How could I load a specific model from all saved models?

[VectorZhao]

I trained a model using the following code:

automl = AutoML(
    results_path="AutoML_model_CaNaK_unscaled",
    ml_task="multiclass_classification",
    algorithms=["CatBoost", "Xgboost", "LightGBM"],
    model_time_limit=30*60,
    start_random_models=10,
    hill_climbing_steps=3,
    top_models_to_improve=3,
    golden_features=False,
    features_selection=False,
#     stack_models=True,
#     train_ensemble=True,
    explain_level=2,
    validation_strategy={
        "validation_type": "split",
        "train_ratio": 0.75,
        "shuffle": True,
        "stratify": True
    }
)

and now I got lots of models in the "AutoML_model_CaNaK_unscaled" path.
I want to get shap values for a specific model such as 16_LightGBM following the shap git code like the following:

import xgboost
import shap

# train an XGBoost model
X, y = shap.datasets.boston()
model = xgboost.XGBRegressor().fit(X, y)

# explain the model's predictions using SHAP
# (same syntax works for LightGBM, CatBoost, scikit-learn, transformers, Spark, etc.)
explainer = shap.Explainer(model)
shap_values = explainer(X)

I know that using AutoML(results_path="AutoML_model_CaNaK_unscaled") can load the well trained model, but loading a specific model from saved models is what I want.

[pplonski]

Hi @VectorZhao,

It is not yet implemented. There is an issue with adding functionality to load only selected model for computing predictions #423.

You can try to reuse the parameters of the selected model and train only it with some custom script.

[VectorZhao]

Hi @VectorZhao,

It is not yet implemented. There is an issue with adding functionality to load only selected model for computing predictions #423.

You can try to reuse the parameters of the selected model and train only it with some custom script.

Ok, many thanks for your reply.