Spacy: Multiprocessing example not working due to Pickle error

Created on 25 Jan 2019  路  8Comments  路  Source: explosion/spaCy

How to reproduce the behaviour

Running the exact multi_processing example given here is not working for me on Windows 10 Python 3.6 or 3.7.

PS C:\Python\projects\spacy-multithread> pipenv run python .\main.py .\data\output\ en_core_web_sm -n 4
Loaded model 'en_core_web_sm'
Loading IMDB data...
Processing texts...
joblib.externals.loky.process_executor._RemoteTraceback:
"""
Traceback (most recent call last):
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\externals\loky\backend\queues.py", line 150, in _feed
    obj_ = dumps(obj, reducers=reducers)
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\externals\loky\backend\reduction.py", line 243, in dumps
    dump(obj, buf, reducers=reducers, protocol=protocol)
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\externals\loky\backend\reduction.py", line 236, in dump
    _LokyPickler(file, reducers=reducers, protocol=protocol).dump(obj)
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\externals\cloudpickle\cloudpickle.py", line 284, in dump
    return Pickler.dump(self, obj)
  File "c:\python\python37\Lib\pickle.py", line 437, in dump
    self.save(obj)
  File "c:\python\python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "c:\python\python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 856, in save_dict
    self._batch_setitems(obj.items())
  File "c:\python\python37\Lib\pickle.py", line 882, in _batch_setitems
    save(v)
  File "c:\python\python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "c:\python\python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 856, in save_dict
    self._batch_setitems(obj.items())
  File "c:\python\python37\Lib\pickle.py", line 887, in _batch_setitems
    save(v)
  File "c:\python\python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "c:\python\python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 856, in save_dict
    self._batch_setitems(obj.items())
  File "c:\python\python37\Lib\pickle.py", line 882, in _batch_setitems
    save(v)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 816, in save_list
    self._batch_appends(obj)
  File "c:\python\python37\Lib\pickle.py", line 843, in _batch_appends
    save(tmp[0])
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 771, in save_tuple
    save(element)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 786, in save_tuple
    save(element)
  File "c:\python\python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "c:\python\python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 856, in save_dict
    self._batch_setitems(obj.items())
  File "c:\python\python37\Lib\pickle.py", line 882, in _batch_setitems
    save(v)
  File "c:\python\python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "c:\python\python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 786, in save_tuple
    save(element)
  File "c:\python\python37\Lib\pickle.py", line 524, in save
    rv = reduce(self.proto)
  File "stringsource", line 2, in preshed.maps.PreshMap.__reduce_cython__
TypeError: self.c_map cannot be converted to a Python object for pickling
"""

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File ".\main.py", line 77, in <module>
    plac.call(main)
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\plac_core.py", line 328, in call
    cmd, result = parser.consume(arglist)
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\plac_core.py", line 207, in consume
    return cmd, self.func(*(args + varargs + extraopts), **kwargs)
  File ".\main.py", line 41, in main
    executor(tasks)
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\parallel.py", line 934, in __call__
    self.retrieve()
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\parallel.py", line 833, in retrieve
    self._output.extend(job.get(timeout=self.timeout))
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\_parallel_backends.py", line 521, in wrap_future_result
    return future.result(timeout=timeout)
  File "c:\python\python37\Lib\concurrent\futures\_base.py", line 432, in result
    return self.__get_result()
  File "c:\python\python37\Lib\concurrent\futures\_base.py", line 384, in __get_result
    raise self._exception
_pickle.PicklingError: Could not pickle the task to send it to the workers.

Info about spaCy

  • spaCy version: 2.0.18
  • Platform: Windows-10-10.0.17134-SP0
  • Python version: 3.7.1 and 3.6.8
examples 馃寵 nightly
Source
picture BramVanroy

Most helpful comment

Not positive but I think this is related to a change in joblib. On my machine it's working if I stipulate the "multiprocessing" backend in joblib.Parallel, instead of using their new loky library. Change here: 582be874

picture honnibal on 21 Feb 2019
👍3

All 8 comments

I tried the nightly as suggested here https://github.com/explosion/spaCy/issues/1669 but the issue remains:

Trace

PS C:\Python\projects\spacy-multithread> pipenv run python .\main.py .\data\output\ en_core_web_sm -n 4
Loaded model 'en_core_web_sm'
Loading IMDB data...
Processing texts...
joblib.externals.loky.process_executor._RemoteTraceback:
"""
Traceback (most recent call last):
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\externals\loky\backend\queues.py", line 150, in _feed
    obj_ = dumps(obj, reducers=reducers)
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\externals\loky\backend\reduction.py", line 243, in dumps
    dump(obj, buf, reducers=reducers, protocol=protocol)
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\externals\loky\backend\reduction.py", line 236, in dump
    _LokyPickler(file, reducers=reducers, protocol=protocol).dump(obj)
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\externals\cloudpickle\cloudpickle.py", line 284, in dump
    return Pickler.dump(self, obj)
  File "c:\python\python37\Lib\pickle.py", line 437, in dump
    self.save(obj)
  File "c:\python\python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "c:\python\python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 856, in save_dict
    self._batch_setitems(obj.items())
  File "c:\python\python37\Lib\pickle.py", line 882, in _batch_setitems
    save(v)
  File "c:\python\python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "c:\python\python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 856, in save_dict
    self._batch_setitems(obj.items())
  File "c:\python\python37\Lib\pickle.py", line 887, in _batch_setitems
    save(v)
  File "c:\python\python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "c:\python\python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 856, in save_dict
    self._batch_setitems(obj.items())
  File "c:\python\python37\Lib\pickle.py", line 882, in _batch_setitems
    save(v)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 816, in save_list
    self._batch_appends(obj)
  File "c:\python\python37\Lib\pickle.py", line 843, in _batch_appends
    save(tmp[0])
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 771, in save_tuple
    save(element)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 786, in save_tuple
    save(element)
  File "c:\python\python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "c:\python\python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 856, in save_dict
    self._batch_setitems(obj.items())
  File "c:\python\python37\Lib\pickle.py", line 882, in _batch_setitems
    save(v)
  File "c:\python\python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "c:\python\python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "c:\python\python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "c:\python\python37\Lib\pickle.py", line 786, in save_tuple
    save(element)
  File "c:\python\python37\Lib\pickle.py", line 524, in save
    rv = reduce(self.proto)
  File "stringsource", line 2, in preshed.maps.PreshMap.__reduce_cython__
TypeError: self.c_map cannot be converted to a Python object for pickling
"""

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File ".\main.py", line 77, in <module>
    plac.call(main)
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\plac_core.py", line 328, in call
    cmd, result = parser.consume(arglist)
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\plac_core.py", line 207, in consume
    return cmd, self.func(*(args + varargs + extraopts), **kwargs)
  File ".\main.py", line 41, in main
    executor(tasks)
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\parallel.py", line 934, in __call__
    self.retrieve()
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\parallel.py", line 833, in retrieve
    self._output.extend(job.get(timeout=self.timeout))
  File "C:\Users\Bram\.virtualenvs\spacy-multithread-0ZRRAX-8\lib\site-packages\joblib\_parallel_backends.py", line 521, in wrap_future_result
    return future.result(timeout=timeout)
  File "c:\python\python37\Lib\concurrent\futures\_base.py", line 432, in result
    return self.__get_result()
  File "c:\python\python37\Lib\concurrent\futures\_base.py", line 384, in __get_result
    raise self._exception
_pickle.PicklingError: Could not pickle the task to send it to the workers.

Info about spaCy

  • spaCy version: 2.1.0a6
  • Platform: Windows-10-10.0.17134-SP0
  • Python version: 3.7.1
BramVanroy picture BramVanroy on 25 Jan 2019
👍3

I'm running into the same issue

kurtjanssensai picture kurtjanssensai on 1 Feb 2019

Hmm. Could you just do a quick test for me?

>>> import spacy
>>> nlp = spacy.load('en_core_web_sm')
>>> import pickle
>>> b = pickle.dumps(nlp)

This is working for me, so I'm not sure what's going wrong. Would be interested to see whether this is failing for you.

honnibal picture honnibal on 7 Feb 2019

The pickling of the model itself seems to go fine. I can do this without issue:

import spacy
import pickle

nlp = spacy.load('en_core_web_sm')
b = pickle.dumps(nlp)
c = pickle.loads(b)

But something else causes an issue, it seems:

Loaded model 'en_core_web_sm'
Loading IMDB data...
Downloading data from http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz
84123648/84125825 [============================>.] - ETA: 0s Untaring file...



Processing texts...
joblib.externals.loky.process_executor._RemoteTraceback: 
"""
Traceback (most recent call last):
  File "C:\Users\bmvroy\.virtualenvs\spacy_mp_test-4MLbpdkD\lib\site-packages\joblib\externals\loky\backend\queues.py", line 150, in _feed
    obj_ = dumps(obj, reducers=reducers)
  File "C:\Users\bmvroy\.virtualenvs\spacy_mp_test-4MLbpdkD\lib\site-packages\joblib\externals\loky\backend\reduction.py", line 243, in dumps
    dump(obj, buf, reducers=reducers, protocol=protocol)
  File "C:\Users\bmvroy\.virtualenvs\spacy_mp_test-4MLbpdkD\lib\site-packages\joblib\externals\loky\backend\reduction.py", line 236, in dump
    _LokyPickler(file, reducers=reducers, protocol=protocol).dump(obj)
  File "C:\Users\bmvroy\.virtualenvs\spacy_mp_test-4MLbpdkD\lib\site-packages\joblib\externals\cloudpickle\cloudpickle.py", line 284, in dump
    return Pickler.dump(self, obj)
  File "C:\Python\Python37\Lib\pickle.py", line 437, in dump
    self.save(obj)
  File "C:\Python\Python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "C:\Python\Python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "C:\Python\Python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "C:\Python\Python37\Lib\pickle.py", line 856, in save_dict
    self._batch_setitems(obj.items())
  File "C:\Python\Python37\Lib\pickle.py", line 882, in _batch_setitems
    save(v)
  File "C:\Python\Python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "C:\Python\Python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "C:\Python\Python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "C:\Python\Python37\Lib\pickle.py", line 856, in save_dict
    self._batch_setitems(obj.items())
  File "C:\Python\Python37\Lib\pickle.py", line 887, in _batch_setitems
    save(v)
  File "C:\Python\Python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "C:\Python\Python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "C:\Python\Python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "C:\Python\Python37\Lib\pickle.py", line 856, in save_dict
    self._batch_setitems(obj.items())
  File "C:\Python\Python37\Lib\pickle.py", line 882, in _batch_setitems
    save(v)
  File "C:\Python\Python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "C:\Python\Python37\Lib\pickle.py", line 816, in save_list
    self._batch_appends(obj)
  File "C:\Python\Python37\Lib\pickle.py", line 843, in _batch_appends
    save(tmp[0])
  File "C:\Python\Python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "C:\Python\Python37\Lib\pickle.py", line 771, in save_tuple
    save(element)
  File "C:\Python\Python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "C:\Python\Python37\Lib\pickle.py", line 786, in save_tuple
    save(element)
  File "C:\Python\Python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "C:\Python\Python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "C:\Python\Python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "C:\Python\Python37\Lib\pickle.py", line 856, in save_dict
    self._batch_setitems(obj.items())
  File "C:\Python\Python37\Lib\pickle.py", line 882, in _batch_setitems
    save(v)
  File "C:\Python\Python37\Lib\pickle.py", line 549, in save
    self.save_reduce(obj=obj, *rv)
  File "C:\Python\Python37\Lib\pickle.py", line 662, in save_reduce
    save(state)
  File "C:\Python\Python37\Lib\pickle.py", line 504, in save
    f(self, obj) # Call unbound method with explicit self
  File "C:\Python\Python37\Lib\pickle.py", line 786, in save_tuple
    save(element)
  File "C:\Python\Python37\Lib\pickle.py", line 524, in save
    rv = reduce(self.proto)
  File "stringsource", line 2, in preshed.maps.PreshMap.__reduce_cython__
TypeError: self.c_map cannot be converted to a Python object for pickling
"""

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "C:/Python/projects/spacy_mp_test/main.py", line 77, in <module>
    plac.call(main)
  File "C:\Users\bmvroy\.virtualenvs\spacy_mp_test-4MLbpdkD\lib\site-packages\plac_core.py", line 328, in call
    cmd, result = parser.consume(arglist)
  File "C:\Users\bmvroy\.virtualenvs\spacy_mp_test-4MLbpdkD\lib\site-packages\plac_core.py", line 207, in consume
    return cmd, self.func(*(args + varargs + extraopts), **kwargs)
  File "C:/Python/projects/spacy_mp_test/main.py", line 41, in main
    executor(tasks)
  File "C:\Users\bmvroy\.virtualenvs\spacy_mp_test-4MLbpdkD\lib\site-packages\joblib\parallel.py", line 934, in __call__
    self.retrieve()
  File "C:\Users\bmvroy\.virtualenvs\spacy_mp_test-4MLbpdkD\lib\site-packages\joblib\parallel.py", line 833, in retrieve
    self._output.extend(job.get(timeout=self.timeout))
  File "C:\Users\bmvroy\.virtualenvs\spacy_mp_test-4MLbpdkD\lib\site-packages\joblib\_parallel_backends.py", line 521, in wrap_future_result
    return future.result(timeout=timeout)
  File "C:\Python\Python37\Lib\concurrent\futures\_base.py", line 432, in result
    return self.__get_result()
  File "C:\Python\Python37\Lib\concurrent\futures\_base.py", line 384, in __get_result
    raise self._exception
_pickle.PicklingError: Could not pickle the task to send it to the workers.
BramVanroy picture BramVanroy on 7 Feb 2019

The sample example works only with the latest nightly but not with 2.0.18. However I got the same error as the comment above when running multiprocess.py with the nightly version.

bsatts picture bsatts on 7 Feb 2019

Not positive but I think this is related to a change in joblib. On my machine it's working if I stipulate the "multiprocessing" backend in joblib.Parallel, instead of using their new loky library. Change here: 582be874

honnibal picture honnibal on 21 Feb 2019
👍3

Tested and approved. Works as expected now.

BramVanroy picture BramVanroy on 21 Feb 2019

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.

lock[bot] picture lock[bot] on 23 Mar 2019
Was this page helpful?
0 / 5 - 0 ratings

Related issues

bebelbop picture bebelbop  路  3Comments
TropComplique picture TropComplique  路  3Comments
besirkurtulmus picture besirkurtulmus  路  3Comments
melanietosik picture melanietosik  路  3Comments
cverluise picture cverluise  路  3Comments