for testing and deploying your application
for finding and fixing issues
for empowering human code reviews
from sklearn.model_selection import cross_val_score
from lightgbm import LGBMRegressor
from sklearn.datasets import load_diabetes
from hyperactive import Hyperactive
data = load_diabetes()
X, y = data.data, data.target
def model(opt):
lgbm = LGBMRegressor(
num_leaves=opt["num_leaves"],
bagging_freq=opt["bagging_freq"],
learning_rate=opt["learning_rate"],
)
scores = cross_val_score(lgbm, X, y, cv=3)
return scores.mean()
search_space = {
"num_leaves": list(range(2, 50)),
"bagging_freq": list(range(2, 12)),
"learning_rate": [1e-3, 1e-2, 1e-1, 0.5, 1.0],
}
hyper = Hyperactive()
hyper.add_search(model, search_space, n_iter=20)
hyper.run()
SimonBlanke /
Hyperactive
