in Azure Machine Learning service, how to update python pickle file 's ML model parameter?

윤 광섭 1

after using Azure Auto ML or Designer, model creates Pickle file.

i tried to develop model in python pickle file by tunning some parameter by set_params method

as you know this pickle file has 'Pipeline', 'y_transformer', 'y_transformer_name' key.

and Pipeline has parameter setting

but as i wrote,

a.pipeline.set_params (memory = 'n')

this code is work.

memory parameter is changed.

but

a.pipeline.set_params (XGBoostClassifier__base_score = 0.6)

this code is not work.

please let me know how to change model's hyperparameter

romungi-MSFT 43,246 Reputation points Microsoft Employee

2020-06-15T13:29:15.36+00:00

@윤 광섭 Based on some available info on github forums I came across this issue where xgboost is not installed by default in training which caused issues to users after training. Could you please check if your case is similar?

Also, is it possible to check get_params() is available to check if this param can be set?
romungi-MSFT 43,246 Reputation points Microsoft Employee

2020-06-23T08:12:01.987+00:00

@윤 광섭 Does the above information help to retrain and check if all parameters are available to be set?