Please help me figure it out. I have a pre-trained xgboost model on version 0.6. I want to use this model on version 0.90. To do this, I selected a kernel with version 0.6, loaded the model from pickle and saved it:
with open (‘old_xgboost.pickle’, ‘rb’) as f
model = pickle.load(f)
model._Booster.save_model(‘new_xgboost.model’)
Next, I go back to the kernel with version xgboost = 0.90 and try to load it:
new_model = xgboost.XGBClassifier() #init model
new_model.load_model(‘new_xgboost.model’) #load
When I print this new model, I do not see the parameters that the old model should have had (regular XGBClassifier () ). If I try model = new_model.load_model('new_xgboost.model') I get the type “None”. What am I doing wrong and how can I use this model for prediction?