如何检索用于训练此 xgboost 助推器类型 model 的超参数？

Question

I have an xgboost model that is trained already. It was trained by the xgboost original API. I am trying to find the hyper-parameters upon which the trained model was trained. Most specifically, I want to retrieve the objective of the trained model.

xgb.__versions__ # returns '1.7.2'
type(model) # returns xgboost.core.Booster
model.params() # AttibuteError: 'Booster' object has no attribute 'params'
model.get_params() # AttibuteError: 'Booster' object has no attribute 'get_params'

How can I retrieve the hyper-parameters that were used to train this xgboost booster type model?

Answer 1

You may find these of interest.

model.attributes()

model.num_features

model.feature_names

model.feature_types

Also, several methods for persisting a model can tell you about what it contains, including

• .save_config
• .save_model
• .save_raw

.get_dump() : https://xgboost.readthedocs.io/en/stable/python/python_api.html#xgboost.Booster.get_dump

... the output format is primarily used for visualization or interpretation, hence it's more human readable...

---

Let us know what you found!