Model loading doesn't run because of joblib/sklearn incompatibility?


I have trained my model (privately) on a dataset, then I downloaded it (model.joblib) and I get the following error.

I tried all sorts of solutions I found on internet; e.g. versions incompatibility between sklearn and joblib, etc.

At the end, it comes down to the fact that I don’t know what versions HF used to train the model… it’s not mentioned anywhere in the files it generated…

I already spent hours on this… as a non-ML expert, I thought I’d save time but it seems I spent hours debugging and looking for solutions…

I also checked documentation, I didn’t find anything related to my issue…

FYI sklearn = 1.3.0 & joblib = 1.3.1

import json
import joblib

model = joblib.load('model.joblib')
config = json.load(open('config.json'))

features = config['features']

# data = pd.read_csv("data.csv")
data = data[features]
data.columns = ["feat_" + str(col) for col in data.columns]

predictions = model.predict(data)  # or model.predict_proba(data)

Could you please help ??

KeyError                                  Traceback (most recent call last)
Cell In[5], line 8
      2 import joblib
      3 # import sklearn
      5 # import sklearn.external.joblib as joblib
      6 # import joblib
----> 8 model = joblib.load('model.joblib')
      9 config = json.load(open('config.json'))
     11 features = config['features']

File ~/.local/lib/python3.10/site-packages/joblib/, in load(filename, mmap_mode)
    652             if isinstance(fobj, str):
    653                 # if the returned file object is a string, this means we
    654                 # try to load a pickle file generated with an version of
    655                 # Joblib so we load it with joblib compatibility function.
    656                 return load_compatibility(fobj)
--> 658             obj = _unpickle(fobj, filename, mmap_mode)
    659 return obj

File ~/.local/lib/python3.10/site-packages/joblib/, in _unpickle(fobj, filename, mmap_mode)
    575 obj = None
    576 try:
--> 577     obj = unpickler.load()
-> 1213         dispatch[key[0]](self)
   1214 except _Stop as stopinst:
   1215     return stopinst.value

KeyError: 118

Is my question in the right / accurate forum or should I move it ? I’d really appreciate your help…

I’m having the exact same error, and a month ago my code runs perfectly.

Did you download the model using joblib library or using sklearn library?