现在需要`pydev_umd`吗？

Question

我刚刚升级到macOS蒙特利，正在尝试运行上周在macOS Catalina上运行的Dask/Python代码。我已经从YAML文件中删除并重新创建了我的conda环境。我还为蒙特利将Python升级到了3.9.7版本。现在，我尝试从Dask工作进程在我的本地机器上运行时出现以下错误。

distributed.protocol.core - CRITICAL - Failed to deserialize
Traceback (most recent call last):
  File "/Users/awd/opt/anaconda3/envs/SpikedTensor/lib/python3.9/site-packages/distributed/protocol/core.py", line 111, in loads
    return msgpack.loads(
  File "msgpack/_unpacker.pyx", line 194, in msgpack._cmsgpack.unpackb
  File "/Users/awd/opt/anaconda3/envs/SpikedTensor/lib/python3.9/site-packages/distributed/protocol/core.py", line 103, in _decode_default
    return merge_and_deserialize(
  File "/Users/awd/opt/anaconda3/envs/SpikedTensor/lib/python3.9/site-packages/distributed/protocol/serialize.py", line 488, in merge_and_deserialize
    return deserialize(header, merged_frames, deserializers=deserializers)
  File "/Users/awd/opt/anaconda3/envs/SpikedTensor/lib/python3.9/site-packages/distributed/protocol/serialize.py", line 401, in deserialize
    deserialize(
  File "/Users/awd/opt/anaconda3/envs/SpikedTensor/lib/python3.9/site-packages/distributed/protocol/serialize.py", line 417, in deserialize
    return loads(header, frames)
  File "/Users/awd/opt/anaconda3/envs/SpikedTensor/lib/python3.9/site-packages/distributed/protocol/serialize.py", line 96, in pickle_loads
    return pickle.loads(x, buffers=new)
  File "/Users/awd/opt/anaconda3/envs/SpikedTensor/lib/python3.9/site-packages/distributed/protocol/pickle.py", line 75, in loads
    return pickle.loads(x)
ModuleNotFoundError: No module named 'pydev_umd'

在最新的Dask或其他什么地方看起来是个问题。这是environment.yml

name: SpikedTensor
channels:
  - conda-forge
  - defaults
dependencies:
  - python=3.9
  - numpy
  - pandas
  - dask
  - sqlalchemy
  - pysqlite3
  - pg8000
  - google-api-python-client
  - google-cloud-sdk
  - cloud-sql-python-connector
  - blas[build=mkl]
  - pip
  - pip:
    - dask-saturn>=0.3.0
prefix: /Users/awd/opt/anaconda3/envs/SpikedTensor

这是一种非常无痛的配置。这个错误看起来像是一个依赖问题。

Answer 1

THis似乎是从控制台运行的，而不是从PyCharm内部运行的。因此，我将连接他们的支持并报告他们的解决方案。