5049f57afbab4e6ede9e6b91a26857ed5668307e,examples/tune_hyperparams.py,,,#,7

Before Change

    xg_learning_rate = ContinuousHyperparam(["model","shallow_model","learning_rate"],0,1)
    scale_pos_weight = CategoricalHyperparam(["model","shallow_model","scale_pos_weight"],[1,10.0,100.0])
    num_samples = CategoricalHyperparam(["model","shallow_model","num_samples"],[10000,100000,1000000,1e7])
    tunables = [shallow_model,n_estimators,max_depth,C,kernel,xg_learning_rate,scale_pos_weight,num_samples] //target
else:
    //for DL
    lr = LogContinuousHyperparam(["model","lr"],1e-7,1e-4)

After Change

    xg_learning_rate = ContinuousHyperparam(["model","shallow_model","learning_rate"],0,1)
    scale_pos_weight = CategoricalHyperparam(["model","shallow_model","scale_pos_weight"],[1,10.0,100.0])
    num_samples = CategoricalHyperparam(["model","shallow_model","num_samples"],[10000,100000,1000000,1e7])
    hidden_size = CategoricalHyperparam(["model","final_hidden_layer_size"],[5,10,20])
    hidden_num = CategoricalHyperparam(["model","num_hidden_layers"],[2,4])
    mlp_learning_rate = CategoricalHyperparam(["model","learning_rate_mlp"],[0.001,0.0001,0.00001])
    tunables = [shallow_model,n_estimators,max_depth,C,kernel,xg_learning_rate,scale_pos_weight,num_samples,hidden_num,hidden_size,mlp_learning_rate] //target
else:

Instances