On my laptop I use pickle abundantly with the routine
def read_pickle(source_file): with open(source_file, "rb") as fp: data = pickle.load(fp) fp.close() return data
When using that as
data = read_pickle(datafile)
it throws the error
/usr/local/lib/python3.7/dist-packages/numba/core/serialize.py in _unpickle__CustomPickled(serialized) 206 Uses `NumbaPickler` to load. 207 """ --> 208 ctor, states = loads(serialized) 209 return _CustomPickled(ctor, states) 210 ModuleNotFoundError: No module named 'numba.cloudpickle'
However, all packages are installed. Is this a typical problem for working in colab?. I'm happy with a work around that isn't as fast/efficient etc.
thanks,
Andreas