binary-husky/gpt_academic · error · RuntimeError

用户取消了程序。

Error message

用户取消了程序。

What it means

RuntimeError ('用户取消了程序。') raised by the watchdog in the ChatGPT streaming reader. observe_window[1] holds a timestamp that the Gradio UI refreshes on every tick; if it ages past watch_dog_patience (5 s) while delta content is still arriving, the bridge concludes the user pressed stop (or the UI thread died) and aborts the stream. It is a deliberate cooperative-cancellation check evaluated only when new content chunks arrive.

Source

Thrown at request_llms/bridge_chatgpt.py:214

        if len(delta) == 0:
            is_termination_certain = False
            if (has_choices) and (chunkjson['choices'][0].get('finish_reason', 'null') == 'stop'): is_termination_certain = True
            if is_termination_certain: break
            else: continue # 对于不符合规范的狗屎接口，这里需要继续

        if (not has_content) and has_role: continue
        if (not has_content) and (not has_role): continue # raise RuntimeError("发现不标准的第三方接口："+delta)
        if has_content: # has_role = True/False
            result += delta["content"]
            if not console_silence: print(delta["content"], end='')
            if observe_window is not None:
                # 观测窗，把已经获取的数据显示出去
                if len(observe_window) >= 1:
                    observe_window[0] += delta["content"]
                # 看门狗，如果超过期限没有喂狗，则终止
                if len(observe_window) >= 2:
                    if (time.time()-observe_window[1]) > watch_dog_patience:
                        raise RuntimeError("用户取消了程序。")
        else: raise RuntimeError("意外Json结构："+delta)

    finish_reason = json_data.get('finish_reason', None) if json_data else None
    if finish_reason == 'content_filter':
        raise RuntimeError("由于提问含不合规内容被过滤。")
    if finish_reason == 'length':
        raise ConnectionAbortedError("正常结束，但显示Token不足，导致输出不完整，请削减单次输入的文本量。")

    return result


def predict(inputs:str, llm_kwargs:dict, plugin_kwargs:dict, chatbot:ChatBotWithCookies,
            history:list=[], system_prompt:str='', stream:bool=True, additional_fn:str=None):
    """
    发送至chatGPT，流式获取输出。
    用于基础的对话功能。
    inputs 是本次问询的输入
    top_p, temperature是chatGPT的内部调优参数

View on GitHub (pinned to d6bde0fa54)

Solutions

1. If you stopped generation deliberately, no fix is needed — re-ask the question.
2. Keep the gpt_academic browser tab in the foreground and responsive during generation so the UI keeps feeding the watchdog.
3. Custom callers must run a loop setting observe_window[1] = time.time() at least every few seconds for the whole stream.
4. For systematically slow setups, raise watch_dog_patience in request_llms/bridge_chatgpt.py.

Example fix

# caller-side feeding loop
import threading, time

def feed_watchdog(observe_window, stop_event):
    while not stop_event.is_set():
        observe_window[1] = time.time()
        time.sleep(1)

try:
    reply = predict_no_ui_long_connection(inputs, llm_kwargs, history, sys_prompt, observe_window=win)
except RuntimeError as e:
    if e.args[0] == '用户取消了程序。' and user_pressed_stop:
        acknowledge_cancel()
    else:
        raise  # watchdog fired but nobody cancelled -> UI feeding problem

When it happens

Trigger: User clicks the stop button in the gpt_academic UI during generation; the UI thread stops updating observe_window[1] (browser tab suspended, Gradio queue stalled); the caller passes observe_window with a stale [1] timestamp and never refreshes it while content streams.

Common situations: Intentional stop mid-answer; mobile browser backgrounding the tab freezes the Gradio polling; long generations on slow relays combined with an unresponsive UI thread; plugin code copying the observe_window pattern but forgetting the feeding thread.

Related errors

• 用户取消了程序。
• 用户取消了程序。
• 用户取消了程序。
• 程序终止。
• 用户取消了程序。

AI-assisted analysis of binary-husky/gpt_academic@d6bde0fa54 (2026-08-14). Data as JSON: /api/errors/4edae61431acfb90. Report an issue: GitHub.