使用spyder运行bert的run_classifier文件出现报错“\udcd5”

Question

问题遇到的现象和发生背景

nlp小白使用spyder运行bert模型的glue项目中的MRPC数据集，但是报错UnicodeEncodeError: 'utf-8' codec can't encode character '\udcd5' in position 196: surrogates not allowed

模型输入参数如下：

--task_name="MRPC" 
--do_train="true" 
--do_eval="true" 
--data_dir="…//GLUE//glue_data//MRPC//" 
--vocab_file="…//GLUE//BERT_BASE_DIR//uncased_L-12_H-768_A-12//vocab.txt" 
--bert_config_file="…//GLUE//BERT_BASE_DIR//uncased_L-12_H-768_A-12//bert_config.json" 
--init_checkpoint="…//GLUE//BERT_BASE_DIR//uncased_L-12_H-768_A-12//bert_model.ckpt" 
--max_seq_length="128" 
--train_batch_size="1" 
--learning_rate="2e-5" 
--num_train_epochs="1.0" 
--output_dir="…//GLUE//output"

问题相关代码，请勿粘贴截图

run_classifier.py直接运行，代码在GitHub上google的bert项目即可下载

运行结果及报错内容

ERROR:tornado.general:Uncaught exception in ZMQStream callback
Traceback (most recent call last):
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\zmq\eventloop\zmqstream.py", line 431, in _run_callback
    callback(*args, **kwargs)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\ipykernel\iostream.py", line 126, in _handle_event
    event_f()
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\ipykernel\iostream.py", line 498, in _flush
    parent=self.parent_header, ident=self.topic)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\jupyter_client\session.py", line 742, in send
    to_send = self.serialize(msg, ident)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\jupyter_client\session.py", line 630, in serialize
    content = self.pack(content)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\jupyter_client\session.py", line 83, in <lambda>
    ensure_ascii=False, allow_nan=False,
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\zmq\utils\jsonapi.py", line 25, in dumps
    return json.dumps(o, **kwargs).encode("utf8")
UnicodeEncodeError: 'utf-8' codec can't encode character '\udcd5' in position 196: surrogates not allowed
ERROR:tornado.general:Uncaught exception in zmqstream callback
Traceback (most recent call last):
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\zmq\eventloop\zmqstream.py", line 448, in _handle_events
    self._handle_recv()
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\zmq\eventloop\zmqstream.py", line 477, in _handle_recv
    self._run_callback(callback, msg)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\zmq\eventloop\zmqstream.py", line 431, in _run_callback
    callback(*args, **kwargs)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\ipykernel\iostream.py", line 126, in _handle_event
    event_f()
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\ipykernel\iostream.py", line 498, in _flush
    parent=self.parent_header, ident=self.topic)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\jupyter_client\session.py", line 742, in send
    to_send = self.serialize(msg, ident)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\jupyter_client\session.py", line 630, in serialize
    content = self.pack(content)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\jupyter_client\session.py", line 83, in <lambda>
    ensure_ascii=False, allow_nan=False,
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\zmq\utils\jsonapi.py", line 25, in dumps
    return json.dumps(o, **kwargs).encode("utf8")
UnicodeEncodeError: 'utf-8' codec can't encode character '\udcd5' in position 196: surrogates not allowed
Exception in callback BaseAsyncIOLoop._handle_events(2244, 1)
handle: <Handle BaseAsyncIOLoop._handle_events(2244, 1)>
Traceback (most recent call last):
  File "C:\Users\40701\.conda\envs\test\lib\asyncio\events.py", line 88, in _run
    self._context.run(self._callback, *self._args)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\tornado\platform\asyncio.py", line 189, in _handle_events
    handler_func(fileobj, events)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\zmq\eventloop\zmqstream.py", line 448, in _handle_events
    self._handle_recv()
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\zmq\eventloop\zmqstream.py", line 477, in _handle_recv
    self._run_callback(callback, msg)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\zmq\eventloop\zmqstream.py", line 431, in _run_callback
    callback(*args, **kwargs)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\ipykernel\iostream.py", line 126, in _handle_event
    event_f()
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\ipykernel\iostream.py", line 498, in _flush
    parent=self.parent_header, ident=self.topic)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\jupyter_client\session.py", line 742, in send
    to_send = self.serialize(msg, ident)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\jupyter_client\session.py", line 630, in serialize
    content = self.pack(content)
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\jupyter_client\session.py", line 83, in <lambda>
    ensure_ascii=False, allow_nan=False,
  File "C:\Users\40701\.conda\envs\test\lib\site-packages\zmq\utils\jsonapi.py", line 25, in dumps
    return json.dumps(o, **kwargs).encode("utf8")
UnicodeEncodeError: 'utf-8' codec can't encode character '\udcd5' in position 196: surrogates not allowed

我的解答思路和尝试过的方法

问了别人说是文件里面可能有中文，但是因为代码能力有限，找不到是什么文件。
然后我把MRPC文件夹中的文件用记事本打开以后，找到第196行，并未发现出现中文

我想要达到的结果

希望能把run_classifier文件跑通，不影响后续学习

Answer 1

从网上相关问题帖子看，可能是文件路径问题，检查文件路径，使用单斜杠作路径分隔符试试。