.replace("[[训练时间]]", "2023年")
        else:
            history.append({"role": "assistant", "metadata": metadata, "content": content})
            if history[0]["role"] != "system" or "tools" not in history[0]:
                content = {"name": metadata.strip(), "content": content}
    return content, history


def process_response(output, history):
    content = output
    history = deepcopy(history)

    metadata = history[-1]["metadata"].strip()
    if not metadata:
        return content, history

    content = "\n".join(content.split("\n")[1:-1])

    def tool_call(**kwargs):
        return kwargs

    parameters = eval(content)
    content = {"name": metadata.strip(), "parameters": parameters}

    return content, history

# initialize tokenizer and model
tokenizer = AutoTokenizer.from_pretrained(TOKENIZER_PATH, trust_remote_code=True)
model = AutoModel.from_pretrained(MODEL_PATH, trust_remote_code=True, device_map="auto").eval()

# inject model.process_response
model.process_response = inject_process_response

# chat with ChatGLM
response = ""
for response, history, past_key_values in model.stream_chat(tokenizer, query,
                                                            history=history,
                                                            role=role,
                                                            past_key_values=past_key_values,
                                                            return_past_key_values=True):
    # do anything you like
    print(response, end="\n", flush=True)
# process history
response, history = process_response(response, history)
# that's all

[zRzRzRzRzRzRzR]

using tool_call() 没有必要使用stream chat 呀，这不是要整段输出被读取后执行工具吗

[thundax-lyp]

对没有tool_call()的应答呢，比如只注册了get_weather，但是问了续写

[zRzRzRzRzRzRzR]

这个模型本身也估计没有回答的能力，可以试试