Downloading and running models locally RRS feed

  • Question

  • I am trying to run models that were created in the ML portal preview.
    My use case requires running the models without an internet connection. 

    I used to be able to download the files, unpickle them and activate them locally. 

    Is this still possible ?

    I am able to download the model. 
    I get:
    - 1 Conda YML
    - 1 example py file
    - 1 .pkl file

    I used to be able to unpickle and run the model on my local computer. 
    Now I get a key error:

    >>> scoring_AutoMLc4e14b9343.init()
    Traceback (most recent call last):
      File "<stdin>", line 1, in <module>
      File "C:\Users\<User>\Scratch\AutoMLc4e14b9343\", line 26, in init
        model = joblib.load("C:\\Users\\<User>\\Scratch\\AutoMLc4e14b9343\\model_AutoMLc4e14b9343.pkl")
      File "C:\Users\<User>\Anaconda3\envs\project_environment\lib\site-packages\sklearn\externals\joblib\", line 598, in load
        obj = _unpickle(fobj, filename, mmap_mode)
      File "C:\Users\<User>\Anaconda3\envs\project_environment\lib\site-packages\sklearn\externals\joblib\", line 526, in _unpickle
        obj = unpickler.load()
      File "C:\Users\<User>\Anaconda3\envs\project_environment\lib\", line 1050, in load
    KeyError: 239
    I also can't find any documentation or tutorials on how to run these files locally. 

    Casey Forsyth

    Sunday, January 26, 2020 12:46 AM

All replies