対戦ゲームデータ分析甲子園

目指せ"Another" バトル優勝！

賞金: 100,000 円参加ユーザー数: 606 約4年前に終了

参加する

t_mochizuki0

モード別に学習モデルを分けた場合の結果（参考）

lobby-mode別に学習した場合の結果

1. 前提

スプラ2をプレイした方はご存知の通り、このゲームには以下のモードがあります。
（今回のコンペに関係ないモードは除外）

ナワバリバトル
ガチエリア
ガチアサリ
ガチホコ
ガチヤグラ

各モードは勝利条件が異なるため、それぞれ別のゲームと言えるほど、プレーヤーの行動が変わります。
有利/不利なブキの組み合わせも、モードによって変わる可能性があります。
このため今回のデータの学習も、モード別に実施した方がより正確になるのでは？という動機です。

2. やること

以下の学習結果を比較し、どちらの結果が上か確認します。

全データをまとめて学習
学習データをモード別に分け、それぞれで学習

学習モデルは以下を使用させていただきました。
「LightGBMを使ったBase line」

なお、「2.」は各モード毎に学習データを分けて学習し、最後に結果をまとめています。（力技です）

3. 結果

「2.」の方が、Publicスコアが上がりました。

全データをまとめて学習した結果を提出した際の、Publicスコア： 0.536344
各モード毎に学習した結果をまとめて提出した際の、Publicスコア： 0.541567

学習結果のCVを見ると、特に「ナワバリバトル」はCV=0.5996257033039711と高い値でした。
他のモードと分けた効果が出ていると思われます。

ただし、「4. 課題 (1)」で述べる通り、学習データに欠損値が生じるという問題があります。
このため上記は参考値と捉えてください。

4. 課題

(1) 学習データに欠損値が生じる

全データをまとめて学習した際は欠損値が出なかったのですが、モード別に学習データを分けると欠損値が生じます。
change_to_target2()のforループ内で欠損値が発生するのですが、原因は未確認です。
何かお気づきの点がありましたら、コメントで教えていただけると助かります。

5. 改良案

(1)上記方法の問題点

上記方法はナワバリバトルのCVは高かったものの、ガチエリア〜ガチヤグラのCVは低めでした（特にガチヤグラ）。
データ分割により、学習データ自体が減ってしまったことが原因かもしれません。

ナワバリバトル： CV_score:0.5996257033039711
ガチエリア： CV_score:0.5319582472095907
ガチアサリ： CV_score:0.5350757493959036
ガチホコ： CV_score:0.528967034800994
ガチヤグラ： CV_score:0.5206471494607088
平均： CV_score:0.5432547768342336

(2)改良案

ガチエリア〜ガチヤグラは「ガチマッチ」としてまとめて学習します。これにより学習データ数を確保します。
ナワバリバトルはCVが良かったので、分けて学習します。
これらのモードは学習データの"lobby-mode"で判別できます。

regular：レギュラーマッチ = ナワバリバトル
gachi：ガチマッチ = ガチエリア〜ガチヤグラ

(3)改良案の結果

以下の通り、全体としてはCVが上がりました。
ただしガチマッッチのCVはそれほど向上しなかったので、微妙なところかもしれません。

レギュラーマッチ： CV_score:0.5996257033039711
ガチマッッチ： CV_score:0.5277763340574645
平均： CV_score:0.543471347270915

5. 最後に

LightGBMのモデルを開示していただいたOreginさん、ありがとうございました。
基準となるモデルがあるとこのような比較がやりやすいため、たいへん助かりました。

# ライブラリのインポート
import pandas as pd
import numpy as np
import re
import matplotlib.pyplot as plt
import seaborn as sns
import lightgbm as lgb
from sklearn.model_selection import train_test_split
from sklearn.model_selection import KFold
from sklearn.metrics import accuracy_score
import warnings
warnings.filterwarnings('ignore')

# データの読込
train = pd.read_csv("../data/train_data.csv")
test = pd.read_csv('../data/test_data.csv')

前処理

# 欠損値は、全て「－１」とする。
def fill_all_null(df):
    for col_name in df.columns[df.isnull().sum()!=0]:
        df[col_name] = df[col_name].fillna(-1)

# 訓練データ、テストデータの欠損値を補完
fill_all_null(train)
fill_all_null(test)

# ターゲットエンコーディングの関数定義
def change_to_target2(train_df,test_df,input_column_name,output_column_name):
    from sklearn.model_selection import KFold
    
    # nan埋め処理
    train[input_column_name] = train[input_column_name].fillna('-1').isnull().sum()
    test[input_column_name] = test[input_column_name].fillna('-1').isnull().sum()

    kf = KFold(n_splits=5, shuffle=True, random_state=71)
    #=========================================================#
    c=input_column_name
    # 学習データ全体で各カテゴリにおけるyの平均を計算
    data_tmp = pd.DataFrame({c: train_df[c],'target':train_df['y']})
    target_mean = data_tmp.groupby(c)['target'].mean()

    #テストデータのカテゴリを置換
    test_df[output_column_name] = test_df[c].map(target_mean)
    
    # 変換後の値を格納する配列を準備
    tmp = np.repeat(np.nan, train_df.shape[0])
    
    for i, (train_index, test_index) in enumerate(kf.split(train_df)): # NFOLDS回まわる
        #学習データについて、各カテゴリにおける目的変数の平均を計算
        target_mean = data_tmp.iloc[train_index].groupby(c)['target'].mean()
        #バリデーションデータについて、変換後の値を一時配列に格納
        tmp[test_index] = train_df[c].iloc[test_index].map(target_mean) 
                
    #変換後のデータで元の変数を置換
    train_df[output_column_name] = tmp
    #========================================================#   

#　'period','game-ver','mode','lobby'の各列は、勝敗に大きく影響しないと思われるため削除
train = train.drop(labels=['period', 'game-ver', 'mode', 'lobby'], axis=1)
test = test.drop(labels=['period', 'game-ver', 'mode', 'lobby'], axis=1)

# regular, gachiの各lobby-mode別データを取得する
train_regular = train[train['lobby-mode'] == "regular"]
train_gachi   = train[train['lobby-mode'] == "gachi"]

test_regular = test[test['lobby-mode'] == "regular"]
test_gachi   = test[test['lobby-mode'] == "gachi"]

# オブジェクトの列のリストを作成
object_col_list = train.select_dtypes(include=object).columns

# オブジェクトの列は全てターゲットエンコーディング実施
for col in object_col_list:
    change_to_target2(train_regular, test_regular, col, "enc_"+col)
    change_to_target2(train_gachi,   test_gachi,   col, "enc_"+col)

#　変換前の列を削除
train_regular = train_regular.drop(object_col_list,axis=1)
train_gachi   = train_gachi.drop(object_col_list,axis=1)

test_regular  = test_regular.drop(object_col_list,axis=1)
test_gachi    = test_gachi.drop(object_col_list,axis=1)

# 予測結果をあとでまとめるために、idの列を抽出する
test_regular_id = test_regular[['id']].copy()  # DataFrameとして抽出するために、リストをネストする
test_gachi_id   = test_gachi[['id']].copy()

#　'id'の列を削除
train_regular = train_regular.drop('id',axis=1)
train_gachi   = train_gachi.drop('id',axis=1)

test_regular  = test_regular.drop('id',axis=1)
test_gachi    = test_gachi.drop('id',axis=1)

データの確認

#　訓練データに欠損がないことの確認（欠損値がある項目数を表示）
print(train_regular.isnull().values.sum())
print(train_gachi.isnull().values.sum())

#　テストデータに欠損がないことの確認（欠損値がある項目数を表示）
print(test_regular.isnull().values.sum())
print(test_gachi.isnull().values.sum())

学習の準備

# 訓練データを説明変数と目的変数に分割
target_regular = train_regular['y']
train_x_regular = train_regular.drop('y',axis=1)

target_gachi = train_gachi['y']
train_x_gachi = train_gachi.drop('y',axis=1)

# LGBMのパラメータを設定
params = {
    # 二値分類問題
    'objective': 'binary',
    # 損失関数は二値のlogloss
    #'metric': 'auc',
    'metric': 'binary_logloss',
    # 最大イテレーション回数指定
    'num_iterations' : 1000,
    # early_stopping 回数指定
    'early_stopping_rounds' : 100,
}

学習・予測の実行

def predict(train_x, target, test):

    # k-分割交差検証を使って学習＆予測（K=10）
    FOLD_NUM = 10
    kf = KFold(n_splits=FOLD_NUM,
                  random_state=42)

    #検証時のスコアを初期化
    scores = []

    #テストデータの予測値を初期化
    pred_cv = np.zeros(len(test.index))

    #lgbmのラウンド数を定義
    num_round = 10000

    for i, (tdx, vdx) in enumerate(kf.split(train_x, target)):
        print(f'Fold : {i}')
        # 訓練用データと検証用データに分割
        X_train, X_valid, y_train, y_valid = train_x.iloc[tdx], train_x.iloc[vdx], target.values[tdx], target.values[vdx]
        lgb_train = lgb.Dataset(X_train, y_train)
        lgb_valid = lgb.Dataset(X_valid, y_valid)

        # 学習の実行
        model = lgb.train(params, lgb_train, num_boost_round=num_round,
                          valid_names=["train", "valid"], valid_sets=[lgb_train, lgb_valid],
                          verbose_eval=100)

        # 検証データに対する予測値を求めて、勝敗（０　or　１）に変換
        va_pred = np.round(model.predict(X_valid,num_iteration=model.best_iteration))

        # accuracyスコアを計算
        score_ = accuracy_score(y_valid, va_pred)

        # フォールド毎の検証時のスコアを格納
        scores.append(score_)

        #テストデータに対する予測値を求める
        submission = model.predict(test,num_iteration=model.best_iteration)

        #テストデータに対する予測値をフォールド数で割って蓄積
        #(フォールド毎の予測値の平均値を求めることと同じ)
        pred_cv += submission/FOLD_NUM

    # 最終的なテストデータに対する予測値を勝敗（０　or　１）に変換
    pred_cv = np.round(pred_cv)

    # 最終的なaccuracyスコアを平均値で出力
    print('')
    print('################################')
    print('CV_score:'+ str(np.mean(scores)))
    print()
    
    return(scores, pred_cv)


# 各モード毎のスコアを初期化
total_scores = []

# 各モードの予測を実行
results = predict(train_x_regular, target_regular, test_regular)
total_scores.append(results[0])
test_regular_id['y'] = results[1]

results = predict(train_x_gachi,   target_gachi,   test_gachi)
total_scores.append(results[0])
test_gachi_id['y'] = results[1]

# 最終的なaccuracyスコアを出力
print('---------------------------------')
print('total CV_score:'+ str(np.mean(total_scores)))

Fold : 0
[I 2020-09-27 04:38:19,970] A new study created in memory with name: no-name-4ce5d708-0a14-4cfc-bc23-b74b1ae8bc8d
feature_fraction, val_score: inf:   0%|          | 0/7 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000784 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.574988	valid's binary_logloss: 0.650038
[200]	train's binary_logloss: 0.51996	valid's binary_logloss: 0.651275
Early stopping, best iteration is:
[137]	train's binary_logloss: 0.552891	valid's binary_logloss: 0.649537
feature_fraction, val_score: 0.649537:  14%|#4        | 1/7 [00:00<00:04,  1.41it/s][I 2020-09-27 04:38:20,696] Trial 0 finished with value: 0.6495366523489564 and parameters: {'feature_fraction': 0.7}. Best is trial 0 with value: 0.6495366523489564.
feature_fraction, val_score: 0.649537:  14%|#4        | 1/7 [00:00<00:04,  1.41it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000262 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579394	valid's binary_logloss: 0.650899
Early stopping, best iteration is:
[92]	train's binary_logloss: 0.584248	valid's binary_logloss: 0.650087
feature_fraction, val_score: 0.649537:  29%|##8       | 2/7 [00:01<00:03,  1.58it/s][I 2020-09-27 04:38:21,152] Trial 1 finished with value: 0.6500867396462259 and parameters: {'feature_fraction': 0.5}. Best is trial 0 with value: 0.6495366523489564.
feature_fraction, val_score: 0.649537:  29%|##8       | 2/7 [00:01<00:03,  1.58it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000583 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.583335	valid's binary_logloss: 0.652707
Early stopping, best iteration is:
[88]	train's binary_logloss: 0.590815	valid's binary_logloss: 0.652179
feature_fraction, val_score: 0.649537:  43%|####2     | 3/7 [00:01<00:02,  1.77it/s][I 2020-09-27 04:38:21,556] Trial 2 finished with value: 0.652179314698687 and parameters: {'feature_fraction': 0.4}. Best is trial 0 with value: 0.6495366523489564.
feature_fraction, val_score: 0.649537:  43%|####2     | 3/7 [00:01<00:02,  1.77it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000873 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.573386	valid's binary_logloss: 0.65332
Early stopping, best iteration is:
[72]	train's binary_logloss: 0.592163	valid's binary_logloss: 0.652672
feature_fraction, val_score: 0.649537:  57%|#####7    | 4/7 [00:02<00:01,  1.90it/s][I 2020-09-27 04:38:21,994] Trial 3 finished with value: 0.6526718623826693 and parameters: {'feature_fraction': 0.8}. Best is trial 0 with value: 0.6495366523489564.
feature_fraction, val_score: 0.649537:  57%|#####7    | 4/7 [00:02<00:01,  1.90it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000825 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.573309	valid's binary_logloss: 0.653575
Early stopping, best iteration is:
[67]	train's binary_logloss: 0.59526	valid's binary_logloss: 0.652553
feature_fraction, val_score: 0.649537:  71%|#######1  | 5/7 [00:02<00:01,  1.58it/s][I 2020-09-27 04:38:22,880] Trial 4 finished with value: 0.6525525222524156 and parameters: {'feature_fraction': 0.8999999999999999}. Best is trial 0 with value: 0.6495366523489564.
feature_fraction, val_score: 0.649537:  71%|#######1  | 5/7 [00:02<00:01,  1.58it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.006985 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.576836	valid's binary_logloss: 0.653247
Early stopping, best iteration is:
[70]	train's binary_logloss: 0.596318	valid's binary_logloss: 0.651989
feature_fraction, val_score: 0.649537:  86%|########5 | 6/7 [00:03<00:00,  1.79it/s][I 2020-09-27 04:38:23,260] Trial 5 finished with value: 0.6519889736642206 and parameters: {'feature_fraction': 0.6}. Best is trial 0 with value: 0.6495366523489564.
feature_fraction, val_score: 0.649537:  86%|########5 | 6/7 [00:03<00:00,  1.79it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000890 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.571148	valid's binary_logloss: 0.654809
Early stopping, best iteration is:
[46]	train's binary_logloss: 0.611505	valid's binary_logloss: 0.653731
feature_fraction, val_score: 0.649537: 100%|##########| 7/7 [00:03<00:00,  1.96it/s][I 2020-09-27 04:38:23,655] Trial 6 finished with value: 0.6537305766184812 and parameters: {'feature_fraction': 1.0}. Best is trial 0 with value: 0.6495366523489564.
feature_fraction, val_score: 0.649537: 100%|##########| 7/7 [00:03<00:00,  1.91it/s]
num_leaves, val_score: 0.649537:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000620 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.57822	valid's binary_logloss: 0.649469
Early stopping, best iteration is:
[85]	train's binary_logloss: 0.587383	valid's binary_logloss: 0.649227
num_leaves, val_score: 0.649227:   5%|5         | 1/20 [00:00<00:09,  2.10it/s][I 2020-09-27 04:38:24,142] Trial 7 finished with value: 0.6492270930479551 and parameters: {'num_leaves': 30}. Best is trial 7 with value: 0.6492270930479551.
num_leaves, val_score: 0.649227:   5%|5         | 1/20 [00:00<00:09,  2.10it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000836 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.514021	valid's binary_logloss: 0.655753
Early stopping, best iteration is:
[46]	train's binary_logloss: 0.579148	valid's binary_logloss: 0.653741
num_leaves, val_score: 0.649227:  10%|#         | 2/20 [00:01<00:08,  2.04it/s][I 2020-09-27 04:38:24,666] Trial 8 finished with value: 0.6537410492395737 and parameters: {'num_leaves': 60}. Best is trial 7 with value: 0.6492270930479551.
num_leaves, val_score: 0.649227:  10%|#         | 2/20 [00:01<00:08,  2.04it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.003688 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635569	valid's binary_logloss: 0.650064
[200]	train's binary_logloss: 0.617663	valid's binary_logloss: 0.647802
[300]	train's binary_logloss: 0.603448	valid's binary_logloss: 0.648074
Early stopping, best iteration is:
[204]	train's binary_logloss: 0.617119	valid's binary_logloss: 0.647726
num_leaves, val_score: 0.647726:  15%|#5        | 3/20 [00:01<00:08,  2.08it/s][I 2020-09-27 04:38:25,128] Trial 9 finished with value: 0.6477259914776615 and parameters: {'num_leaves': 8}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  15%|#5        | 3/20 [00:01<00:08,  2.08it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.003613 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.295966	valid's binary_logloss: 0.669876
Early stopping, best iteration is:
[28]	train's binary_logloss: 0.505319	valid's binary_logloss: 0.658492
num_leaves, val_score: 0.647726:  20%|##        | 4/20 [00:02<00:11,  1.33it/s][I 2020-09-27 04:38:26,505] Trial 10 finished with value: 0.6584916550281572 and parameters: {'num_leaves': 229}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  20%|##        | 4/20 [00:02<00:11,  1.33it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.003890 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.614941	valid's binary_logloss: 0.650113
[200]	train's binary_logloss: 0.584546	valid's binary_logloss: 0.649317
[300]	train's binary_logloss: 0.558777	valid's binary_logloss: 0.650116
Early stopping, best iteration is:
[244]	train's binary_logloss: 0.57318	valid's binary_logloss: 0.649088
num_leaves, val_score: 0.647726:  25%|##5       | 5/20 [00:03<00:11,  1.29it/s][I 2020-09-27 04:38:27,335] Trial 11 finished with value: 0.6490878119710577 and parameters: {'num_leaves': 15}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  25%|##5       | 5/20 [00:03<00:11,  1.29it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000790 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.646526	valid's binary_logloss: 0.653926
[200]	train's binary_logloss: 0.634088	valid's binary_logloss: 0.649382
[300]	train's binary_logloss: 0.62524	valid's binary_logloss: 0.648291
[400]	train's binary_logloss: 0.617685	valid's binary_logloss: 0.648421
Early stopping, best iteration is:
[324]	train's binary_logloss: 0.623449	valid's binary_logloss: 0.648061
num_leaves, val_score: 0.647726:  30%|###       | 6/20 [00:04<00:10,  1.34it/s][I 2020-09-27 04:38:28,014] Trial 12 finished with value: 0.6480614052342452 and parameters: {'num_leaves': 5}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  30%|###       | 6/20 [00:04<00:10,  1.34it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008792 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.414675	valid's binary_logloss: 0.664247
Early stopping, best iteration is:
[26]	train's binary_logloss: 0.571385	valid's binary_logloss: 0.659552
num_leaves, val_score: 0.647726:  35%|###5      | 7/20 [00:05<00:09,  1.37it/s][I 2020-09-27 04:38:28,706] Trial 13 finished with value: 0.6595522413168151 and parameters: {'num_leaves': 122}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  35%|###5      | 7/20 [00:05<00:09,  1.37it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000826 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.470281	valid's binary_logloss: 0.657387
Early stopping, best iteration is:
[37]	train's binary_logloss: 0.570263	valid's binary_logloss: 0.654921
num_leaves, val_score: 0.647726:  40%|####      | 8/20 [00:05<00:08,  1.36it/s][I 2020-09-27 04:38:29,454] Trial 14 finished with value: 0.654921475007968 and parameters: {'num_leaves': 85}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  40%|####      | 8/20 [00:05<00:08,  1.36it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004565 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635569	valid's binary_logloss: 0.650064
[200]	train's binary_logloss: 0.617663	valid's binary_logloss: 0.647802
[300]	train's binary_logloss: 0.603448	valid's binary_logloss: 0.648074
Early stopping, best iteration is:
[204]	train's binary_logloss: 0.617119	valid's binary_logloss: 0.647726
num_leaves, val_score: 0.647726:  45%|####5     | 9/20 [00:06<00:08,  1.24it/s][I 2020-09-27 04:38:30,428] Trial 15 finished with value: 0.6477259914776615 and parameters: {'num_leaves': 8}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  45%|####5     | 9/20 [00:06<00:08,  1.24it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000847 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.291012	valid's binary_logloss: 0.673269
Early stopping, best iteration is:
[36]	train's binary_logloss: 0.469537	valid's binary_logloss: 0.661474
num_leaves, val_score: 0.647726:  50%|#####     | 10/20 [00:07<00:09,  1.07it/s][I 2020-09-27 04:38:31,655] Trial 16 finished with value: 0.661474106398792 and parameters: {'num_leaves': 233}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  50%|#####     | 10/20 [00:07<00:09,  1.07it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.007106 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.35312	valid's binary_logloss: 0.66936
Early stopping, best iteration is:
[30]	train's binary_logloss: 0.529202	valid's binary_logloss: 0.659327
num_leaves, val_score: 0.647726:  55%|#####5    | 11/20 [00:08<00:08,  1.12it/s][I 2020-09-27 04:38:32,457] Trial 17 finished with value: 0.6593265762527931 and parameters: {'num_leaves': 171}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  55%|#####5    | 11/20 [00:08<00:08,  1.12it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.003451 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.516771	valid's binary_logloss: 0.654055
Early stopping, best iteration is:
[73]	train's binary_logloss: 0.545672	valid's binary_logloss: 0.65329
num_leaves, val_score: 0.647726:  60%|######    | 12/20 [00:09<00:06,  1.25it/s][I 2020-09-27 04:38:33,041] Trial 18 finished with value: 0.6532896945252986 and parameters: {'num_leaves': 59}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  60%|######    | 12/20 [00:09<00:06,  1.25it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000421 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.642488	valid's binary_logloss: 0.652362
[200]	train's binary_logloss: 0.628687	valid's binary_logloss: 0.649023
[300]	train's binary_logloss: 0.617877	valid's binary_logloss: 0.649151
Early stopping, best iteration is:
[279]	train's binary_logloss: 0.620188	valid's binary_logloss: 0.648624
num_leaves, val_score: 0.647726:  65%|######5   | 13/20 [00:10<00:05,  1.31it/s][I 2020-09-27 04:38:33,727] Trial 19 finished with value: 0.6486244754568631 and parameters: {'num_leaves': 6}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  65%|######5   | 13/20 [00:10<00:05,  1.31it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.009248 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.423908	valid's binary_logloss: 0.659234
Early stopping, best iteration is:
[44]	train's binary_logloss: 0.527404	valid's binary_logloss: 0.655427
num_leaves, val_score: 0.647726:  70%|#######   | 14/20 [00:11<00:05,  1.07it/s][I 2020-09-27 04:38:35,066] Trial 20 finished with value: 0.6554267059223545 and parameters: {'num_leaves': 116}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  70%|#######   | 14/20 [00:11<00:05,  1.07it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004318 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.669306	valid's binary_logloss: 0.670416
[200]	train's binary_logloss: 0.659732	valid's binary_logloss: 0.660321
[300]	train's binary_logloss: 0.654518	valid's binary_logloss: 0.655319
[400]	train's binary_logloss: 0.651377	valid's binary_logloss: 0.652506
[500]	train's binary_logloss: 0.649375	valid's binary_logloss: 0.650671
[600]	train's binary_logloss: 0.648029	valid's binary_logloss: 0.649863
[700]	train's binary_logloss: 0.647077	valid's binary_logloss: 0.649248
[800]	train's binary_logloss: 0.64636	valid's binary_logloss: 0.649015
[900]	train's binary_logloss: 0.645776	valid's binary_logloss: 0.649025
[1000]	train's binary_logloss: 0.645301	valid's binary_logloss: 0.649111
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.645301	valid's binary_logloss: 0.649111
num_leaves, val_score: 0.647726:  75%|#######5  | 15/20 [00:12<00:05,  1.02s/it][I 2020-09-27 04:38:36,284] Trial 21 finished with value: 0.649110563220017 and parameters: {'num_leaves': 2}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  75%|#######5  | 15/20 [00:12<00:05,  1.02s/it][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.011497 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.560366	valid's binary_logloss: 0.653399
[200]	train's binary_logloss: 0.496525	valid's binary_logloss: 0.653927
Early stopping, best iteration is:
[123]	train's binary_logloss: 0.544369	valid's binary_logloss: 0.65141
num_leaves, val_score: 0.647726:  80%|########  | 16/20 [00:13<00:03,  1.10it/s][I 2020-09-27 04:38:36,924] Trial 22 finished with value: 0.6514101116698465 and parameters: {'num_leaves': 38}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  80%|########  | 16/20 [00:13<00:03,  1.10it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005527 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.669306	valid's binary_logloss: 0.670416
[200]	train's binary_logloss: 0.659732	valid's binary_logloss: 0.660321
[300]	train's binary_logloss: 0.654518	valid's binary_logloss: 0.655319
[400]	train's binary_logloss: 0.651377	valid's binary_logloss: 0.652506
[500]	train's binary_logloss: 0.649375	valid's binary_logloss: 0.650671
[600]	train's binary_logloss: 0.648029	valid's binary_logloss: 0.649863
[700]	train's binary_logloss: 0.647077	valid's binary_logloss: 0.649248
[800]	train's binary_logloss: 0.64636	valid's binary_logloss: 0.649015
[900]	train's binary_logloss: 0.645776	valid's binary_logloss: 0.649025
[1000]	train's binary_logloss: 0.645301	valid's binary_logloss: 0.649111
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.645301	valid's binary_logloss: 0.649111
num_leaves, val_score: 0.647726:  85%|########5 | 17/20 [00:14<00:03,  1.12s/it][I 2020-09-27 04:38:38,549] Trial 23 finished with value: 0.6491105632200171 and parameters: {'num_leaves': 2}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  85%|########5 | 17/20 [00:14<00:03,  1.12s/it][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005133 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.53935	valid's binary_logloss: 0.652897
Early stopping, best iteration is:
[92]	train's binary_logloss: 0.546487	valid's binary_logloss: 0.652502
num_leaves, val_score: 0.647726:  90%|######### | 18/20 [00:15<00:01,  1.04it/s][I 2020-09-27 04:38:39,143] Trial 24 finished with value: 0.6525017174749979 and parameters: {'num_leaves': 48}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  90%|######### | 18/20 [00:15<00:01,  1.04it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.003834 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.669306	valid's binary_logloss: 0.670416
[200]	train's binary_logloss: 0.659732	valid's binary_logloss: 0.660321
[300]	train's binary_logloss: 0.654518	valid's binary_logloss: 0.655319
[400]	train's binary_logloss: 0.651377	valid's binary_logloss: 0.652506
[500]	train's binary_logloss: 0.649375	valid's binary_logloss: 0.650671
[600]	train's binary_logloss: 0.648029	valid's binary_logloss: 0.649863
[700]	train's binary_logloss: 0.647077	valid's binary_logloss: 0.649248
[800]	train's binary_logloss: 0.64636	valid's binary_logloss: 0.649015
[900]	train's binary_logloss: 0.645776	valid's binary_logloss: 0.649025
[1000]	train's binary_logloss: 0.645301	valid's binary_logloss: 0.649111
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.645301	valid's binary_logloss: 0.649111
num_leaves, val_score: 0.647726:  95%|#########5| 19/20 [00:16<00:01,  1.02s/it][I 2020-09-27 04:38:40,291] Trial 25 finished with value: 0.649110563220017 and parameters: {'num_leaves': 2}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726:  95%|#########5| 19/20 [00:16<00:01,  1.02s/it][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000543 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.468108	valid's binary_logloss: 0.654681
Early stopping, best iteration is:
[69]	train's binary_logloss: 0.512521	valid's binary_logloss: 0.652052
num_leaves, val_score: 0.647726: 100%|##########| 20/20 [00:17<00:00,  1.05it/s][I 2020-09-27 04:38:41,075] Trial 26 finished with value: 0.652051918071773 and parameters: {'num_leaves': 86}. Best is trial 9 with value: 0.6477259914776615.
num_leaves, val_score: 0.647726: 100%|##########| 20/20 [00:17<00:00,  1.15it/s]
bagging, val_score: 0.647726:   0%|          | 0/10 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.003535 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.63651	valid's binary_logloss: 0.650977
[200]	train's binary_logloss: 0.620185	valid's binary_logloss: 0.652728
Early stopping, best iteration is:
[150]	train's binary_logloss: 0.627724	valid's binary_logloss: 0.650131
bagging, val_score: 0.647726:  10%|#         | 1/10 [00:00<00:04,  1.99it/s][I 2020-09-27 04:38:41,600] Trial 27 finished with value: 0.650131071613353 and parameters: {'bagging_fraction': 0.4586199697400575, 'bagging_freq': 3}. Best is trial 27 with value: 0.650131071613353.
bagging, val_score: 0.647726:  10%|#         | 1/10 [00:00<00:04,  1.99it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000853 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635382	valid's binary_logloss: 0.651145
[200]	train's binary_logloss: 0.61788	valid's binary_logloss: 0.647841
[300]	train's binary_logloss: 0.603222	valid's binary_logloss: 0.647911
[400]	train's binary_logloss: 0.589945	valid's binary_logloss: 0.647612
Early stopping, best iteration is:
[340]	train's binary_logloss: 0.597917	valid's binary_logloss: 0.647006
bagging, val_score: 0.647006:  20%|##        | 2/10 [00:01<00:05,  1.33it/s][I 2020-09-27 04:38:42,919] Trial 28 finished with value: 0.6470063208679462 and parameters: {'bagging_fraction': 0.9958662267298061, 'bagging_freq': 7}. Best is trial 28 with value: 0.6470063208679462.
bagging, val_score: 0.647006:  20%|##        | 2/10 [00:01<00:05,  1.33it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.003784 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635259	valid's binary_logloss: 0.651142
[200]	train's binary_logloss: 0.617565	valid's binary_logloss: 0.64962
Early stopping, best iteration is:
[195]	train's binary_logloss: 0.618361	valid's binary_logloss: 0.649343
bagging, val_score: 0.647006:  30%|###       | 3/10 [00:02<00:04,  1.45it/s][I 2020-09-27 04:38:43,462] Trial 29 finished with value: 0.6493431744945917 and parameters: {'bagging_fraction': 0.9799916779746923, 'bagging_freq': 7}. Best is trial 28 with value: 0.6470063208679462.
bagging, val_score: 0.647006:  30%|###       | 3/10 [00:02<00:04,  1.45it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000779 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635241	valid's binary_logloss: 0.651269
[200]	train's binary_logloss: 0.618036	valid's binary_logloss: 0.649672
[300]	train's binary_logloss: 0.603932	valid's binary_logloss: 0.649408
[400]	train's binary_logloss: 0.590656	valid's binary_logloss: 0.648685
Early stopping, best iteration is:
[368]	train's binary_logloss: 0.594891	valid's binary_logloss: 0.648347
bagging, val_score: 0.647006:  40%|####      | 4/10 [00:03<00:04,  1.37it/s][I 2020-09-27 04:38:44,294] Trial 30 finished with value: 0.6483471798474496 and parameters: {'bagging_fraction': 0.9986614103818544, 'bagging_freq': 7}. Best is trial 28 with value: 0.6470063208679462.
bagging, val_score: 0.647006:  40%|####      | 4/10 [00:03<00:04,  1.37it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.007385 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.63573	valid's binary_logloss: 0.649902
[200]	train's binary_logloss: 0.618646	valid's binary_logloss: 0.650492
Early stopping, best iteration is:
[170]	train's binary_logloss: 0.623387	valid's binary_logloss: 0.648591
bagging, val_score: 0.647006:  50%|#####     | 5/10 [00:03<00:03,  1.52it/s][I 2020-09-27 04:38:44,781] Trial 31 finished with value: 0.6485905878341305 and parameters: {'bagging_fraction': 0.7328660882461872, 'bagging_freq': 5}. Best is trial 28 with value: 0.6470063208679462.
bagging, val_score: 0.647006:  50%|#####     | 5/10 [00:03<00:03,  1.52it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000825 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635272	valid's binary_logloss: 0.650422
[200]	train's binary_logloss: 0.61747	valid's binary_logloss: 0.648941
[300]	train's binary_logloss: 0.602462	valid's binary_logloss: 0.650582
Early stopping, best iteration is:
[227]	train's binary_logloss: 0.613139	valid's binary_logloss: 0.648451
bagging, val_score: 0.647006:  60%|######    | 6/10 [00:04<00:02,  1.49it/s][I 2020-09-27 04:38:45,485] Trial 32 finished with value: 0.6484510872456424 and parameters: {'bagging_fraction': 0.7900181913598195, 'bagging_freq': 1}. Best is trial 28 with value: 0.6470063208679462.
bagging, val_score: 0.647006:  60%|######    | 6/10 [00:04<00:02,  1.49it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.003362 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.636504	valid's binary_logloss: 0.650927
[200]	train's binary_logloss: 0.620547	valid's binary_logloss: 0.651909
Early stopping, best iteration is:
[115]	train's binary_logloss: 0.633633	valid's binary_logloss: 0.6507
bagging, val_score: 0.647006:  70%|#######   | 7/10 [00:05<00:02,  1.40it/s][I 2020-09-27 04:38:46,298] Trial 33 finished with value: 0.6506995210705391 and parameters: {'bagging_fraction': 0.47267376312879716, 'bagging_freq': 5}. Best is trial 28 with value: 0.6470063208679462.
bagging, val_score: 0.647006:  70%|#######   | 7/10 [00:05<00:02,  1.40it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005528 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635448	valid's binary_logloss: 0.651834
[200]	train's binary_logloss: 0.617764	valid's binary_logloss: 0.650603
[300]	train's binary_logloss: 0.603475	valid's binary_logloss: 0.652326
Early stopping, best iteration is:
[226]	train's binary_logloss: 0.614069	valid's binary_logloss: 0.649886
bagging, val_score: 0.647006:  80%|########  | 8/10 [00:05<00:01,  1.47it/s][I 2020-09-27 04:38:46,894] Trial 34 finished with value: 0.649885537943031 and parameters: {'bagging_fraction': 0.876875549392964, 'bagging_freq': 7}. Best is trial 28 with value: 0.6470063208679462.
bagging, val_score: 0.647006:  80%|########  | 8/10 [00:05<00:01,  1.47it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000544 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635851	valid's binary_logloss: 0.649962
[200]	train's binary_logloss: 0.6186	valid's binary_logloss: 0.647048
[300]	train's binary_logloss: 0.603846	valid's binary_logloss: 0.649943
Early stopping, best iteration is:
[207]	train's binary_logloss: 0.617528	valid's binary_logloss: 0.646928
bagging, val_score: 0.646928:  90%|######### | 9/10 [00:06<00:00,  1.56it/s][I 2020-09-27 04:38:47,444] Trial 35 finished with value: 0.6469279610503994 and parameters: {'bagging_fraction': 0.6178680364111143, 'bagging_freq': 3}. Best is trial 35 with value: 0.6469279610503994.
bagging, val_score: 0.646928:  90%|######### | 9/10 [00:06<00:00,  1.56it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000490 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635909	valid's binary_logloss: 0.648293
[200]	train's binary_logloss: 0.618922	valid's binary_logloss: 0.649973
Early stopping, best iteration is:
[149]	train's binary_logloss: 0.627266	valid's binary_logloss: 0.647531
bagging, val_score: 0.646928: 100%|##########| 10/10 [00:06<00:00,  1.68it/s][I 2020-09-27 04:38:47,929] Trial 36 finished with value: 0.6475307026936266 and parameters: {'bagging_fraction': 0.5956628615183736, 'bagging_freq': 2}. Best is trial 35 with value: 0.6469279610503994.
bagging, val_score: 0.646928: 100%|##########| 10/10 [00:06<00:00,  1.46it/s]
feature_fraction_stage2, val_score: 0.646928:   0%|          | 0/6 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000428 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.634952	valid's binary_logloss: 0.65056
[200]	train's binary_logloss: 0.617682	valid's binary_logloss: 0.649553
[300]	train's binary_logloss: 0.603161	valid's binary_logloss: 0.652262
Early stopping, best iteration is:
[241]	train's binary_logloss: 0.611549	valid's binary_logloss: 0.649321
feature_fraction_stage2, val_score: 0.646928:  17%|#6        | 1/6 [00:00<00:03,  1.60it/s][I 2020-09-27 04:38:48,575] Trial 37 finished with value: 0.6493214800336689 and parameters: {'feature_fraction': 0.748}. Best is trial 37 with value: 0.6493214800336689.
feature_fraction_stage2, val_score: 0.646928:  17%|#6        | 1/6 [00:00<00:03,  1.60it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001333 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635533	valid's binary_logloss: 0.650663
[200]	train's binary_logloss: 0.618659	valid's binary_logloss: 0.651432
Early stopping, best iteration is:
[141]	train's binary_logloss: 0.628172	valid's binary_logloss: 0.650017
feature_fraction_stage2, val_score: 0.646928:  33%|###3      | 2/6 [00:01<00:02,  1.67it/s][I 2020-09-27 04:38:49,111] Trial 38 finished with value: 0.6500171564584806 and parameters: {'feature_fraction': 0.652}. Best is trial 37 with value: 0.6493214800336689.
feature_fraction_stage2, val_score: 0.646928:  33%|###3      | 2/6 [00:01<00:02,  1.67it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000674 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635851	valid's binary_logloss: 0.649962
[200]	train's binary_logloss: 0.6186	valid's binary_logloss: 0.647048
[300]	train's binary_logloss: 0.603846	valid's binary_logloss: 0.649943
Early stopping, best iteration is:
[207]	train's binary_logloss: 0.617528	valid's binary_logloss: 0.646928
feature_fraction_stage2, val_score: 0.646928:  50%|#####     | 3/6 [00:02<00:02,  1.33it/s][I 2020-09-27 04:38:50,213] Trial 39 finished with value: 0.6469279610503994 and parameters: {'feature_fraction': 0.6839999999999999}. Best is trial 39 with value: 0.6469279610503994.
feature_fraction_stage2, val_score: 0.646928:  50%|#####     | 3/6 [00:02<00:02,  1.33it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000915 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635444	valid's binary_logloss: 0.649405
[200]	train's binary_logloss: 0.618068	valid's binary_logloss: 0.648194
Early stopping, best iteration is:
[173]	train's binary_logloss: 0.622667	valid's binary_logloss: 0.647565
feature_fraction_stage2, val_score: 0.646928:  67%|######6   | 4/6 [00:02<00:01,  1.46it/s][I 2020-09-27 04:38:50,749] Trial 40 finished with value: 0.6475654368255805 and parameters: {'feature_fraction': 0.7799999999999999}. Best is trial 39 with value: 0.6469279610503994.
feature_fraction_stage2, val_score: 0.646928:  67%|######6   | 4/6 [00:02<00:01,  1.46it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004998 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.63624	valid's binary_logloss: 0.650875
[200]	train's binary_logloss: 0.619436	valid's binary_logloss: 0.649352
Early stopping, best iteration is:
[165]	train's binary_logloss: 0.62485	valid's binary_logloss: 0.648991
feature_fraction_stage2, val_score: 0.646928:  83%|########3 | 5/6 [00:03<00:00,  1.59it/s][I 2020-09-27 04:38:51,248] Trial 41 finished with value: 0.6489914374612618 and parameters: {'feature_fraction': 0.62}. Best is trial 39 with value: 0.6469279610503994.
feature_fraction_stage2, val_score: 0.646928:  83%|########3 | 5/6 [00:03<00:00,  1.59it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000866 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.634952	valid's binary_logloss: 0.65056
[200]	train's binary_logloss: 0.617682	valid's binary_logloss: 0.649553
[300]	train's binary_logloss: 0.603161	valid's binary_logloss: 0.652262
Early stopping, best iteration is:
[241]	train's binary_logloss: 0.611549	valid's binary_logloss: 0.649321
feature_fraction_stage2, val_score: 0.646928: 100%|##########| 6/6 [00:03<00:00,  1.56it/s][I 2020-09-27 04:38:51,915] Trial 42 finished with value: 0.6493214800336689 and parameters: {'feature_fraction': 0.716}. Best is trial 39 with value: 0.6469279610503994.
feature_fraction_stage2, val_score: 0.646928: 100%|##########| 6/6 [00:03<00:00,  1.51it/s]
regularization_factors, val_score: 0.646928:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000644 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635945	valid's binary_logloss: 0.65002
[200]	train's binary_logloss: 0.61936	valid's binary_logloss: 0.648569
Early stopping, best iteration is:
[161]	train's binary_logloss: 0.625243	valid's binary_logloss: 0.647758
regularization_factors, val_score: 0.646928:   5%|5         | 1/20 [00:00<00:10,  1.73it/s][I 2020-09-27 04:38:52,510] Trial 43 finished with value: 0.6477575542610232 and parameters: {'lambda_l1': 0.5305532619182826, 'lambda_l2': 2.2940234848178292e-05}. Best is trial 43 with value: 0.6477575542610232.
regularization_factors, val_score: 0.646928:   5%|5         | 1/20 [00:00<00:10,  1.73it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004925 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637159	valid's binary_logloss: 0.648561
[200]	train's binary_logloss: 0.62222	valid's binary_logloss: 0.646184
[300]	train's binary_logloss: 0.610053	valid's binary_logloss: 0.646865
Early stopping, best iteration is:
[250]	train's binary_logloss: 0.616221	valid's binary_logloss: 0.645506
regularization_factors, val_score: 0.645506:  10%|#         | 2/20 [00:01<00:12,  1.48it/s][I 2020-09-27 04:38:53,419] Trial 44 finished with value: 0.6455057317841788 and parameters: {'lambda_l1': 1.5729907152074253e-08, 'lambda_l2': 6.408315304180313}. Best is trial 44 with value: 0.6455057317841788.
regularization_factors, val_score: 0.645506:  10%|#         | 2/20 [00:01<00:12,  1.48it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010288 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637347	valid's binary_logloss: 0.64928
[200]	train's binary_logloss: 0.623416	valid's binary_logloss: 0.647003
[300]	train's binary_logloss: 0.611546	valid's binary_logloss: 0.648015
Early stopping, best iteration is:
[225]	train's binary_logloss: 0.620614	valid's binary_logloss: 0.646489
regularization_factors, val_score: 0.645506:  15%|#5        | 3/20 [00:02<00:12,  1.34it/s][I 2020-09-27 04:38:54,330] Trial 45 finished with value: 0.6464889390748203 and parameters: {'lambda_l1': 1.792278679780671e-08, 'lambda_l2': 8.810233526119564}. Best is trial 44 with value: 0.6455057317841788.
regularization_factors, val_score: 0.645506:  15%|#5        | 3/20 [00:02<00:12,  1.34it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000443 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.63666	valid's binary_logloss: 0.649464
[200]	train's binary_logloss: 0.62174	valid's binary_logloss: 0.647351
[300]	train's binary_logloss: 0.609288	valid's binary_logloss: 0.648443
Early stopping, best iteration is:
[254]	train's binary_logloss: 0.615011	valid's binary_logloss: 0.646435
regularization_factors, val_score: 0.645506:  20%|##        | 4/20 [00:03<00:11,  1.37it/s][I 2020-09-27 04:38:55,027] Trial 46 finished with value: 0.6464351970725228 and parameters: {'lambda_l1': 2.2856982982765707e-08, 'lambda_l2': 4.5894526330871095}. Best is trial 44 with value: 0.6455057317841788.
regularization_factors, val_score: 0.645506:  20%|##        | 4/20 [00:03<00:11,  1.37it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.007799 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637474	valid's binary_logloss: 0.648583
[200]	train's binary_logloss: 0.62337	valid's binary_logloss: 0.645859
[300]	train's binary_logloss: 0.61123	valid's binary_logloss: 0.645293
[400]	train's binary_logloss: 0.599843	valid's binary_logloss: 0.646225
Early stopping, best iteration is:
[317]	train's binary_logloss: 0.609063	valid's binary_logloss: 0.644762
regularization_factors, val_score: 0.644762:  25%|##5       | 5/20 [00:03<00:11,  1.32it/s][I 2020-09-27 04:38:55,839] Trial 47 finished with value: 0.6447624933131433 and parameters: {'lambda_l1': 1.2755111716437187e-08, 'lambda_l2': 9.850714874340618}. Best is trial 47 with value: 0.6447624933131433.
regularization_factors, val_score: 0.644762:  25%|##5       | 5/20 [00:03<00:11,  1.32it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004054 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.63754	valid's binary_logloss: 0.648978
[200]	train's binary_logloss: 0.623228	valid's binary_logloss: 0.646323
[300]	train's binary_logloss: 0.611727	valid's binary_logloss: 0.64615
[400]	train's binary_logloss: 0.600473	valid's binary_logloss: 0.646062
Early stopping, best iteration is:
[347]	train's binary_logloss: 0.606258	valid's binary_logloss: 0.645207
regularization_factors, val_score: 0.644762:  30%|###       | 6/20 [00:04<00:11,  1.26it/s][I 2020-09-27 04:38:56,723] Trial 48 finished with value: 0.6452071554754294 and parameters: {'lambda_l1': 1.3011629397041354e-08, 'lambda_l2': 9.988192496777456}. Best is trial 47 with value: 0.6447624933131433.
regularization_factors, val_score: 0.644762:  30%|###       | 6/20 [00:04<00:11,  1.26it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.007536 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.6371	valid's binary_logloss: 0.6496
[200]	train's binary_logloss: 0.622819	valid's binary_logloss: 0.647511
[300]	train's binary_logloss: 0.610769	valid's binary_logloss: 0.647971
Early stopping, best iteration is:
[232]	train's binary_logloss: 0.618821	valid's binary_logloss: 0.646619
regularization_factors, val_score: 0.644762:  35%|###5      | 7/20 [00:05<00:11,  1.14it/s][I 2020-09-27 04:38:57,789] Trial 49 finished with value: 0.6466188865283253 and parameters: {'lambda_l1': 1.5462932963049097e-08, 'lambda_l2': 8.71269927175754}. Best is trial 47 with value: 0.6447624933131433.
regularization_factors, val_score: 0.644762:  35%|###5      | 7/20 [00:05<00:11,  1.14it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004606 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637133	valid's binary_logloss: 0.649344
[200]	train's binary_logloss: 0.622572	valid's binary_logloss: 0.647963
[300]	train's binary_logloss: 0.61055	valid's binary_logloss: 0.647491
Early stopping, best iteration is:
[255]	train's binary_logloss: 0.616042	valid's binary_logloss: 0.646564
regularization_factors, val_score: 0.644762:  40%|####      | 8/20 [00:06<00:09,  1.21it/s][I 2020-09-27 04:38:58,493] Trial 50 finished with value: 0.6465642008648574 and parameters: {'lambda_l1': 1.1089576217805702e-08, 'lambda_l2': 7.312599056585437}. Best is trial 47 with value: 0.6447624933131433.
regularization_factors, val_score: 0.644762:  40%|####      | 8/20 [00:06<00:09,  1.21it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000434 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637479	valid's binary_logloss: 0.648584
[200]	train's binary_logloss: 0.62341	valid's binary_logloss: 0.645728
[300]	train's binary_logloss: 0.611552	valid's binary_logloss: 0.64603
Early stopping, best iteration is:
[254]	train's binary_logloss: 0.617107	valid's binary_logloss: 0.644789
regularization_factors, val_score: 0.644762:  45%|####5     | 9/20 [00:07<00:08,  1.27it/s][I 2020-09-27 04:38:59,196] Trial 51 finished with value: 0.6447887069638869 and parameters: {'lambda_l1': 1.7016389032742813e-08, 'lambda_l2': 9.898211761649065}. Best is trial 47 with value: 0.6447624933131433.
regularization_factors, val_score: 0.644762:  45%|####5     | 9/20 [00:07<00:08,  1.27it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.007899 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637278	valid's binary_logloss: 0.649089
[200]	train's binary_logloss: 0.622576	valid's binary_logloss: 0.647208
[300]	train's binary_logloss: 0.610418	valid's binary_logloss: 0.648157
Early stopping, best iteration is:
[255]	train's binary_logloss: 0.615802	valid's binary_logloss: 0.646398
regularization_factors, val_score: 0.644762:  50%|#####     | 10/20 [00:07<00:07,  1.31it/s][I 2020-09-27 04:38:59,894] Trial 52 finished with value: 0.6463976805891822 and parameters: {'lambda_l1': 2.2240739211030176e-08, 'lambda_l2': 7.404538655662823}. Best is trial 47 with value: 0.6447624933131433.
regularization_factors, val_score: 0.644762:  50%|#####     | 10/20 [00:07<00:07,  1.31it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000572 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637182	valid's binary_logloss: 0.647682
[200]	train's binary_logloss: 0.623009	valid's binary_logloss: 0.646084
[300]	train's binary_logloss: 0.611035	valid's binary_logloss: 0.645152
Early stopping, best iteration is:
[260]	train's binary_logloss: 0.615822	valid's binary_logloss: 0.643798
regularization_factors, val_score: 0.643798:  55%|#####5    | 11/20 [00:08<00:06,  1.35it/s][I 2020-09-27 04:39:00,581] Trial 53 finished with value: 0.6437977363337001 and parameters: {'lambda_l1': 2.039143182350295e-08, 'lambda_l2': 9.400199053113356}. Best is trial 53 with value: 0.6437977363337001.
regularization_factors, val_score: 0.643798:  55%|#####5    | 11/20 [00:08<00:06,  1.35it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004032 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637206	valid's binary_logloss: 0.647684
[200]	train's binary_logloss: 0.623151	valid's binary_logloss: 0.646089
[300]	train's binary_logloss: 0.611403	valid's binary_logloss: 0.645967
Early stopping, best iteration is:
[255]	train's binary_logloss: 0.616582	valid's binary_logloss: 0.644531
regularization_factors, val_score: 0.643798:  60%|######    | 12/20 [00:09<00:06,  1.17it/s][I 2020-09-27 04:39:01,697] Trial 54 finished with value: 0.6445311936163728 and parameters: {'lambda_l1': 1.6270643281831753e-08, 'lambda_l2': 9.628307902916124}. Best is trial 53 with value: 0.6437977363337001.
regularization_factors, val_score: 0.643798:  60%|######    | 12/20 [00:09<00:06,  1.17it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000874 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637281	valid's binary_logloss: 0.649044
[200]	train's binary_logloss: 0.622584	valid's binary_logloss: 0.647021
[300]	train's binary_logloss: 0.610763	valid's binary_logloss: 0.647525
Early stopping, best iteration is:
[254]	train's binary_logloss: 0.616357	valid's binary_logloss: 0.645998
regularization_factors, val_score: 0.643798:  65%|######5   | 13/20 [00:10<00:05,  1.22it/s][I 2020-09-27 04:39:02,449] Trial 55 finished with value: 0.6459984456397755 and parameters: {'lambda_l1': 1.0240697224966498e-08, 'lambda_l2': 8.012028500403863}. Best is trial 53 with value: 0.6437977363337001.
regularization_factors, val_score: 0.643798:  65%|######5   | 13/20 [00:10<00:05,  1.22it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000558 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635858	valid's binary_logloss: 0.649961
[200]	train's binary_logloss: 0.618625	valid's binary_logloss: 0.647047
[300]	train's binary_logloss: 0.603932	valid's binary_logloss: 0.648748
Early stopping, best iteration is:
[207]	train's binary_logloss: 0.617554	valid's binary_logloss: 0.646927
regularization_factors, val_score: 0.643798:  70%|#######   | 14/20 [00:11<00:04,  1.32it/s][I 2020-09-27 04:39:03,059] Trial 56 finished with value: 0.6469267717887545 and parameters: {'lambda_l1': 1.2959832015054276e-06, 'lambda_l2': 0.01861979600536283}. Best is trial 53 with value: 0.6437977363337001.
regularization_factors, val_score: 0.643798:  70%|#######   | 14/20 [00:11<00:04,  1.32it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010899 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635708	valid's binary_logloss: 0.649563
[200]	train's binary_logloss: 0.61871	valid's binary_logloss: 0.647652
[300]	train's binary_logloss: 0.604402	valid's binary_logloss: 0.650241
Early stopping, best iteration is:
[206]	train's binary_logloss: 0.6179	valid's binary_logloss: 0.647417
regularization_factors, val_score: 0.643798:  75%|#######5  | 15/20 [00:11<00:03,  1.43it/s][I 2020-09-27 04:39:03,628] Trial 57 finished with value: 0.6474166124124395 and parameters: {'lambda_l1': 5.765532893767189e-07, 'lambda_l2': 0.14788147344128014}. Best is trial 53 with value: 0.6437977363337001.
regularization_factors, val_score: 0.643798:  75%|#######5  | 15/20 [00:11<00:03,  1.43it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005282 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635911	valid's binary_logloss: 0.650092
[200]	train's binary_logloss: 0.618697	valid's binary_logloss: 0.649148
[300]	train's binary_logloss: 0.604075	valid's binary_logloss: 0.650674
Early stopping, best iteration is:
[250]	train's binary_logloss: 0.611662	valid's binary_logloss: 0.648499
regularization_factors, val_score: 0.643798:  80%|########  | 16/20 [00:12<00:02,  1.44it/s][I 2020-09-27 04:39:04,302] Trial 58 finished with value: 0.6484991507925789 and parameters: {'lambda_l1': 5.203220654613759e-07, 'lambda_l2': 0.30810693370096226}. Best is trial 53 with value: 0.6437977363337001.
regularization_factors, val_score: 0.643798:  80%|########  | 16/20 [00:12<00:02,  1.44it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004372 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.63618	valid's binary_logloss: 0.650527
[200]	train's binary_logloss: 0.619801	valid's binary_logloss: 0.649123
Early stopping, best iteration is:
[160]	train's binary_logloss: 0.625979	valid's binary_logloss: 0.648367
regularization_factors, val_score: 0.643798:  85%|########5 | 17/20 [00:13<00:02,  1.37it/s][I 2020-09-27 04:39:05,120] Trial 59 finished with value: 0.6483666505535219 and parameters: {'lambda_l1': 1.1681262794154124e-07, 'lambda_l2': 0.9991529121249367}. Best is trial 53 with value: 0.6437977363337001.
regularization_factors, val_score: 0.643798:  85%|########5 | 17/20 [00:13<00:02,  1.37it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008489 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635851	valid's binary_logloss: 0.649962
[200]	train's binary_logloss: 0.6186	valid's binary_logloss: 0.647048
[300]	train's binary_logloss: 0.603848	valid's binary_logloss: 0.649943
Early stopping, best iteration is:
[207]	train's binary_logloss: 0.617528	valid's binary_logloss: 0.646928
regularization_factors, val_score: 0.643798:  90%|######### | 18/20 [00:13<00:01,  1.40it/s][I 2020-09-27 04:39:05,803] Trial 60 finished with value: 0.6469279610385475 and parameters: {'lambda_l1': 1.213400855943739e-08, 'lambda_l2': 1.534183918800275e-07}. Best is trial 53 with value: 0.6437977363337001.
regularization_factors, val_score: 0.643798:  90%|######### | 18/20 [00:13<00:01,  1.40it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.009995 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637219	valid's binary_logloss: 0.650017
[200]	train's binary_logloss: 0.622187	valid's binary_logloss: 0.647211
[300]	train's binary_logloss: 0.609711	valid's binary_logloss: 0.647263
Early stopping, best iteration is:
[254]	train's binary_logloss: 0.615481	valid's binary_logloss: 0.64613
regularization_factors, val_score: 0.643798:  95%|#########5| 19/20 [00:14<00:00,  1.44it/s][I 2020-09-27 04:39:06,446] Trial 61 finished with value: 0.6461303865450015 and parameters: {'lambda_l1': 1.0232750932997174e-08, 'lambda_l2': 5.61104076150738}. Best is trial 53 with value: 0.6437977363337001.
regularization_factors, val_score: 0.643798:  95%|#########5| 19/20 [00:14<00:00,  1.44it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000915 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637448	valid's binary_logloss: 0.649065
[200]	train's binary_logloss: 0.623392	valid's binary_logloss: 0.647489
[300]	train's binary_logloss: 0.611817	valid's binary_logloss: 0.647888
Early stopping, best iteration is:
[236]	train's binary_logloss: 0.61913	valid's binary_logloss: 0.647044
regularization_factors, val_score: 0.643798: 100%|##########| 20/20 [00:15<00:00,  1.48it/s][I 2020-09-27 04:39:07,086] Trial 62 finished with value: 0.6470438910627877 and parameters: {'lambda_l1': 1.4832304242797658e-08, 'lambda_l2': 9.997088293767975}. Best is trial 53 with value: 0.6437977363337001.
regularization_factors, val_score: 0.643798: 100%|##########| 20/20 [00:15<00:00,  1.32it/s]
min_data_in_leaf, val_score: 0.643798:   0%|          | 0/5 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.003521 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637289	valid's binary_logloss: 0.648183
[200]	train's binary_logloss: 0.62309	valid's binary_logloss: 0.64706
[300]	train's binary_logloss: 0.611616	valid's binary_logloss: 0.64692
Early stopping, best iteration is:
[254]	train's binary_logloss: 0.616891	valid's binary_logloss: 0.645943
min_data_in_leaf, val_score: 0.643798:  20%|##        | 1/5 [00:00<00:02,  1.35it/s][I 2020-09-27 04:39:07,843] Trial 63 finished with value: 0.6459430156769583 and parameters: {'min_child_samples': 50}. Best is trial 63 with value: 0.6459430156769583.
min_data_in_leaf, val_score: 0.643798:  20%|##        | 1/5 [00:00<00:02,  1.35it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000832 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637182	valid's binary_logloss: 0.647682
[200]	train's binary_logloss: 0.622869	valid's binary_logloss: 0.646623
[300]	train's binary_logloss: 0.611124	valid's binary_logloss: 0.646202
Early stopping, best iteration is:
[254]	train's binary_logloss: 0.616639	valid's binary_logloss: 0.645162
min_data_in_leaf, val_score: 0.643798:  40%|####      | 2/5 [00:01<00:02,  1.27it/s][I 2020-09-27 04:39:08,731] Trial 64 finished with value: 0.6451618189346408 and parameters: {'min_child_samples': 10}. Best is trial 64 with value: 0.6451618189346408.
min_data_in_leaf, val_score: 0.643798:  40%|####      | 2/5 [00:01<00:02,  1.27it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012841 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637182	valid's binary_logloss: 0.647682
[200]	train's binary_logloss: 0.623074	valid's binary_logloss: 0.64581
[300]	train's binary_logloss: 0.61124	valid's binary_logloss: 0.645222
Early stopping, best iteration is:
[255]	train's binary_logloss: 0.616649	valid's binary_logloss: 0.643974
min_data_in_leaf, val_score: 0.643798:  60%|######    | 3/5 [00:02<00:01,  1.11it/s][I 2020-09-27 04:39:09,890] Trial 65 finished with value: 0.6439740196693694 and parameters: {'min_child_samples': 25}. Best is trial 65 with value: 0.6439740196693694.
min_data_in_leaf, val_score: 0.643798:  60%|######    | 3/5 [00:02<00:01,  1.11it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004742 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637182	valid's binary_logloss: 0.647682
[200]	train's binary_logloss: 0.622869	valid's binary_logloss: 0.646623
[300]	train's binary_logloss: 0.610951	valid's binary_logloss: 0.646227
Early stopping, best iteration is:
[243]	train's binary_logloss: 0.617796	valid's binary_logloss: 0.645499
min_data_in_leaf, val_score: 0.643798:  80%|########  | 4/5 [00:03<00:00,  1.19it/s][I 2020-09-27 04:39:10,600] Trial 66 finished with value: 0.6454986987713198 and parameters: {'min_child_samples': 5}. Best is trial 65 with value: 0.6439740196693694.
min_data_in_leaf, val_score: 0.643798:  80%|########  | 4/5 [00:03<00:00,  1.19it/s][LightGBM] [Info] Number of positive: 13186, number of negative: 12813
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000441 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.507173 -> initscore=0.028695
[LightGBM] [Info] Start training from score 0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637729	valid's binary_logloss: 0.649258
[200]	train's binary_logloss: 0.62325	valid's binary_logloss: 0.646308
[300]	train's binary_logloss: 0.611957	valid's binary_logloss: 0.647641
Early stopping, best iteration is:
[238]	train's binary_logloss: 0.618817	valid's binary_logloss: 0.645531
min_data_in_leaf, val_score: 0.643798: 100%|##########| 5/5 [00:04<00:00,  1.26it/s][I 2020-09-27 04:39:11,288] Trial 67 finished with value: 0.6455312591688435 and parameters: {'min_child_samples': 100}. Best is trial 65 with value: 0.6439740196693694.
min_data_in_leaf, val_score: 0.643798: 100%|##########| 5/5 [00:04<00:00,  1.19it/s]
Fold : 1
[I 2020-09-27 04:39:11,344] A new study created in memory with name: no-name-8ec748bd-7c24-485d-8dfe-5099f051a9d9
feature_fraction, val_score: inf:   0%|          | 0/7 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001101 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.575663	valid's binary_logloss: 0.661357
Early stopping, best iteration is:
[56]	train's binary_logloss: 0.605835	valid's binary_logloss: 0.659163
feature_fraction, val_score: 0.659163:  14%|#4        | 1/7 [00:00<00:03,  1.89it/s][I 2020-09-27 04:39:11,890] Trial 0 finished with value: 0.6591626166519308 and parameters: {'feature_fraction': 0.7}. Best is trial 0 with value: 0.6591626166519308.
feature_fraction, val_score: 0.659163:  14%|#4        | 1/7 [00:00<00:03,  1.89it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000533 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.577716	valid's binary_logloss: 0.658007
Early stopping, best iteration is:
[72]	train's binary_logloss: 0.596211	valid's binary_logloss: 0.65673
feature_fraction, val_score: 0.656730:  29%|##8       | 2/7 [00:00<00:02,  1.98it/s][I 2020-09-27 04:39:12,340] Trial 1 finished with value: 0.6567303283064773 and parameters: {'feature_fraction': 0.5}. Best is trial 1 with value: 0.6567303283064773.
feature_fraction, val_score: 0.656730:  29%|##8       | 2/7 [00:00<00:02,  1.98it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001310 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.576271	valid's binary_logloss: 0.660374
Early stopping, best iteration is:
[66]	train's binary_logloss: 0.598062	valid's binary_logloss: 0.658574
feature_fraction, val_score: 0.656730:  43%|####2     | 3/7 [00:01<00:02,  1.53it/s][I 2020-09-27 04:39:13,335] Trial 2 finished with value: 0.6585740014120882 and parameters: {'feature_fraction': 0.6}. Best is trial 1 with value: 0.6567303283064773.
feature_fraction, val_score: 0.656730:  43%|####2     | 3/7 [00:01<00:02,  1.53it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000370 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.582075	valid's binary_logloss: 0.6565
Early stopping, best iteration is:
[94]	train's binary_logloss: 0.58569	valid's binary_logloss: 0.655992
feature_fraction, val_score: 0.655992:  57%|#####7    | 4/7 [00:02<00:01,  1.67it/s][I 2020-09-27 04:39:13,802] Trial 3 finished with value: 0.6559924908506093 and parameters: {'feature_fraction': 0.4}. Best is trial 3 with value: 0.6559924908506093.
feature_fraction, val_score: 0.655992:  57%|#####7    | 4/7 [00:02<00:01,  1.67it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000609 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.571094	valid's binary_logloss: 0.661412
Early stopping, best iteration is:
[54]	train's binary_logloss: 0.604656	valid's binary_logloss: 0.657544
feature_fraction, val_score: 0.655992:  71%|#######1  | 5/7 [00:02<00:01,  1.78it/s][I 2020-09-27 04:39:14,285] Trial 4 finished with value: 0.6575438872927727 and parameters: {'feature_fraction': 0.8999999999999999}. Best is trial 3 with value: 0.6559924908506093.
feature_fraction, val_score: 0.655992:  71%|#######1  | 5/7 [00:02<00:01,  1.78it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000923 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.569935	valid's binary_logloss: 0.659342
Early stopping, best iteration is:
[54]	train's binary_logloss: 0.602845	valid's binary_logloss: 0.658926
feature_fraction, val_score: 0.655992:  86%|########5 | 6/7 [00:03<00:00,  1.86it/s][I 2020-09-27 04:39:14,768] Trial 5 finished with value: 0.658926225620644 and parameters: {'feature_fraction': 1.0}. Best is trial 3 with value: 0.6559924908506093.
feature_fraction, val_score: 0.655992:  86%|########5 | 6/7 [00:03<00:00,  1.86it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000465 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.574321	valid's binary_logloss: 0.659912
Early stopping, best iteration is:
[66]	train's binary_logloss: 0.595979	valid's binary_logloss: 0.657826
feature_fraction, val_score: 0.655992: 100%|##########| 7/7 [00:03<00:00,  1.91it/s][I 2020-09-27 04:39:15,259] Trial 6 finished with value: 0.65782592600512 and parameters: {'feature_fraction': 0.8}. Best is trial 3 with value: 0.6559924908506093.
feature_fraction, val_score: 0.655992: 100%|##########| 7/7 [00:03<00:00,  1.79it/s]
num_leaves, val_score: 0.655992:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000434 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.631152	valid's binary_logloss: 0.657113
[200]	train's binary_logloss: 0.611555	valid's binary_logloss: 0.657151
Early stopping, best iteration is:
[114]	train's binary_logloss: 0.627933	valid's binary_logloss: 0.656291
num_leaves, val_score: 0.655992:   5%|5         | 1/20 [00:00<00:07,  2.47it/s][I 2020-09-27 04:39:15,682] Trial 7 finished with value: 0.6562905644151281 and parameters: {'num_leaves': 10}. Best is trial 7 with value: 0.6562905644151281.
num_leaves, val_score: 0.655992:   5%|5         | 1/20 [00:00<00:07,  2.47it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000313 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.313854	valid's binary_logloss: 0.672207
Early stopping, best iteration is:
[34]	train's binary_logloss: 0.498751	valid's binary_logloss: 0.663462
num_leaves, val_score: 0.655992:  10%|#         | 2/20 [00:01<00:13,  1.32it/s][I 2020-09-27 04:39:17,270] Trial 8 finished with value: 0.6634620451568378 and parameters: {'num_leaves': 238}. Best is trial 7 with value: 0.6562905644151281.
num_leaves, val_score: 0.655992:  10%|#         | 2/20 [00:02<00:13,  1.32it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000229 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.442275	valid's binary_logloss: 0.665841
Early stopping, best iteration is:
[40]	train's binary_logloss: 0.552251	valid's binary_logloss: 0.660374
num_leaves, val_score: 0.655992:  15%|#5        | 3/20 [00:02<00:12,  1.37it/s][I 2020-09-27 04:39:17,923] Trial 9 finished with value: 0.6603744399138358 and parameters: {'num_leaves': 116}. Best is trial 7 with value: 0.6562905644151281.
num_leaves, val_score: 0.655992:  15%|#5        | 3/20 [00:02<00:12,  1.37it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010332 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.308823	valid's binary_logloss: 0.670809
Early stopping, best iteration is:
[36]	train's binary_logloss: 0.48791	valid's binary_logloss: 0.665738
num_leaves, val_score: 0.655992:  20%|##        | 4/20 [00:03<00:12,  1.27it/s][I 2020-09-27 04:39:18,840] Trial 10 finished with value: 0.6657380834337528 and parameters: {'num_leaves': 244}. Best is trial 7 with value: 0.6562905644151281.
num_leaves, val_score: 0.655992:  20%|##        | 4/20 [00:03<00:12,  1.27it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000322 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.658235	valid's binary_logloss: 0.664731
[200]	train's binary_logloss: 0.647762	valid's binary_logloss: 0.657647
[300]	train's binary_logloss: 0.642542	valid's binary_logloss: 0.655465
[400]	train's binary_logloss: 0.638941	valid's binary_logloss: 0.654486
[500]	train's binary_logloss: 0.636035	valid's binary_logloss: 0.653981
[600]	train's binary_logloss: 0.63343	valid's binary_logloss: 0.65371
Early stopping, best iteration is:
[555]	train's binary_logloss: 0.634597	valid's binary_logloss: 0.653553
num_leaves, val_score: 0.653553:  25%|##5       | 5/20 [00:04<00:11,  1.26it/s][I 2020-09-27 04:39:19,645] Trial 11 finished with value: 0.6535525819216924 and parameters: {'num_leaves': 3}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  25%|##5       | 5/20 [00:04<00:11,  1.26it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000332 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.620698	valid's binary_logloss: 0.657525
[200]	train's binary_logloss: 0.594023	valid's binary_logloss: 0.657636
Early stopping, best iteration is:
[151]	train's binary_logloss: 0.606458	valid's binary_logloss: 0.657287
num_leaves, val_score: 0.653553:  30%|###       | 6/20 [00:05<00:10,  1.32it/s][I 2020-09-27 04:39:20,325] Trial 12 finished with value: 0.6572874929871441 and parameters: {'num_leaves': 14}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  30%|###       | 6/20 [00:05<00:10,  1.32it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.011876 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.467946	valid's binary_logloss: 0.670734
Early stopping, best iteration is:
[45]	train's binary_logloss: 0.556356	valid's binary_logloss: 0.664088
num_leaves, val_score: 0.653553:  35%|###5      | 7/20 [00:05<00:10,  1.25it/s][I 2020-09-27 04:39:21,228] Trial 13 finished with value: 0.6640878662544877 and parameters: {'num_leaves': 97}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  35%|###5      | 7/20 [00:05<00:10,  1.25it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000489 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.524406	valid's binary_logloss: 0.663032
Early stopping, best iteration is:
[41]	train's binary_logloss: 0.595701	valid's binary_logloss: 0.662191
num_leaves, val_score: 0.653553:  40%|####      | 8/20 [00:06<00:08,  1.39it/s][I 2020-09-27 04:39:21,759] Trial 14 finished with value: 0.6621905066342226 and parameters: {'num_leaves': 62}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  40%|####      | 8/20 [00:06<00:08,  1.39it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000238 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.367271	valid's binary_logloss: 0.673756
Early stopping, best iteration is:
[34]	train's binary_logloss: 0.528584	valid's binary_logloss: 0.666502
num_leaves, val_score: 0.653553:  45%|####5     | 9/20 [00:07<00:08,  1.30it/s][I 2020-09-27 04:39:22,638] Trial 15 finished with value: 0.6665016344232115 and parameters: {'num_leaves': 180}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  45%|####5     | 9/20 [00:07<00:08,  1.30it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000232 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.36627	valid's binary_logloss: 0.674896
Early stopping, best iteration is:
[36]	train's binary_logloss: 0.521716	valid's binary_logloss: 0.66343
num_leaves, val_score: 0.653553:  50%|#####     | 10/20 [00:08<00:08,  1.21it/s][I 2020-09-27 04:39:23,590] Trial 16 finished with value: 0.6634300540848975 and parameters: {'num_leaves': 183}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  50%|#####     | 10/20 [00:08<00:08,  1.21it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000415 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.536123	valid's binary_logloss: 0.658196
Early stopping, best iteration is:
[92]	train's binary_logloss: 0.543103	valid's binary_logloss: 0.657473
num_leaves, val_score: 0.653553:  55%|#####5    | 11/20 [00:09<00:07,  1.13it/s][I 2020-09-27 04:39:24,613] Trial 17 finished with value: 0.6574725321803361 and parameters: {'num_leaves': 55}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  55%|#####5    | 11/20 [00:09<00:07,  1.13it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000386 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.386366	valid's binary_logloss: 0.670881
Early stopping, best iteration is:
[32]	train's binary_logloss: 0.54618	valid's binary_logloss: 0.663828
num_leaves, val_score: 0.653553:  60%|######    | 12/20 [00:10<00:07,  1.14it/s][I 2020-09-27 04:39:25,475] Trial 18 finished with value: 0.663827571242285 and parameters: {'num_leaves': 162}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  60%|######    | 12/20 [00:10<00:07,  1.14it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000380 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.543127	valid's binary_logloss: 0.660244
Early stopping, best iteration is:
[92]	train's binary_logloss: 0.549667	valid's binary_logloss: 0.659503
num_leaves, val_score: 0.653553:  65%|######5   | 13/20 [00:10<00:05,  1.28it/s][I 2020-09-27 04:39:26,025] Trial 19 finished with value: 0.6595031085541457 and parameters: {'num_leaves': 51}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  65%|######5   | 13/20 [00:10<00:05,  1.28it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000521 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.636989	valid's binary_logloss: 0.657697
[200]	train's binary_logloss: 0.620568	valid's binary_logloss: 0.656088
Early stopping, best iteration is:
[171]	train's binary_logloss: 0.624795	valid's binary_logloss: 0.655485
num_leaves, val_score: 0.653553:  70%|#######   | 14/20 [00:11<00:04,  1.49it/s][I 2020-09-27 04:39:26,451] Trial 20 finished with value: 0.6554852556407339 and parameters: {'num_leaves': 8}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  70%|#######   | 14/20 [00:11<00:04,  1.49it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004552 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.643142	valid's binary_logloss: 0.658677
[200]	train's binary_logloss: 0.62998	valid's binary_logloss: 0.656522
[300]	train's binary_logloss: 0.620354	valid's binary_logloss: 0.655925
Early stopping, best iteration is:
[297]	train's binary_logloss: 0.620584	valid's binary_logloss: 0.655776
num_leaves, val_score: 0.653553:  75%|#######5  | 15/20 [00:11<00:03,  1.58it/s][I 2020-09-27 04:39:26,995] Trial 21 finished with value: 0.655776407729715 and parameters: {'num_leaves': 6}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  75%|#######5  | 15/20 [00:11<00:03,  1.58it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000245 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.639989	valid's binary_logloss: 0.657478
[200]	train's binary_logloss: 0.62545	valid's binary_logloss: 0.65644
Early stopping, best iteration is:
[140]	train's binary_logloss: 0.633249	valid's binary_logloss: 0.655769
num_leaves, val_score: 0.653553:  80%|########  | 16/20 [00:12<00:02,  1.77it/s][I 2020-09-27 04:39:27,399] Trial 22 finished with value: 0.6557689832038824 and parameters: {'num_leaves': 7}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  80%|########  | 16/20 [00:12<00:02,  1.77it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004511 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.575815	valid's binary_logloss: 0.658024
[200]	train's binary_logloss: 0.522157	valid's binary_logloss: 0.657932
Early stopping, best iteration is:
[154]	train's binary_logloss: 0.544614	valid's binary_logloss: 0.656365
num_leaves, val_score: 0.653553:  85%|########5 | 17/20 [00:13<00:02,  1.40it/s][I 2020-09-27 04:39:28,462] Trial 23 finished with value: 0.6563652624863247 and parameters: {'num_leaves': 34}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  85%|########5 | 17/20 [00:13<00:02,  1.40it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.009938 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.647155	valid's binary_logloss: 0.660323
[200]	train's binary_logloss: 0.635295	valid's binary_logloss: 0.655924
Early stopping, best iteration is:
[188]	train's binary_logloss: 0.63642	valid's binary_logloss: 0.655643
num_leaves, val_score: 0.653553:  90%|######### | 18/20 [00:13<00:01,  1.59it/s][I 2020-09-27 04:39:28,891] Trial 24 finished with value: 0.655643287434117 and parameters: {'num_leaves': 5}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  90%|######### | 18/20 [00:13<00:01,  1.59it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000491 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.488408	valid's binary_logloss: 0.664489
Early stopping, best iteration is:
[78]	train's binary_logloss: 0.517222	valid's binary_logloss: 0.661867
num_leaves, val_score: 0.653553:  95%|#########5| 19/20 [00:14<00:00,  1.53it/s][I 2020-09-27 04:39:29,600] Trial 25 finished with value: 0.6618666948452248 and parameters: {'num_leaves': 84}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553:  95%|#########5| 19/20 [00:14<00:00,  1.53it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000236 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.580318	valid's binary_logloss: 0.659611
Early stopping, best iteration is:
[67]	train's binary_logloss: 0.601799	valid's binary_logloss: 0.658994
num_leaves, val_score: 0.653553: 100%|##########| 20/20 [00:14<00:00,  1.71it/s][I 2020-09-27 04:39:30,031] Trial 26 finished with value: 0.6589942104081719 and parameters: {'num_leaves': 32}. Best is trial 11 with value: 0.6535525819216924.
num_leaves, val_score: 0.653553: 100%|##########| 20/20 [00:14<00:00,  1.35it/s]
bagging, val_score: 0.653553:   0%|          | 0/10 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.020148 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657525	valid's binary_logloss: 0.662893
[200]	train's binary_logloss: 0.647047	valid's binary_logloss: 0.656095
[300]	train's binary_logloss: 0.641935	valid's binary_logloss: 0.654303
[400]	train's binary_logloss: 0.638294	valid's binary_logloss: 0.653695
[500]	train's binary_logloss: 0.635159	valid's binary_logloss: 0.653205
[600]	train's binary_logloss: 0.632193	valid's binary_logloss: 0.652803
[700]	train's binary_logloss: 0.629395	valid's binary_logloss: 0.651799
[800]	train's binary_logloss: 0.626828	valid's binary_logloss: 0.652063
Early stopping, best iteration is:
[745]	train's binary_logloss: 0.628188	valid's binary_logloss: 0.651395
bagging, val_score: 0.651395:  10%|#         | 1/10 [00:01<00:10,  1.19s/it][I 2020-09-27 04:39:31,244] Trial 27 finished with value: 0.6513952391116833 and parameters: {'bagging_fraction': 0.8097719374771206, 'bagging_freq': 2}. Best is trial 27 with value: 0.6513952391116833.
bagging, val_score: 0.651395:  10%|#         | 1/10 [00:01<00:10,  1.19s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000403 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657637	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642211	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638597	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635489	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632512	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.62981	valid's binary_logloss: 0.651852
[800]	train's binary_logloss: 0.627226	valid's binary_logloss: 0.651697
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628657	valid's binary_logloss: 0.651255
bagging, val_score: 0.651255:  20%|##        | 2/10 [00:02<00:10,  1.35s/it][I 2020-09-27 04:39:32,959] Trial 28 finished with value: 0.6512546217026939 and parameters: {'bagging_fraction': 0.8280339557117807, 'bagging_freq': 2}. Best is trial 28 with value: 0.6512546217026939.
bagging, val_score: 0.651255:  20%|##        | 2/10 [00:02<00:10,  1.35s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000249 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657845	valid's binary_logloss: 0.663325
[200]	train's binary_logloss: 0.647265	valid's binary_logloss: 0.656205
[300]	train's binary_logloss: 0.642195	valid's binary_logloss: 0.653918
[400]	train's binary_logloss: 0.638463	valid's binary_logloss: 0.653807
[500]	train's binary_logloss: 0.635325	valid's binary_logloss: 0.653223
[600]	train's binary_logloss: 0.632547	valid's binary_logloss: 0.652408
[700]	train's binary_logloss: 0.629962	valid's binary_logloss: 0.652142
[800]	train's binary_logloss: 0.627395	valid's binary_logloss: 0.652339
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628893	valid's binary_logloss: 0.651858
bagging, val_score: 0.651255:  30%|###       | 3/10 [00:04<00:09,  1.32s/it][I 2020-09-27 04:39:34,223] Trial 29 finished with value: 0.6518582592610941 and parameters: {'bagging_fraction': 0.8423665601800546, 'bagging_freq': 2}. Best is trial 28 with value: 0.6512546217026939.
bagging, val_score: 0.651255:  30%|###       | 3/10 [00:04<00:09,  1.32s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000300 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657643	valid's binary_logloss: 0.663425
[200]	train's binary_logloss: 0.647094	valid's binary_logloss: 0.65631
[300]	train's binary_logloss: 0.642016	valid's binary_logloss: 0.653742
[400]	train's binary_logloss: 0.638407	valid's binary_logloss: 0.653279
Early stopping, best iteration is:
[377]	train's binary_logloss: 0.639168	valid's binary_logloss: 0.653138
bagging, val_score: 0.651255:  40%|####      | 4/10 [00:04<00:06,  1.14s/it][I 2020-09-27 04:39:34,939] Trial 30 finished with value: 0.6531380481905572 and parameters: {'bagging_fraction': 0.8282561902576878, 'bagging_freq': 2}. Best is trial 28 with value: 0.6512546217026939.
bagging, val_score: 0.651255:  40%|####      | 4/10 [00:04<00:06,  1.14s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000244 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.65787	valid's binary_logloss: 0.662913
[200]	train's binary_logloss: 0.647202	valid's binary_logloss: 0.656389
[300]	train's binary_logloss: 0.642181	valid's binary_logloss: 0.65432
[400]	train's binary_logloss: 0.638556	valid's binary_logloss: 0.653816
Early stopping, best iteration is:
[361]	train's binary_logloss: 0.639969	valid's binary_logloss: 0.653683
bagging, val_score: 0.651255:  50%|#####     | 5/10 [00:06<00:05,  1.14s/it][I 2020-09-27 04:39:36,081] Trial 31 finished with value: 0.6536825951065313 and parameters: {'bagging_fraction': 0.8381544500351495, 'bagging_freq': 2}. Best is trial 28 with value: 0.6512546217026939.
bagging, val_score: 0.651255:  50%|#####     | 5/10 [00:06<00:05,  1.14s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000455 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657707	valid's binary_logloss: 0.662425
[200]	train's binary_logloss: 0.64726	valid's binary_logloss: 0.655364
[300]	train's binary_logloss: 0.64231	valid's binary_logloss: 0.653736
[400]	train's binary_logloss: 0.638629	valid's binary_logloss: 0.653251
[500]	train's binary_logloss: 0.63557	valid's binary_logloss: 0.652996
[600]	train's binary_logloss: 0.632836	valid's binary_logloss: 0.653094
[700]	train's binary_logloss: 0.630172	valid's binary_logloss: 0.652558
[800]	train's binary_logloss: 0.627493	valid's binary_logloss: 0.652618
Early stopping, best iteration is:
[738]	train's binary_logloss: 0.629206	valid's binary_logloss: 0.65221
bagging, val_score: 0.651255:  60%|######    | 6/10 [00:07<00:04,  1.16s/it][I 2020-09-27 04:39:37,277] Trial 32 finished with value: 0.6522103084600892 and parameters: {'bagging_fraction': 0.8240973522704712, 'bagging_freq': 2}. Best is trial 28 with value: 0.6512546217026939.
bagging, val_score: 0.651255:  60%|######    | 6/10 [00:07<00:04,  1.16s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000240 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657759	valid's binary_logloss: 0.6633
[200]	train's binary_logloss: 0.647172	valid's binary_logloss: 0.656277
[300]	train's binary_logloss: 0.642123	valid's binary_logloss: 0.654052
[400]	train's binary_logloss: 0.638549	valid's binary_logloss: 0.653547
[500]	train's binary_logloss: 0.635603	valid's binary_logloss: 0.653766
Early stopping, best iteration is:
[419]	train's binary_logloss: 0.637971	valid's binary_logloss: 0.653468
bagging, val_score: 0.651255:  70%|#######   | 7/10 [00:08<00:03,  1.04s/it][I 2020-09-27 04:39:38,055] Trial 33 finished with value: 0.6534681190347887 and parameters: {'bagging_fraction': 0.8133668384521671, 'bagging_freq': 2}. Best is trial 28 with value: 0.6512546217026939.
bagging, val_score: 0.651255:  70%|#######   | 7/10 [00:08<00:03,  1.04s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000309 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.65767	valid's binary_logloss: 0.663092
[200]	train's binary_logloss: 0.646954	valid's binary_logloss: 0.65622
[300]	train's binary_logloss: 0.641882	valid's binary_logloss: 0.653856
[400]	train's binary_logloss: 0.638359	valid's binary_logloss: 0.653238
[500]	train's binary_logloss: 0.635264	valid's binary_logloss: 0.652485
[600]	train's binary_logloss: 0.632569	valid's binary_logloss: 0.652572
Early stopping, best iteration is:
[500]	train's binary_logloss: 0.635264	valid's binary_logloss: 0.652485
bagging, val_score: 0.651255:  80%|########  | 8/10 [00:08<00:01,  1.01it/s][I 2020-09-27 04:39:38,927] Trial 34 finished with value: 0.6524845330402806 and parameters: {'bagging_fraction': 0.833189344448115, 'bagging_freq': 2}. Best is trial 28 with value: 0.6512546217026939.
bagging, val_score: 0.651255:  80%|########  | 8/10 [00:08<00:01,  1.01it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000447 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657357	valid's binary_logloss: 0.663082
[200]	train's binary_logloss: 0.646962	valid's binary_logloss: 0.656317
[300]	train's binary_logloss: 0.642048	valid's binary_logloss: 0.65432
Early stopping, best iteration is:
[285]	train's binary_logloss: 0.642706	valid's binary_logloss: 0.653911
bagging, val_score: 0.651255:  90%|######### | 9/10 [00:09<00:01,  1.02s/it][I 2020-09-27 04:39:40,016] Trial 35 finished with value: 0.653910601498707 and parameters: {'bagging_fraction': 0.6953833757673704, 'bagging_freq': 3}. Best is trial 28 with value: 0.6512546217026939.
bagging, val_score: 0.651255:  90%|######### | 9/10 [00:09<00:01,  1.02s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000404 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.658127	valid's binary_logloss: 0.664339
[200]	train's binary_logloss: 0.647507	valid's binary_logloss: 0.656812
[300]	train's binary_logloss: 0.642315	valid's binary_logloss: 0.65461
[400]	train's binary_logloss: 0.638734	valid's binary_logloss: 0.653365
[500]	train's binary_logloss: 0.635615	valid's binary_logloss: 0.653259
Early stopping, best iteration is:
[439]	train's binary_logloss: 0.637483	valid's binary_logloss: 0.652984
bagging, val_score: 0.651255: 100%|##########| 10/10 [00:10<00:00,  1.03it/s][I 2020-09-27 04:39:40,868] Trial 36 finished with value: 0.6529843808564952 and parameters: {'bagging_fraction': 0.9750898583135423, 'bagging_freq': 5}. Best is trial 28 with value: 0.6512546217026939.
bagging, val_score: 0.651255: 100%|##########| 10/10 [00:10<00:00,  1.08s/it]
feature_fraction_stage2, val_score: 0.651255:   0%|          | 0/3 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000412 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657426	valid's binary_logloss: 0.66294
[200]	train's binary_logloss: 0.646763	valid's binary_logloss: 0.65579
[300]	train's binary_logloss: 0.641842	valid's binary_logloss: 0.653608
[400]	train's binary_logloss: 0.638277	valid's binary_logloss: 0.653005
[500]	train's binary_logloss: 0.635065	valid's binary_logloss: 0.652871
Early stopping, best iteration is:
[437]	train's binary_logloss: 0.637058	valid's binary_logloss: 0.652763
feature_fraction_stage2, val_score: 0.651255:  33%|###3      | 1/3 [00:00<00:01,  1.20it/s][I 2020-09-27 04:39:41,719] Trial 37 finished with value: 0.6527633350213652 and parameters: {'feature_fraction': 0.41600000000000004}. Best is trial 37 with value: 0.6527633350213652.
feature_fraction_stage2, val_score: 0.651255:  33%|###3      | 1/3 [00:00<00:01,  1.20it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000510 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657376	valid's binary_logloss: 0.66298
[200]	train's binary_logloss: 0.646944	valid's binary_logloss: 0.656526
[300]	train's binary_logloss: 0.641961	valid's binary_logloss: 0.653998
[400]	train's binary_logloss: 0.638308	valid's binary_logloss: 0.653033
[500]	train's binary_logloss: 0.634963	valid's binary_logloss: 0.652584
[600]	train's binary_logloss: 0.632079	valid's binary_logloss: 0.652716
[700]	train's binary_logloss: 0.629142	valid's binary_logloss: 0.652471
[800]	train's binary_logloss: 0.62641	valid's binary_logloss: 0.652186
Early stopping, best iteration is:
[758]	train's binary_logloss: 0.627519	valid's binary_logloss: 0.651784
feature_fraction_stage2, val_score: 0.651255:  67%|######6   | 2/3 [00:02<00:00,  1.00it/s][I 2020-09-27 04:39:43,108] Trial 38 finished with value: 0.6517839548740617 and parameters: {'feature_fraction': 0.44800000000000006}. Best is trial 38 with value: 0.6517839548740617.
feature_fraction_stage2, val_score: 0.651255:  67%|######6   | 2/3 [00:02<00:00,  1.00it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.011259 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657376	valid's binary_logloss: 0.66298
[200]	train's binary_logloss: 0.646944	valid's binary_logloss: 0.656526
[300]	train's binary_logloss: 0.641961	valid's binary_logloss: 0.653998
[400]	train's binary_logloss: 0.638308	valid's binary_logloss: 0.653033
[500]	train's binary_logloss: 0.634963	valid's binary_logloss: 0.652584
[600]	train's binary_logloss: 0.632079	valid's binary_logloss: 0.652716
[700]	train's binary_logloss: 0.629142	valid's binary_logloss: 0.652471
[800]	train's binary_logloss: 0.62641	valid's binary_logloss: 0.652186
Early stopping, best iteration is:
[758]	train's binary_logloss: 0.627519	valid's binary_logloss: 0.651784
feature_fraction_stage2, val_score: 0.651255: 100%|##########| 3/3 [00:03<00:00,  1.15s/it][I 2020-09-27 04:39:44,601] Trial 39 finished with value: 0.6517839548740617 and parameters: {'feature_fraction': 0.48000000000000004}. Best is trial 38 with value: 0.6517839548740617.
feature_fraction_stage2, val_score: 0.651255: 100%|##########| 3/3 [00:03<00:00,  1.24s/it]
regularization_factors, val_score: 0.651255:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000460 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657637	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642211	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638597	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635489	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632512	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.62981	valid's binary_logloss: 0.651852
[800]	train's binary_logloss: 0.627227	valid's binary_logloss: 0.651697
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628658	valid's binary_logloss: 0.651255
regularization_factors, val_score: 0.651255:   5%|5         | 1/20 [00:01<00:23,  1.23s/it][I 2020-09-27 04:39:45,848] Trial 40 finished with value: 0.6512546032948433 and parameters: {'lambda_l1': 3.7407304516474708e-06, 'lambda_l2': 0.0003236321792160619}. Best is trial 40 with value: 0.6512546032948433.
regularization_factors, val_score: 0.651255:   5%|5         | 1/20 [00:01<00:23,  1.23s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000479 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657637	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642211	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638597	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635489	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632512	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.62981	valid's binary_logloss: 0.651852
[800]	train's binary_logloss: 0.627227	valid's binary_logloss: 0.651697
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628658	valid's binary_logloss: 0.651255
regularization_factors, val_score: 0.651255:  10%|#         | 2/20 [00:02<00:23,  1.33s/it][I 2020-09-27 04:39:47,419] Trial 41 finished with value: 0.6512546004999574 and parameters: {'lambda_l1': 1.58423580909119e-06, 'lambda_l2': 0.00037329518153918976}. Best is trial 41 with value: 0.6512546004999574.
regularization_factors, val_score: 0.651255:  10%|#         | 2/20 [00:02<00:23,  1.33s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000390 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657637	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642211	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638597	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635489	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632512	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.62981	valid's binary_logloss: 0.651852
[800]	train's binary_logloss: 0.627227	valid's binary_logloss: 0.651697
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628658	valid's binary_logloss: 0.651255
regularization_factors, val_score: 0.651255:  15%|#5        | 3/20 [00:04<00:22,  1.33s/it][I 2020-09-27 04:39:48,734] Trial 42 finished with value: 0.6512545924218361 and parameters: {'lambda_l1': 1.3626757884196743e-06, 'lambda_l2': 0.0005160151192105561}. Best is trial 42 with value: 0.6512545924218361.
regularization_factors, val_score: 0.651255:  15%|#5        | 3/20 [00:04<00:22,  1.33s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000381 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657637	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642211	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638597	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635489	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632512	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.62981	valid's binary_logloss: 0.651852
[800]	train's binary_logloss: 0.627227	valid's binary_logloss: 0.651697
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628658	valid's binary_logloss: 0.651255
regularization_factors, val_score: 0.651255:  20%|##        | 4/20 [00:05<00:20,  1.29s/it][I 2020-09-27 04:39:49,939] Trial 43 finished with value: 0.6512546024195541 and parameters: {'lambda_l1': 1.826384761461434e-06, 'lambda_l2': 0.000339404582931327}. Best is trial 42 with value: 0.6512545924218361.
regularization_factors, val_score: 0.651255:  20%|##        | 4/20 [00:05<00:20,  1.29s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000442 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657637	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642211	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638597	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635489	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632512	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.62981	valid's binary_logloss: 0.651852
[800]	train's binary_logloss: 0.627227	valid's binary_logloss: 0.651697
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628658	valid's binary_logloss: 0.651255
regularization_factors, val_score: 0.651255:  25%|##5       | 5/20 [00:07<00:21,  1.42s/it][I 2020-09-27 04:39:51,672] Trial 44 finished with value: 0.6512545971181739 and parameters: {'lambda_l1': 9.830713052753886e-07, 'lambda_l2': 0.00043290604666443164}. Best is trial 42 with value: 0.6512545924218361.
regularization_factors, val_score: 0.651255:  25%|##5       | 5/20 [00:07<00:21,  1.42s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000564 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657637	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642211	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638597	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635489	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632512	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.62981	valid's binary_logloss: 0.651852
[800]	train's binary_logloss: 0.627227	valid's binary_logloss: 0.651697
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628658	valid's binary_logloss: 0.651255
regularization_factors, val_score: 0.651255:  30%|###       | 6/20 [00:08<00:19,  1.36s/it][I 2020-09-27 04:39:52,891] Trial 45 finished with value: 0.6512546006093322 and parameters: {'lambda_l1': 7.770048321005081e-07, 'lambda_l2': 0.00037155770796455103}. Best is trial 42 with value: 0.6512545924218361.
regularization_factors, val_score: 0.651255:  30%|###       | 6/20 [00:08<00:19,  1.36s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000386 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657637	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642211	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638597	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635489	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632512	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.62981	valid's binary_logloss: 0.651852
[800]	train's binary_logloss: 0.627227	valid's binary_logloss: 0.651697
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628658	valid's binary_logloss: 0.651255
regularization_factors, val_score: 0.651255:  35%|###5      | 7/20 [00:09<00:17,  1.32s/it][I 2020-09-27 04:39:54,130] Trial 46 finished with value: 0.65125458769479 and parameters: {'lambda_l1': 5.65299654936304e-07, 'lambda_l2': 0.0005992895040580498}. Best is trial 46 with value: 0.65125458769479.
regularization_factors, val_score: 0.651255:  35%|###5      | 7/20 [00:09<00:17,  1.32s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000244 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657637	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642211	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638597	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635489	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632512	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.62981	valid's binary_logloss: 0.651852
[800]	train's binary_logloss: 0.627227	valid's binary_logloss: 0.651697
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628658	valid's binary_logloss: 0.651255
regularization_factors, val_score: 0.651255:  40%|####      | 8/20 [00:11<00:17,  1.42s/it][I 2020-09-27 04:39:55,765] Trial 47 finished with value: 0.651254582843981 and parameters: {'lambda_l1': 5.113147117976975e-07, 'lambda_l2': 0.0006849316707389379}. Best is trial 47 with value: 0.651254582843981.
regularization_factors, val_score: 0.651255:  40%|####      | 8/20 [00:11<00:17,  1.42s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000452 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657637	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642211	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638597	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635489	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632512	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.62981	valid's binary_logloss: 0.651852
[800]	train's binary_logloss: 0.627227	valid's binary_logloss: 0.651697
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628658	valid's binary_logloss: 0.651255
regularization_factors, val_score: 0.651255:  45%|####5     | 9/20 [00:12<00:14,  1.35s/it][I 2020-09-27 04:39:56,969] Trial 48 finished with value: 0.6512545741011545 and parameters: {'lambda_l1': 3.656814382369128e-07, 'lambda_l2': 0.0008392662361736358}. Best is trial 48 with value: 0.6512545741011545.
regularization_factors, val_score: 0.651255:  45%|####5     | 9/20 [00:12<00:14,  1.35s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000266 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657637	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642211	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638597	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635489	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632513	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.629811	valid's binary_logloss: 0.651852
[800]	train's binary_logloss: 0.627228	valid's binary_logloss: 0.651696
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628659	valid's binary_logloss: 0.651254
regularization_factors, val_score: 0.651254:  50%|#####     | 10/20 [00:13<00:13,  1.31s/it][I 2020-09-27 04:39:58,184] Trial 49 finished with value: 0.65125447186593 and parameters: {'lambda_l1': 4.097168681166982e-08, 'lambda_l2': 0.002642450368328453}. Best is trial 49 with value: 0.65125447186593.
regularization_factors, val_score: 0.651254:  50%|#####     | 10/20 [00:13<00:13,  1.31s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000234 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657642	valid's binary_logloss: 0.663104
[200]	train's binary_logloss: 0.647161	valid's binary_logloss: 0.6561
[300]	train's binary_logloss: 0.642211	valid's binary_logloss: 0.653789
[400]	train's binary_logloss: 0.638567	valid's binary_logloss: 0.652995
Early stopping, best iteration is:
[378]	train's binary_logloss: 0.639273	valid's binary_logloss: 0.65285
regularization_factors, val_score: 0.651254:  55%|#####5    | 11/20 [00:14<00:11,  1.26s/it][I 2020-09-27 04:39:59,310] Trial 50 finished with value: 0.6528496531447342 and parameters: {'lambda_l1': 1.8926948433638843e-08, 'lambda_l2': 0.18218516904146864}. Best is trial 49 with value: 0.65125447186593.
regularization_factors, val_score: 0.651254:  55%|#####5    | 11/20 [00:14<00:11,  1.26s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000499 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657637	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642211	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638597	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635489	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632513	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.629811	valid's binary_logloss: 0.651852
[800]	train's binary_logloss: 0.627228	valid's binary_logloss: 0.651697
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628659	valid's binary_logloss: 0.651254
regularization_factors, val_score: 0.651254:  60%|######    | 12/20 [00:15<00:10,  1.25s/it][I 2020-09-27 04:40:00,559] Trial 51 finished with value: 0.651254491688544 and parameters: {'lambda_l1': 9.464855403688449e-08, 'lambda_l2': 0.0022922152533675836}. Best is trial 49 with value: 0.65125447186593.
regularization_factors, val_score: 0.651254:  60%|######    | 12/20 [00:15<00:10,  1.25s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000381 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657638	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642213	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.6386	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635493	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632518	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.629817	valid's binary_logloss: 0.651851
[800]	train's binary_logloss: 0.627236	valid's binary_logloss: 0.651695
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628666	valid's binary_logloss: 0.651254
regularization_factors, val_score: 0.651254:  65%|######5   | 13/20 [00:17<00:09,  1.34s/it][I 2020-09-27 04:40:02,095] Trial 52 finished with value: 0.651253816244402 and parameters: {'lambda_l1': 1.1047734679593184e-08, 'lambda_l2': 0.014255881410398455}. Best is trial 52 with value: 0.651253816244402.
regularization_factors, val_score: 0.651254:  65%|######5   | 13/20 [00:17<00:09,  1.34s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000248 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657638	valid's binary_logloss: 0.663102
[200]	train's binary_logloss: 0.647156	valid's binary_logloss: 0.656097
[300]	train's binary_logloss: 0.642216	valid's binary_logloss: 0.653703
[400]	train's binary_logloss: 0.638607	valid's binary_logloss: 0.652863
[500]	train's binary_logloss: 0.6355	valid's binary_logloss: 0.652355
[600]	train's binary_logloss: 0.632662	valid's binary_logloss: 0.652694
Early stopping, best iteration is:
[531]	train's binary_logloss: 0.634567	valid's binary_logloss: 0.652228
regularization_factors, val_score: 0.651254:  70%|#######   | 14/20 [00:19<00:09,  1.53s/it][I 2020-09-27 04:40:04,065] Trial 53 finished with value: 0.6522275662505942 and parameters: {'lambda_l1': 1.4618448056501589e-08, 'lambda_l2': 0.04375470564403219}. Best is trial 52 with value: 0.651253816244402.
regularization_factors, val_score: 0.651254:  70%|#######   | 14/20 [00:19<00:09,  1.53s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000394 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657638	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647154	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642213	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.6386	valid's binary_logloss: 0.652862
[500]	train's binary_logloss: 0.635493	valid's binary_logloss: 0.652354
[600]	train's binary_logloss: 0.632518	valid's binary_logloss: 0.652123
[700]	train's binary_logloss: 0.629817	valid's binary_logloss: 0.651851
[800]	train's binary_logloss: 0.627236	valid's binary_logloss: 0.651695
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628666	valid's binary_logloss: 0.651254
regularization_factors, val_score: 0.651254:  75%|#######5  | 15/20 [00:20<00:07,  1.45s/it][I 2020-09-27 04:40:05,343] Trial 54 finished with value: 0.6512538204787435 and parameters: {'lambda_l1': 1.0600410018512844e-08, 'lambda_l2': 0.014180637901064958}. Best is trial 52 with value: 0.651253816244402.
regularization_factors, val_score: 0.651254:  75%|#######5  | 15/20 [00:20<00:07,  1.45s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000400 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657638	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647155	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642214	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638602	valid's binary_logloss: 0.652863
[500]	train's binary_logloss: 0.635496	valid's binary_logloss: 0.652353
[600]	train's binary_logloss: 0.632522	valid's binary_logloss: 0.652122
[700]	train's binary_logloss: 0.629822	valid's binary_logloss: 0.65185
[800]	train's binary_logloss: 0.627242	valid's binary_logloss: 0.651695
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628671	valid's binary_logloss: 0.651253
regularization_factors, val_score: 0.651253:  80%|########  | 16/20 [00:22<00:06,  1.51s/it][I 2020-09-27 04:40:06,971] Trial 55 finished with value: 0.6512532800647542 and parameters: {'lambda_l1': 1.1653169345025846e-08, 'lambda_l2': 0.02381661498148718}. Best is trial 55 with value: 0.6512532800647542.
regularization_factors, val_score: 0.651253:  80%|########  | 16/20 [00:22<00:06,  1.51s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.014470 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657638	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647155	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642213	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638601	valid's binary_logloss: 0.652863
[500]	train's binary_logloss: 0.635495	valid's binary_logloss: 0.652353
[600]	train's binary_logloss: 0.632521	valid's binary_logloss: 0.652122
[700]	train's binary_logloss: 0.62982	valid's binary_logloss: 0.65185
[800]	train's binary_logloss: 0.62724	valid's binary_logloss: 0.651695
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.628669	valid's binary_logloss: 0.651253
regularization_factors, val_score: 0.651253:  85%|########5 | 17/20 [00:23<00:04,  1.46s/it][I 2020-09-27 04:40:08,339] Trial 56 finished with value: 0.6512534743967614 and parameters: {'lambda_l1': 1.0004229968657813e-08, 'lambda_l2': 0.020345235153092046}. Best is trial 55 with value: 0.6512532800647542.
regularization_factors, val_score: 0.651253:  85%|########5 | 17/20 [00:23<00:04,  1.46s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000378 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657638	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647155	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642213	valid's binary_logloss: 0.653702
[400]	train's binary_logloss: 0.638602	valid's binary_logloss: 0.652863
[500]	train's binary_logloss: 0.635495	valid's binary_logloss: 0.652353
[600]	train's binary_logloss: 0.632521	valid's binary_logloss: 0.652122
[700]	train's binary_logloss: 0.629821	valid's binary_logloss: 0.65185
[800]	train's binary_logloss: 0.62724	valid's binary_logloss: 0.651695
Early stopping, best iteration is:
[743]	train's binary_logloss: 0.62867	valid's binary_logloss: 0.651253
regularization_factors, val_score: 0.651253:  90%|######### | 18/20 [00:25<00:02,  1.42s/it][I 2020-09-27 04:40:09,647] Trial 57 finished with value: 0.6512534186191336 and parameters: {'lambda_l1': 1.573306324418443e-08, 'lambda_l2': 0.021340714973965415}. Best is trial 55 with value: 0.6512532800647542.
regularization_factors, val_score: 0.651253:  90%|######### | 18/20 [00:25<00:02,  1.42s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004407 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657638	valid's binary_logloss: 0.663102
[200]	train's binary_logloss: 0.647156	valid's binary_logloss: 0.656097
[300]	train's binary_logloss: 0.642216	valid's binary_logloss: 0.653703
[400]	train's binary_logloss: 0.638606	valid's binary_logloss: 0.652863
[500]	train's binary_logloss: 0.635499	valid's binary_logloss: 0.652355
[600]	train's binary_logloss: 0.632662	valid's binary_logloss: 0.652694
Early stopping, best iteration is:
[531]	train's binary_logloss: 0.634567	valid's binary_logloss: 0.652228
regularization_factors, val_score: 0.651253:  95%|#########5| 19/20 [00:26<00:01,  1.41s/it][I 2020-09-27 04:40:11,033] Trial 58 finished with value: 0.6522275964597894 and parameters: {'lambda_l1': 1.1442661122184645e-08, 'lambda_l2': 0.04237276613246146}. Best is trial 55 with value: 0.6512532800647542.
regularization_factors, val_score: 0.651253:  95%|#########5| 19/20 [00:26<00:01,  1.41s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.002336 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657713	valid's binary_logloss: 0.663154
[200]	train's binary_logloss: 0.647209	valid's binary_logloss: 0.655905
[300]	train's binary_logloss: 0.64221	valid's binary_logloss: 0.65381
[400]	train's binary_logloss: 0.638641	valid's binary_logloss: 0.653241
Early stopping, best iteration is:
[351]	train's binary_logloss: 0.640312	valid's binary_logloss: 0.65316
regularization_factors, val_score: 0.651253: 100%|##########| 20/20 [00:27<00:00,  1.24s/it][I 2020-09-27 04:40:11,871] Trial 59 finished with value: 0.6531604651549241 and parameters: {'lambda_l1': 0.7583136516531738, 'lambda_l2': 0.017657519269218453}. Best is trial 55 with value: 0.6512532800647542.
regularization_factors, val_score: 0.651253: 100%|##########| 20/20 [00:27<00:00,  1.36s/it]
min_data_in_leaf, val_score: 0.651253:   0%|          | 0/5 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000390 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657638	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647129	valid's binary_logloss: 0.656026
[300]	train's binary_logloss: 0.642165	valid's binary_logloss: 0.653612
[400]	train's binary_logloss: 0.638561	valid's binary_logloss: 0.653007
[500]	train's binary_logloss: 0.635388	valid's binary_logloss: 0.652537
[600]	train's binary_logloss: 0.63262	valid's binary_logloss: 0.652704
Early stopping, best iteration is:
[575]	train's binary_logloss: 0.633278	valid's binary_logloss: 0.652405
min_data_in_leaf, val_score: 0.651253:  20%|##        | 1/5 [00:01<00:04,  1.04s/it][I 2020-09-27 04:40:12,924] Trial 60 finished with value: 0.6524054109680009 and parameters: {'min_child_samples': 5}. Best is trial 60 with value: 0.6524054109680009.
min_data_in_leaf, val_score: 0.651253:  20%|##        | 1/5 [00:01<00:04,  1.04s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000685 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657638	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647173	valid's binary_logloss: 0.656292
[300]	train's binary_logloss: 0.64235	valid's binary_logloss: 0.653435
[400]	train's binary_logloss: 0.639047	valid's binary_logloss: 0.652707
[500]	train's binary_logloss: 0.636172	valid's binary_logloss: 0.651812
[600]	train's binary_logloss: 0.633542	valid's binary_logloss: 0.651717
[700]	train's binary_logloss: 0.631145	valid's binary_logloss: 0.651603
[800]	train's binary_logloss: 0.628853	valid's binary_logloss: 0.651423
Early stopping, best iteration is:
[737]	train's binary_logloss: 0.630317	valid's binary_logloss: 0.651187
min_data_in_leaf, val_score: 0.651187:  40%|####      | 2/5 [00:02<00:03,  1.09s/it][I 2020-09-27 04:40:14,129] Trial 61 finished with value: 0.6511874541615081 and parameters: {'min_child_samples': 100}. Best is trial 61 with value: 0.6511874541615081.
min_data_in_leaf, val_score: 0.651187:  40%|####      | 2/5 [00:02<00:03,  1.09s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.006001 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657638	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647155	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.64222	valid's binary_logloss: 0.653679
[400]	train's binary_logloss: 0.638577	valid's binary_logloss: 0.653154
[500]	train's binary_logloss: 0.635489	valid's binary_logloss: 0.652771
[600]	train's binary_logloss: 0.632657	valid's binary_logloss: 0.652758
Early stopping, best iteration is:
[531]	train's binary_logloss: 0.634621	valid's binary_logloss: 0.65239
min_data_in_leaf, val_score: 0.651187:  60%|######    | 3/5 [00:03<00:02,  1.18s/it][I 2020-09-27 04:40:15,534] Trial 62 finished with value: 0.6523903635256941 and parameters: {'min_child_samples': 25}. Best is trial 61 with value: 0.6511874541615081.
min_data_in_leaf, val_score: 0.651187:  60%|######    | 3/5 [00:03<00:02,  1.18s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000824 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657638	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647129	valid's binary_logloss: 0.656026
[300]	train's binary_logloss: 0.642173	valid's binary_logloss: 0.653464
[400]	train's binary_logloss: 0.638672	valid's binary_logloss: 0.652989
[500]	train's binary_logloss: 0.635515	valid's binary_logloss: 0.652344
[600]	train's binary_logloss: 0.632548	valid's binary_logloss: 0.652081
[700]	train's binary_logloss: 0.629711	valid's binary_logloss: 0.651529
[800]	train's binary_logloss: 0.627105	valid's binary_logloss: 0.651486
Early stopping, best iteration is:
[749]	train's binary_logloss: 0.628438	valid's binary_logloss: 0.651112
min_data_in_leaf, val_score: 0.651112:  80%|########  | 4/5 [00:04<00:01,  1.19s/it][I 2020-09-27 04:40:16,745] Trial 63 finished with value: 0.6511117328857268 and parameters: {'min_child_samples': 10}. Best is trial 63 with value: 0.6511117328857268.
min_data_in_leaf, val_score: 0.651112:  80%|########  | 4/5 [00:04<00:01,  1.19s/it][LightGBM] [Info] Number of positive: 13151, number of negative: 12848
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000400 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4237
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505827 -> initscore=0.023310
[LightGBM] [Info] Start training from score 0.023310
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657638	valid's binary_logloss: 0.663101
[200]	train's binary_logloss: 0.647155	valid's binary_logloss: 0.656096
[300]	train's binary_logloss: 0.642357	valid's binary_logloss: 0.653658
[400]	train's binary_logloss: 0.638938	valid's binary_logloss: 0.653056
[500]	train's binary_logloss: 0.636126	valid's binary_logloss: 0.652589
[600]	train's binary_logloss: 0.633479	valid's binary_logloss: 0.652767
[700]	train's binary_logloss: 0.630836	valid's binary_logloss: 0.652316
Early stopping, best iteration is:
[693]	train's binary_logloss: 0.631059	valid's binary_logloss: 0.652118
min_data_in_leaf, val_score: 0.651112: 100%|##########| 5/5 [00:06<00:00,  1.19s/it][I 2020-09-27 04:40:17,925] Trial 64 finished with value: 0.6521175158063565 and parameters: {'min_child_samples': 50}. Best is trial 63 with value: 0.6511117328857268.
min_data_in_leaf, val_score: 0.651112: 100%|##########| 5/5 [00:06<00:00,  1.21s/it]
Fold : 2
[I 2020-09-27 04:40:18,017] A new study created in memory with name: no-name-75b2cbed-db22-45df-bfc2-635871fa6cfc
feature_fraction, val_score: inf:   0%|          | 0/7 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.007450 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.578542	valid's binary_logloss: 0.656491
Early stopping, best iteration is:
[79]	train's binary_logloss: 0.59215	valid's binary_logloss: 0.655948
feature_fraction, val_score: 0.655948:  14%|#4        | 1/7 [00:01<00:06,  1.04s/it][I 2020-09-27 04:40:19,072] Trial 0 finished with value: 0.655947777974299 and parameters: {'feature_fraction': 0.5}. Best is trial 0 with value: 0.655947777974299.
feature_fraction, val_score: 0.655948:  14%|#4        | 1/7 [00:01<00:06,  1.04s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.007324 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.572993	valid's binary_logloss: 0.65828
Early stopping, best iteration is:
[79]	train's binary_logloss: 0.587161	valid's binary_logloss: 0.656458
feature_fraction, val_score: 0.655948:  29%|##8       | 2/7 [00:01<00:04,  1.12it/s][I 2020-09-27 04:40:19,615] Trial 1 finished with value: 0.6564576518870222 and parameters: {'feature_fraction': 0.8}. Best is trial 0 with value: 0.655947777974299.
feature_fraction, val_score: 0.655948:  29%|##8       | 2/7 [00:01<00:04,  1.12it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.006440 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.570115	valid's binary_logloss: 0.66021
Early stopping, best iteration is:
[64]	train's binary_logloss: 0.595593	valid's binary_logloss: 0.658255
feature_fraction, val_score: 0.655948:  43%|####2     | 3/7 [00:02<00:03,  1.29it/s][I 2020-09-27 04:40:20,131] Trial 2 finished with value: 0.6582553269377266 and parameters: {'feature_fraction': 1.0}. Best is trial 0 with value: 0.655947777974299.
feature_fraction, val_score: 0.655948:  43%|####2     | 3/7 [00:02<00:03,  1.29it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010302 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.575364	valid's binary_logloss: 0.659046
Early stopping, best iteration is:
[68]	train's binary_logloss: 0.596133	valid's binary_logloss: 0.65791
feature_fraction, val_score: 0.655948:  57%|#####7    | 4/7 [00:02<00:02,  1.47it/s][I 2020-09-27 04:40:20,586] Trial 3 finished with value: 0.6579096104153804 and parameters: {'feature_fraction': 0.7}. Best is trial 0 with value: 0.655947777974299.
feature_fraction, val_score: 0.655948:  57%|#####7    | 4/7 [00:02<00:02,  1.47it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000900 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.572299	valid's binary_logloss: 0.658723
Early stopping, best iteration is:
[74]	train's binary_logloss: 0.590022	valid's binary_logloss: 0.656515
feature_fraction, val_score: 0.655948:  71%|#######1  | 5/7 [00:03<00:01,  1.58it/s][I 2020-09-27 04:40:21,113] Trial 4 finished with value: 0.6565150373220784 and parameters: {'feature_fraction': 0.8999999999999999}. Best is trial 0 with value: 0.655947777974299.
feature_fraction, val_score: 0.655948:  71%|#######1  | 5/7 [00:03<00:01,  1.58it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.009831 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.576537	valid's binary_logloss: 0.658655
Early stopping, best iteration is:
[67]	train's binary_logloss: 0.598042	valid's binary_logloss: 0.657232
feature_fraction, val_score: 0.655948:  86%|########5 | 6/7 [00:03<00:00,  1.74it/s][I 2020-09-27 04:40:21,552] Trial 5 finished with value: 0.6572319404658664 and parameters: {'feature_fraction': 0.6}. Best is trial 0 with value: 0.655947777974299.
feature_fraction, val_score: 0.655948:  86%|########5 | 6/7 [00:03<00:00,  1.74it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000333 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.583115	valid's binary_logloss: 0.656735
Early stopping, best iteration is:
[85]	train's binary_logloss: 0.592606	valid's binary_logloss: 0.655167
feature_fraction, val_score: 0.655167: 100%|##########| 7/7 [00:04<00:00,  1.37it/s][I 2020-09-27 04:40:22,641] Trial 6 finished with value: 0.6551672639343036 and parameters: {'feature_fraction': 0.4}. Best is trial 6 with value: 0.6551672639343036.
feature_fraction, val_score: 0.655167: 100%|##########| 7/7 [00:04<00:00,  1.52it/s]
num_leaves, val_score: 0.655167:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004364 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.362	valid's binary_logloss: 0.666306
Early stopping, best iteration is:
[62]	train's binary_logloss: 0.44475	valid's binary_logloss: 0.66107
num_leaves, val_score: 0.655167:   5%|5         | 1/20 [00:00<00:17,  1.09it/s][I 2020-09-27 04:40:23,577] Trial 7 finished with value: 0.6610695796505417 and parameters: {'num_leaves': 186}. Best is trial 7 with value: 0.6610695796505417.
num_leaves, val_score: 0.655167:   5%|5         | 1/20 [00:00<00:17,  1.09it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000378 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.509613	valid's binary_logloss: 0.655007
Early stopping, best iteration is:
[63]	train's binary_logloss: 0.555305	valid's binary_logloss: 0.653724
num_leaves, val_score: 0.653724:  10%|#         | 2/20 [00:01<00:14,  1.22it/s][I 2020-09-27 04:40:24,159] Trial 8 finished with value: 0.6537238118183929 and parameters: {'num_leaves': 71}. Best is trial 8 with value: 0.6537238118183929.
num_leaves, val_score: 0.653724:  10%|#         | 2/20 [00:01<00:14,  1.22it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000384 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.486851	valid's binary_logloss: 0.659048
Early stopping, best iteration is:
[64]	train's binary_logloss: 0.536523	valid's binary_logloss: 0.655529
num_leaves, val_score: 0.653724:  15%|#5        | 3/20 [00:02<00:13,  1.29it/s][I 2020-09-27 04:40:24,823] Trial 9 finished with value: 0.6555293293116815 and parameters: {'num_leaves': 85}. Best is trial 8 with value: 0.6537238118183929.
num_leaves, val_score: 0.653724:  15%|#5        | 3/20 [00:02<00:13,  1.29it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000459 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.668828	valid's binary_logloss: 0.671652
[200]	train's binary_logloss: 0.659076	valid's binary_logloss: 0.663493
[300]	train's binary_logloss: 0.653776	valid's binary_logloss: 0.659129
[400]	train's binary_logloss: 0.65062	valid's binary_logloss: 0.656814
[500]	train's binary_logloss: 0.648621	valid's binary_logloss: 0.655373
[600]	train's binary_logloss: 0.647288	valid's binary_logloss: 0.654268
[700]	train's binary_logloss: 0.646365	valid's binary_logloss: 0.653758
[800]	train's binary_logloss: 0.645671	valid's binary_logloss: 0.653488
[900]	train's binary_logloss: 0.645111	valid's binary_logloss: 0.653271
[1000]	train's binary_logloss: 0.644638	valid's binary_logloss: 0.653148
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.644638	valid's binary_logloss: 0.653148
num_leaves, val_score: 0.653148:  20%|##        | 4/20 [00:03<00:15,  1.02it/s][I 2020-09-27 04:40:26,293] Trial 10 finished with value: 0.6531478306778341 and parameters: {'num_leaves': 2}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  20%|##        | 4/20 [00:03<00:15,  1.02it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001421 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.647264	valid's binary_logloss: 0.65894
[200]	train's binary_logloss: 0.635718	valid's binary_logloss: 0.655218
[300]	train's binary_logloss: 0.627699	valid's binary_logloss: 0.654772
[400]	train's binary_logloss: 0.620605	valid's binary_logloss: 0.65463
Early stopping, best iteration is:
[337]	train's binary_logloss: 0.625076	valid's binary_logloss: 0.654329
num_leaves, val_score: 0.653148:  25%|##5       | 5/20 [00:04<00:14,  1.05it/s][I 2020-09-27 04:40:27,188] Trial 11 finished with value: 0.6543286934013544 and parameters: {'num_leaves': 5}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  25%|##5       | 5/20 [00:04<00:14,  1.05it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000570 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.631583	valid's binary_logloss: 0.655099
[200]	train's binary_logloss: 0.611849	valid's binary_logloss: 0.653955
Early stopping, best iteration is:
[189]	train's binary_logloss: 0.613791	valid's binary_logloss: 0.653569
num_leaves, val_score: 0.653148:  30%|###       | 6/20 [00:05<00:11,  1.22it/s][I 2020-09-27 04:40:27,696] Trial 12 finished with value: 0.6535687066231268 and parameters: {'num_leaves': 10}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  30%|###       | 6/20 [00:05<00:11,  1.22it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000425 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.634671	valid's binary_logloss: 0.655145
[200]	train's binary_logloss: 0.616535	valid's binary_logloss: 0.655474
Early stopping, best iteration is:
[160]	train's binary_logloss: 0.62285	valid's binary_logloss: 0.654646
num_leaves, val_score: 0.653148:  35%|###5      | 7/20 [00:05<00:09,  1.39it/s][I 2020-09-27 04:40:28,179] Trial 13 finished with value: 0.6546457645585857 and parameters: {'num_leaves': 9}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  35%|###5      | 7/20 [00:05<00:09,  1.39it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000412 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.668828	valid's binary_logloss: 0.671652
[200]	train's binary_logloss: 0.659076	valid's binary_logloss: 0.663493
[300]	train's binary_logloss: 0.653776	valid's binary_logloss: 0.659129
[400]	train's binary_logloss: 0.65062	valid's binary_logloss: 0.656814
[500]	train's binary_logloss: 0.648621	valid's binary_logloss: 0.655373
[600]	train's binary_logloss: 0.647288	valid's binary_logloss: 0.654268
[700]	train's binary_logloss: 0.646365	valid's binary_logloss: 0.653758
[800]	train's binary_logloss: 0.645671	valid's binary_logloss: 0.653488
[900]	train's binary_logloss: 0.645111	valid's binary_logloss: 0.653271
[1000]	train's binary_logloss: 0.644638	valid's binary_logloss: 0.653148
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.644638	valid's binary_logloss: 0.653148
num_leaves, val_score: 0.653148:  40%|####      | 8/20 [00:06<00:10,  1.12it/s][I 2020-09-27 04:40:29,465] Trial 14 finished with value: 0.6531478306778341 and parameters: {'num_leaves': 2}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  40%|####      | 8/20 [00:06<00:10,  1.12it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000471 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.306465	valid's binary_logloss: 0.670298
Early stopping, best iteration is:
[44]	train's binary_logloss: 0.45679	valid's binary_logloss: 0.663698
num_leaves, val_score: 0.653148:  45%|####5     | 9/20 [00:08<00:12,  1.15s/it][I 2020-09-27 04:40:31,226] Trial 15 finished with value: 0.663697988152403 and parameters: {'num_leaves': 248}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  45%|####5     | 9/20 [00:08<00:12,  1.15s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004276 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.520782	valid's binary_logloss: 0.657921
Early stopping, best iteration is:
[85]	train's binary_logloss: 0.536614	valid's binary_logloss: 0.656043
num_leaves, val_score: 0.653148:  50%|#####     | 10/20 [00:09<00:09,  1.03it/s][I 2020-09-27 04:40:31,778] Trial 16 finished with value: 0.6560430059842651 and parameters: {'num_leaves': 64}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  50%|#####     | 10/20 [00:09<00:09,  1.03it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000459 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.414427	valid's binary_logloss: 0.657771
Early stopping, best iteration is:
[82]	train's binary_logloss: 0.446068	valid's binary_logloss: 0.655491
num_leaves, val_score: 0.653148:  55%|#####5    | 11/20 [00:10<00:08,  1.03it/s][I 2020-09-27 04:40:32,746] Trial 17 finished with value: 0.6554905655085496 and parameters: {'num_leaves': 138}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  55%|#####5    | 11/20 [00:10<00:08,  1.03it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000382 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.57465	valid's binary_logloss: 0.656774
[200]	train's binary_logloss: 0.519325	valid's binary_logloss: 0.66119
Early stopping, best iteration is:
[115]	train's binary_logloss: 0.565412	valid's binary_logloss: 0.656102
num_leaves, val_score: 0.653148:  60%|######    | 12/20 [00:10<00:06,  1.20it/s][I 2020-09-27 04:40:33,269] Trial 18 finished with value: 0.6561023872985662 and parameters: {'num_leaves': 35}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  60%|######    | 12/20 [00:10<00:06,  1.20it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000377 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.431599	valid's binary_logloss: 0.661998
Early stopping, best iteration is:
[51]	train's binary_logloss: 0.522437	valid's binary_logloss: 0.656248
num_leaves, val_score: 0.653148:  65%|######5   | 13/20 [00:11<00:06,  1.03it/s][I 2020-09-27 04:40:34,569] Trial 19 finished with value: 0.6562478660337796 and parameters: {'num_leaves': 124}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  65%|######5   | 13/20 [00:11<00:06,  1.03it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000466 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.563816	valid's binary_logloss: 0.657169
Early stopping, best iteration is:
[88]	train's binary_logloss: 0.572908	valid's binary_logloss: 0.656464
num_leaves, val_score: 0.653148:  70%|#######   | 14/20 [00:12<00:05,  1.20it/s][I 2020-09-27 04:40:35,073] Trial 20 finished with value: 0.6564643963450036 and parameters: {'num_leaves': 40}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  70%|#######   | 14/20 [00:12<00:05,  1.20it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000380 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.668828	valid's binary_logloss: 0.671652
[200]	train's binary_logloss: 0.659076	valid's binary_logloss: 0.663493
[300]	train's binary_logloss: 0.653776	valid's binary_logloss: 0.659129
[400]	train's binary_logloss: 0.65062	valid's binary_logloss: 0.656814
[500]	train's binary_logloss: 0.648621	valid's binary_logloss: 0.655373
[600]	train's binary_logloss: 0.647288	valid's binary_logloss: 0.654268
[700]	train's binary_logloss: 0.646365	valid's binary_logloss: 0.653758
[800]	train's binary_logloss: 0.645671	valid's binary_logloss: 0.653488
[900]	train's binary_logloss: 0.645111	valid's binary_logloss: 0.653271
[1000]	train's binary_logloss: 0.644638	valid's binary_logloss: 0.653148
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.644638	valid's binary_logloss: 0.653148
num_leaves, val_score: 0.653148:  75%|#######5  | 15/20 [00:13<00:04,  1.06it/s][I 2020-09-27 04:40:36,259] Trial 21 finished with value: 0.6531478306778341 and parameters: {'num_leaves': 2}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  75%|#######5  | 15/20 [00:13<00:04,  1.06it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010841 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.63738	valid's binary_logloss: 0.656329
[200]	train's binary_logloss: 0.621173	valid's binary_logloss: 0.654961
[300]	train's binary_logloss: 0.608196	valid's binary_logloss: 0.655376
Early stopping, best iteration is:
[225]	train's binary_logloss: 0.617766	valid's binary_logloss: 0.654241
num_leaves, val_score: 0.653148:  80%|########  | 16/20 [00:14<00:03,  1.14it/s][I 2020-09-27 04:40:36,999] Trial 22 finished with value: 0.654241444152379 and parameters: {'num_leaves': 8}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  80%|########  | 16/20 [00:14<00:03,  1.14it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000584 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.576413	valid's binary_logloss: 0.656352
Early stopping, best iteration is:
[85]	train's binary_logloss: 0.58641	valid's binary_logloss: 0.655563
num_leaves, val_score: 0.653148:  85%|########5 | 17/20 [00:15<00:02,  1.02it/s][I 2020-09-27 04:40:38,198] Trial 23 finished with value: 0.6555629950352285 and parameters: {'num_leaves': 34}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  85%|########5 | 17/20 [00:15<00:02,  1.02it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000677 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.647264	valid's binary_logloss: 0.65894
[200]	train's binary_logloss: 0.635718	valid's binary_logloss: 0.655218
[300]	train's binary_logloss: 0.627699	valid's binary_logloss: 0.654772
[400]	train's binary_logloss: 0.620605	valid's binary_logloss: 0.65463
Early stopping, best iteration is:
[337]	train's binary_logloss: 0.625076	valid's binary_logloss: 0.654329
num_leaves, val_score: 0.653148:  90%|######### | 18/20 [00:16<00:01,  1.07it/s][I 2020-09-27 04:40:39,032] Trial 24 finished with value: 0.6543286934013544 and parameters: {'num_leaves': 5}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  90%|######### | 18/20 [00:16<00:01,  1.07it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000412 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.557014	valid's binary_logloss: 0.658549
Early stopping, best iteration is:
[46]	train's binary_logloss: 0.606508	valid's binary_logloss: 0.656484
num_leaves, val_score: 0.653148:  95%|#########5| 19/20 [00:16<00:00,  1.23it/s][I 2020-09-27 04:40:39,562] Trial 25 finished with value: 0.6564836527501194 and parameters: {'num_leaves': 44}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148:  95%|#########5| 19/20 [00:16<00:00,  1.23it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000389 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.455025	valid's binary_logloss: 0.660325
Early stopping, best iteration is:
[39]	train's binary_logloss: 0.562754	valid's binary_logloss: 0.655052
num_leaves, val_score: 0.653148: 100%|##########| 20/20 [00:17<00:00,  1.29it/s][I 2020-09-27 04:40:40,252] Trial 26 finished with value: 0.655051684902722 and parameters: {'num_leaves': 107}. Best is trial 10 with value: 0.6531478306778341.
num_leaves, val_score: 0.653148: 100%|##########| 20/20 [00:17<00:00,  1.14it/s]
bagging, val_score: 0.653148:   0%|          | 0/10 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000406 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.668417	valid's binary_logloss: 0.671074
[200]	train's binary_logloss: 0.658108	valid's binary_logloss: 0.662187
[300]	train's binary_logloss: 0.652572	valid's binary_logloss: 0.657758
[400]	train's binary_logloss: 0.649457	valid's binary_logloss: 0.655824
[500]	train's binary_logloss: 0.64748	valid's binary_logloss: 0.654202
[600]	train's binary_logloss: 0.646194	valid's binary_logloss: 0.653541
[700]	train's binary_logloss: 0.645284	valid's binary_logloss: 0.65328
[800]	train's binary_logloss: 0.644551	valid's binary_logloss: 0.653263
[900]	train's binary_logloss: 0.64391	valid's binary_logloss: 0.653167
[1000]	train's binary_logloss: 0.643382	valid's binary_logloss: 0.653181
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643382	valid's binary_logloss: 0.653181
bagging, val_score: 0.653148:  10%|#         | 1/10 [00:01<00:14,  1.60s/it][I 2020-09-27 04:40:41,874] Trial 27 finished with value: 0.6531811973594188 and parameters: {'bagging_fraction': 0.8306103567609042, 'bagging_freq': 5}. Best is trial 27 with value: 0.6531811973594188.
bagging, val_score: 0.653148:  10%|#         | 1/10 [00:01<00:14,  1.60s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000338 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.666533	valid's binary_logloss: 0.669291
[200]	train's binary_logloss: 0.655913	valid's binary_logloss: 0.659959
[300]	train's binary_logloss: 0.650521	valid's binary_logloss: 0.656005
[400]	train's binary_logloss: 0.647915	valid's binary_logloss: 0.654692
[500]	train's binary_logloss: 0.646471	valid's binary_logloss: 0.65411
[600]	train's binary_logloss: 0.645402	valid's binary_logloss: 0.654374
Early stopping, best iteration is:
[511]	train's binary_logloss: 0.646347	valid's binary_logloss: 0.653867
bagging, val_score: 0.653148:  20%|##        | 2/10 [00:02<00:11,  1.42s/it][I 2020-09-27 04:40:42,870] Trial 28 finished with value: 0.6538669216016616 and parameters: {'bagging_fraction': 0.4517567998193014, 'bagging_freq': 1}. Best is trial 27 with value: 0.6531811973594188.
bagging, val_score: 0.653148:  20%|##        | 2/10 [00:02<00:11,  1.42s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000347 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667071	valid's binary_logloss: 0.669326
[200]	train's binary_logloss: 0.656588	valid's binary_logloss: 0.660429
[300]	train's binary_logloss: 0.651211	valid's binary_logloss: 0.656184
[400]	train's binary_logloss: 0.648478	valid's binary_logloss: 0.653762
[500]	train's binary_logloss: 0.646843	valid's binary_logloss: 0.653211
[600]	train's binary_logloss: 0.645633	valid's binary_logloss: 0.652424
Early stopping, best iteration is:
[573]	train's binary_logloss: 0.645934	valid's binary_logloss: 0.652252
bagging, val_score: 0.652252:  30%|###       | 3/10 [00:03<00:08,  1.25s/it][I 2020-09-27 04:40:43,705] Trial 29 finished with value: 0.65225160633841 and parameters: {'bagging_fraction': 0.4737096463396763, 'bagging_freq': 7}. Best is trial 29 with value: 0.65225160633841.
bagging, val_score: 0.652252:  30%|###       | 3/10 [00:03<00:08,  1.25s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000502 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.666936	valid's binary_logloss: 0.668846
[200]	train's binary_logloss: 0.656163	valid's binary_logloss: 0.659914
[300]	train's binary_logloss: 0.650959	valid's binary_logloss: 0.655368
[400]	train's binary_logloss: 0.648469	valid's binary_logloss: 0.653546
[500]	train's binary_logloss: 0.646883	valid's binary_logloss: 0.653403
Early stopping, best iteration is:
[438]	train's binary_logloss: 0.647744	valid's binary_logloss: 0.652963
bagging, val_score: 0.652252:  40%|####      | 4/10 [00:04<00:06,  1.07s/it][I 2020-09-27 04:40:44,379] Trial 30 finished with value: 0.6529627183444107 and parameters: {'bagging_fraction': 0.4264803439292804, 'bagging_freq': 7}. Best is trial 29 with value: 0.65225160633841.
bagging, val_score: 0.652252:  40%|####      | 4/10 [00:04<00:06,  1.07s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000349 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.666824	valid's binary_logloss: 0.668808
[200]	train's binary_logloss: 0.656165	valid's binary_logloss: 0.65966
[300]	train's binary_logloss: 0.650811	valid's binary_logloss: 0.654901
[400]	train's binary_logloss: 0.648589	valid's binary_logloss: 0.652893
[500]	train's binary_logloss: 0.647086	valid's binary_logloss: 0.653134
[600]	train's binary_logloss: 0.645858	valid's binary_logloss: 0.652547
[700]	train's binary_logloss: 0.644904	valid's binary_logloss: 0.65326
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.64558	valid's binary_logloss: 0.652267
bagging, val_score: 0.652252:  50%|#####     | 5/10 [00:05<00:05,  1.18s/it][I 2020-09-27 04:40:45,808] Trial 31 finished with value: 0.6522673041330311 and parameters: {'bagging_fraction': 0.41942184957156203, 'bagging_freq': 7}. Best is trial 29 with value: 0.65225160633841.
bagging, val_score: 0.652252:  50%|#####     | 5/10 [00:05<00:05,  1.18s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000470 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.666715	valid's binary_logloss: 0.669238
[200]	train's binary_logloss: 0.656103	valid's binary_logloss: 0.660182
[300]	train's binary_logloss: 0.650838	valid's binary_logloss: 0.655485
[400]	train's binary_logloss: 0.648497	valid's binary_logloss: 0.653732
[500]	train's binary_logloss: 0.646909	valid's binary_logloss: 0.653654
Early stopping, best iteration is:
[470]	train's binary_logloss: 0.6473	valid's binary_logloss: 0.652896
bagging, val_score: 0.652252:  60%|######    | 6/10 [00:06<00:04,  1.07s/it][I 2020-09-27 04:40:46,607] Trial 32 finished with value: 0.6528955403221239 and parameters: {'bagging_fraction': 0.41256437409812413, 'bagging_freq': 7}. Best is trial 29 with value: 0.65225160633841.
bagging, val_score: 0.652252:  60%|######    | 6/10 [00:06<00:04,  1.07s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000341 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.666696	valid's binary_logloss: 0.669282
[200]	train's binary_logloss: 0.656159	valid's binary_logloss: 0.660236
[300]	train's binary_logloss: 0.65083	valid's binary_logloss: 0.655589
[400]	train's binary_logloss: 0.648563	valid's binary_logloss: 0.653633
[500]	train's binary_logloss: 0.64694	valid's binary_logloss: 0.653341
Early stopping, best iteration is:
[470]	train's binary_logloss: 0.647303	valid's binary_logloss: 0.652827
bagging, val_score: 0.652252:  70%|#######   | 7/10 [00:06<00:02,  1.07it/s][I 2020-09-27 04:40:47,241] Trial 33 finished with value: 0.6528271831886884 and parameters: {'bagging_fraction': 0.40860834213805497, 'bagging_freq': 7}. Best is trial 29 with value: 0.65225160633841.
bagging, val_score: 0.652252:  70%|#######   | 7/10 [00:06<00:02,  1.07it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000446 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.666727	valid's binary_logloss: 0.669333
[200]	train's binary_logloss: 0.656066	valid's binary_logloss: 0.660295
[300]	train's binary_logloss: 0.650735	valid's binary_logloss: 0.655605
[400]	train's binary_logloss: 0.648394	valid's binary_logloss: 0.653499
[500]	train's binary_logloss: 0.646916	valid's binary_logloss: 0.653049
Early stopping, best iteration is:
[470]	train's binary_logloss: 0.647295	valid's binary_logloss: 0.652461
bagging, val_score: 0.652252:  80%|########  | 8/10 [00:07<00:01,  1.15it/s][I 2020-09-27 04:40:47,944] Trial 34 finished with value: 0.6524608085550598 and parameters: {'bagging_fraction': 0.40094624308690235, 'bagging_freq': 7}. Best is trial 29 with value: 0.65225160633841.
bagging, val_score: 0.652252:  80%|########  | 8/10 [00:07<00:01,  1.15it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000704 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.666735	valid's binary_logloss: 0.669024
[200]	train's binary_logloss: 0.656104	valid's binary_logloss: 0.660308
[300]	train's binary_logloss: 0.650686	valid's binary_logloss: 0.655603
[400]	train's binary_logloss: 0.648444	valid's binary_logloss: 0.653985
[500]	train's binary_logloss: 0.646852	valid's binary_logloss: 0.653461
Early stopping, best iteration is:
[470]	train's binary_logloss: 0.647262	valid's binary_logloss: 0.652979
bagging, val_score: 0.652252:  90%|######### | 9/10 [00:08<00:00,  1.21it/s][I 2020-09-27 04:40:48,677] Trial 35 finished with value: 0.6529787611145468 and parameters: {'bagging_fraction': 0.40896873900368547, 'bagging_freq': 7}. Best is trial 29 with value: 0.65225160633841.
bagging, val_score: 0.652252:  90%|######### | 9/10 [00:08<00:00,  1.21it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010501 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667557	valid's binary_logloss: 0.670365
[200]	train's binary_logloss: 0.656945	valid's binary_logloss: 0.660765
[300]	train's binary_logloss: 0.651474	valid's binary_logloss: 0.656544
[400]	train's binary_logloss: 0.648605	valid's binary_logloss: 0.653988
[500]	train's binary_logloss: 0.646848	valid's binary_logloss: 0.653526
[600]	train's binary_logloss: 0.645737	valid's binary_logloss: 0.653035
Early stopping, best iteration is:
[588]	train's binary_logloss: 0.645878	valid's binary_logloss: 0.652666
bagging, val_score: 0.652252: 100%|##########| 10/10 [00:09<00:00,  1.03it/s][I 2020-09-27 04:40:49,991] Trial 36 finished with value: 0.65266565454767 and parameters: {'bagging_fraction': 0.5936827208546511, 'bagging_freq': 7}. Best is trial 29 with value: 0.65225160633841.
bagging, val_score: 0.652252: 100%|##########| 10/10 [00:09<00:00,  1.03it/s]
feature_fraction_stage2, val_score: 0.652252:   0%|          | 0/3 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.004126 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667157	valid's binary_logloss: 0.669629
[200]	train's binary_logloss: 0.65666	valid's binary_logloss: 0.660691
[300]	train's binary_logloss: 0.651252	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.648542	valid's binary_logloss: 0.653738
[500]	train's binary_logloss: 0.646904	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.645658	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.644752	valid's binary_logloss: 0.652982
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645415	valid's binary_logloss: 0.652211
feature_fraction_stage2, val_score: 0.652211:  33%|###3      | 1/3 [00:00<00:01,  1.14it/s][I 2020-09-27 04:40:50,883] Trial 37 finished with value: 0.6522110872949437 and parameters: {'feature_fraction': 0.41600000000000004}. Best is trial 37 with value: 0.6522110872949437.
feature_fraction_stage2, val_score: 0.652211:  33%|###3      | 1/3 [00:00<00:01,  1.14it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000546 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667214	valid's binary_logloss: 0.669614
[200]	train's binary_logloss: 0.656661	valid's binary_logloss: 0.660306
[300]	train's binary_logloss: 0.651258	valid's binary_logloss: 0.656152
[400]	train's binary_logloss: 0.648532	valid's binary_logloss: 0.653868
[500]	train's binary_logloss: 0.64686	valid's binary_logloss: 0.65353
[600]	train's binary_logloss: 0.645694	valid's binary_logloss: 0.652608
[700]	train's binary_logloss: 0.644833	valid's binary_logloss: 0.653223
Early stopping, best iteration is:
[622]	train's binary_logloss: 0.645442	valid's binary_logloss: 0.652368
feature_fraction_stage2, val_score: 0.652211:  67%|######6   | 2/3 [00:01<00:00,  1.14it/s][I 2020-09-27 04:40:51,765] Trial 38 finished with value: 0.6523684425922491 and parameters: {'feature_fraction': 0.44800000000000006}. Best is trial 37 with value: 0.6522110872949437.
feature_fraction_stage2, val_score: 0.652211:  67%|######6   | 2/3 [00:01<00:00,  1.14it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000441 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667214	valid's binary_logloss: 0.669614
[200]	train's binary_logloss: 0.656661	valid's binary_logloss: 0.660306
[300]	train's binary_logloss: 0.651258	valid's binary_logloss: 0.656152
[400]	train's binary_logloss: 0.648532	valid's binary_logloss: 0.653868
[500]	train's binary_logloss: 0.64686	valid's binary_logloss: 0.65353
[600]	train's binary_logloss: 0.645694	valid's binary_logloss: 0.652608
[700]	train's binary_logloss: 0.644833	valid's binary_logloss: 0.653223
Early stopping, best iteration is:
[622]	train's binary_logloss: 0.645442	valid's binary_logloss: 0.652368
feature_fraction_stage2, val_score: 0.652211: 100%|##########| 3/3 [00:02<00:00,  1.13it/s][I 2020-09-27 04:40:52,662] Trial 39 finished with value: 0.6523684425922491 and parameters: {'feature_fraction': 0.48000000000000004}. Best is trial 37 with value: 0.6522110872949437.
feature_fraction_stage2, val_score: 0.652211: 100%|##########| 3/3 [00:02<00:00,  1.12it/s]
regularization_factors, val_score: 0.652211:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000446 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667157	valid's binary_logloss: 0.669629
[200]	train's binary_logloss: 0.65666	valid's binary_logloss: 0.660691
[300]	train's binary_logloss: 0.651252	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.648542	valid's binary_logloss: 0.653738
[500]	train's binary_logloss: 0.646905	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.645659	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.644752	valid's binary_logloss: 0.652981
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645415	valid's binary_logloss: 0.652211
regularization_factors, val_score: 0.652211:   5%|5         | 1/20 [00:01<00:25,  1.34s/it][I 2020-09-27 04:40:54,031] Trial 40 finished with value: 0.6522109977626771 and parameters: {'lambda_l1': 0.00014164147515972837, 'lambda_l2': 0.005481479902021038}. Best is trial 40 with value: 0.6522109977626771.
regularization_factors, val_score: 0.652211:   5%|5         | 1/20 [00:01<00:25,  1.34s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004749 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667157	valid's binary_logloss: 0.669629
[200]	train's binary_logloss: 0.656661	valid's binary_logloss: 0.660691
[300]	train's binary_logloss: 0.651253	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.648543	valid's binary_logloss: 0.653738
[500]	train's binary_logloss: 0.646906	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.64566	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.644753	valid's binary_logloss: 0.652981
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645416	valid's binary_logloss: 0.652211
regularization_factors, val_score: 0.652211:  10%|#         | 2/20 [00:02<00:21,  1.20s/it][I 2020-09-27 04:40:54,895] Trial 41 finished with value: 0.6522109708484218 and parameters: {'lambda_l1': 0.00010302212398099651, 'lambda_l2': 0.007995173172021716}. Best is trial 41 with value: 0.6522109708484218.
regularization_factors, val_score: 0.652211:  10%|#         | 2/20 [00:02<00:21,  1.20s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000406 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667157	valid's binary_logloss: 0.669629
[200]	train's binary_logloss: 0.656661	valid's binary_logloss: 0.660691
[300]	train's binary_logloss: 0.651252	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.648542	valid's binary_logloss: 0.653738
[500]	train's binary_logloss: 0.646905	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.645659	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.644752	valid's binary_logloss: 0.652981
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645415	valid's binary_logloss: 0.652211
regularization_factors, val_score: 0.652211:  15%|#5        | 3/20 [00:03<00:18,  1.10s/it][I 2020-09-27 04:40:55,766] Trial 42 finished with value: 0.6522109575004142 and parameters: {'lambda_l1': 5.703427719344543e-05, 'lambda_l2': 0.01133484756475216}. Best is trial 42 with value: 0.6522109575004142.
regularization_factors, val_score: 0.652211:  15%|#5        | 3/20 [00:03<00:18,  1.10s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001530 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667157	valid's binary_logloss: 0.669629
[200]	train's binary_logloss: 0.656661	valid's binary_logloss: 0.660691
[300]	train's binary_logloss: 0.651252	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.648542	valid's binary_logloss: 0.653738
[500]	train's binary_logloss: 0.646905	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.645659	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.644753	valid's binary_logloss: 0.652981
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645416	valid's binary_logloss: 0.652211
regularization_factors, val_score: 0.652211:  20%|##        | 4/20 [00:04<00:17,  1.08s/it][I 2020-09-27 04:40:56,793] Trial 43 finished with value: 0.6522109461124184 and parameters: {'lambda_l1': 7.037388835726531e-05, 'lambda_l2': 0.009023034240515858}. Best is trial 43 with value: 0.6522109461124184.
regularization_factors, val_score: 0.652211:  20%|##        | 4/20 [00:04<00:17,  1.08s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.017232 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667157	valid's binary_logloss: 0.669629
[200]	train's binary_logloss: 0.656661	valid's binary_logloss: 0.660691
[300]	train's binary_logloss: 0.651252	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.648542	valid's binary_logloss: 0.653738
[500]	train's binary_logloss: 0.646905	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.645659	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.644753	valid's binary_logloss: 0.652981
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645416	valid's binary_logloss: 0.652211
regularization_factors, val_score: 0.652211:  25%|##5       | 5/20 [00:05<00:16,  1.12s/it][I 2020-09-27 04:40:57,991] Trial 44 finished with value: 0.6522108862289743 and parameters: {'lambda_l1': 4.695264971426557e-05, 'lambda_l2': 0.012990997644376091}. Best is trial 44 with value: 0.6522108862289743.
regularization_factors, val_score: 0.652211:  25%|##5       | 5/20 [00:05<00:16,  1.12s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000418 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667157	valid's binary_logloss: 0.669629
[200]	train's binary_logloss: 0.656661	valid's binary_logloss: 0.660691
[300]	train's binary_logloss: 0.651252	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.648542	valid's binary_logloss: 0.653738
[500]	train's binary_logloss: 0.646905	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.645659	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.644753	valid's binary_logloss: 0.652981
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645416	valid's binary_logloss: 0.652211
regularization_factors, val_score: 0.652211:  30%|###       | 6/20 [00:06<00:14,  1.05s/it][I 2020-09-27 04:40:58,893] Trial 45 finished with value: 0.6522109405266442 and parameters: {'lambda_l1': 6.628550474869429e-05, 'lambda_l2': 0.00939838589902522}. Best is trial 44 with value: 0.6522108862289743.
regularization_factors, val_score: 0.652211:  30%|###       | 6/20 [00:06<00:14,  1.05s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000752 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667157	valid's binary_logloss: 0.669629
[200]	train's binary_logloss: 0.656661	valid's binary_logloss: 0.660691
[300]	train's binary_logloss: 0.651252	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.648542	valid's binary_logloss: 0.653738
[500]	train's binary_logloss: 0.646905	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.645659	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.644753	valid's binary_logloss: 0.652981
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645416	valid's binary_logloss: 0.652211
regularization_factors, val_score: 0.652211:  35%|###5      | 7/20 [00:07<00:13,  1.01s/it][I 2020-09-27 04:40:59,810] Trial 46 finished with value: 0.6522109149315195 and parameters: {'lambda_l1': 5.01427220376058e-05, 'lambda_l2': 0.011108962252046103}. Best is trial 44 with value: 0.6522108862289743.
regularization_factors, val_score: 0.652211:  35%|###5      | 7/20 [00:07<00:13,  1.01s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000545 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667157	valid's binary_logloss: 0.66963
[200]	train's binary_logloss: 0.656661	valid's binary_logloss: 0.660691
[300]	train's binary_logloss: 0.651253	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.648542	valid's binary_logloss: 0.653739
[500]	train's binary_logloss: 0.646905	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.64566	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.644753	valid's binary_logloss: 0.652981
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645416	valid's binary_logloss: 0.652211
regularization_factors, val_score: 0.652211:  40%|####      | 8/20 [00:08<00:12,  1.03s/it][I 2020-09-27 04:41:00,882] Trial 47 finished with value: 0.6522108308126086 and parameters: {'lambda_l1': 4.076885797493761e-05, 'lambda_l2': 0.016633645127292034}. Best is trial 47 with value: 0.6522108308126086.
regularization_factors, val_score: 0.652211:  40%|####      | 8/20 [00:08<00:12,  1.03s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000476 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667157	valid's binary_logloss: 0.66963
[200]	train's binary_logloss: 0.656661	valid's binary_logloss: 0.660691
[300]	train's binary_logloss: 0.651253	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.648543	valid's binary_logloss: 0.653739
[500]	train's binary_logloss: 0.646906	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.64566	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.644754	valid's binary_logloss: 0.652981
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645417	valid's binary_logloss: 0.652211
regularization_factors, val_score: 0.652211:  45%|####5     | 9/20 [00:09<00:11,  1.04s/it][I 2020-09-27 04:41:01,942] Trial 48 finished with value: 0.6522107665018237 and parameters: {'lambda_l1': 4.1633476557442786e-05, 'lambda_l2': 0.020110745479315996}. Best is trial 48 with value: 0.6522107665018237.
regularization_factors, val_score: 0.652211:  45%|####5     | 9/20 [00:09<00:11,  1.04s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000514 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667174	valid's binary_logloss: 0.669606
[200]	train's binary_logloss: 0.656682	valid's binary_logloss: 0.660677
[300]	train's binary_logloss: 0.651269	valid's binary_logloss: 0.656126
[400]	train's binary_logloss: 0.64857	valid's binary_logloss: 0.653596
[500]	train's binary_logloss: 0.646941	valid's binary_logloss: 0.653105
[600]	train's binary_logloss: 0.645643	valid's binary_logloss: 0.652442
[700]	train's binary_logloss: 0.644768	valid's binary_logloss: 0.653208
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645403	valid's binary_logloss: 0.652216
regularization_factors, val_score: 0.652211:  50%|#####     | 10/20 [00:10<00:09,  1.02it/s][I 2020-09-27 04:41:02,794] Trial 49 finished with value: 0.6522161063945356 and parameters: {'lambda_l1': 1.0048108252149604e-05, 'lambda_l2': 0.04595411901882225}. Best is trial 48 with value: 0.6522107665018237.
regularization_factors, val_score: 0.652211:  50%|#####     | 10/20 [00:10<00:09,  1.02it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000250 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667264	valid's binary_logloss: 0.669374
[200]	train's binary_logloss: 0.656765	valid's binary_logloss: 0.660326
[300]	train's binary_logloss: 0.651459	valid's binary_logloss: 0.656068
[400]	train's binary_logloss: 0.648733	valid's binary_logloss: 0.65386
[500]	train's binary_logloss: 0.647116	valid's binary_logloss: 0.653305
[600]	train's binary_logloss: 0.645917	valid's binary_logloss: 0.652237
[700]	train's binary_logloss: 0.645032	valid's binary_logloss: 0.652691
Early stopping, best iteration is:
[622]	train's binary_logloss: 0.645713	valid's binary_logloss: 0.652093
regularization_factors, val_score: 0.652093:  55%|#####5    | 11/20 [00:10<00:08,  1.06it/s][I 2020-09-27 04:41:03,645] Trial 50 finished with value: 0.6520932820946773 and parameters: {'lambda_l1': 5.634084468530463e-06, 'lambda_l2': 2.653747794681544}. Best is trial 50 with value: 0.6520932820946773.
regularization_factors, val_score: 0.652093:  55%|#####5    | 11/20 [00:10<00:08,  1.06it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.002392 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667312	valid's binary_logloss: 0.66942
[200]	train's binary_logloss: 0.656827	valid's binary_logloss: 0.660345
[300]	train's binary_logloss: 0.651466	valid's binary_logloss: 0.656086
[400]	train's binary_logloss: 0.648842	valid's binary_logloss: 0.653957
[500]	train's binary_logloss: 0.647208	valid's binary_logloss: 0.653403
[600]	train's binary_logloss: 0.646029	valid's binary_logloss: 0.652507
[700]	train's binary_logloss: 0.645184	valid's binary_logloss: 0.653114
Early stopping, best iteration is:
[622]	train's binary_logloss: 0.645818	valid's binary_logloss: 0.652395
regularization_factors, val_score: 0.652093:  60%|######    | 12/20 [00:12<00:08,  1.04s/it][I 2020-09-27 04:41:04,925] Trial 51 finished with value: 0.6523945935028854 and parameters: {'lambda_l1': 7.71584097888483e-06, 'lambda_l2': 4.4305490881206495}. Best is trial 50 with value: 0.6520932820946773.
regularization_factors, val_score: 0.652093:  60%|######    | 12/20 [00:12<00:08,  1.04s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.013219 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667193	valid's binary_logloss: 0.669624
[200]	train's binary_logloss: 0.656716	valid's binary_logloss: 0.660578
[300]	train's binary_logloss: 0.651263	valid's binary_logloss: 0.656044
[400]	train's binary_logloss: 0.648561	valid's binary_logloss: 0.654042
[500]	train's binary_logloss: 0.646891	valid's binary_logloss: 0.653546
[600]	train's binary_logloss: 0.645722	valid's binary_logloss: 0.652543
[700]	train's binary_logloss: 0.644815	valid's binary_logloss: 0.653225
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645497	valid's binary_logloss: 0.652335
regularization_factors, val_score: 0.652093:  65%|######5   | 13/20 [00:13<00:08,  1.16s/it][I 2020-09-27 04:41:06,349] Trial 52 finished with value: 0.6523351333047752 and parameters: {'lambda_l1': 7.422157693379738e-06, 'lambda_l2': 0.7332370123602728}. Best is trial 50 with value: 0.6520932820946773.
regularization_factors, val_score: 0.652093:  65%|######5   | 13/20 [00:13<00:08,  1.16s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000421 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667345	valid's binary_logloss: 0.66941
[200]	train's binary_logloss: 0.656873	valid's binary_logloss: 0.660467
[300]	train's binary_logloss: 0.651513	valid's binary_logloss: 0.655838
[400]	train's binary_logloss: 0.648818	valid's binary_logloss: 0.653846
[500]	train's binary_logloss: 0.647188	valid's binary_logloss: 0.653186
[600]	train's binary_logloss: 0.646098	valid's binary_logloss: 0.652539
Early stopping, best iteration is:
[573]	train's binary_logloss: 0.646377	valid's binary_logloss: 0.652312
regularization_factors, val_score: 0.652093:  70%|#######   | 14/20 [00:14<00:06,  1.07s/it][I 2020-09-27 04:41:07,202] Trial 53 finished with value: 0.6523123313741976 and parameters: {'lambda_l1': 1.6736112176387592, 'lambda_l2': 3.1184950223667215e-07}. Best is trial 50 with value: 0.6520932820946773.
regularization_factors, val_score: 0.652093:  70%|#######   | 14/20 [00:14<00:06,  1.07s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.002056 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667157	valid's binary_logloss: 0.669629
[200]	train's binary_logloss: 0.65666	valid's binary_logloss: 0.660691
[300]	train's binary_logloss: 0.651252	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.648542	valid's binary_logloss: 0.653738
[500]	train's binary_logloss: 0.646905	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.645659	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.644752	valid's binary_logloss: 0.652981
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645415	valid's binary_logloss: 0.652211
regularization_factors, val_score: 0.652093:  75%|#######5  | 15/20 [00:15<00:05,  1.02s/it][I 2020-09-27 04:41:08,102] Trial 54 finished with value: 0.6522110253635155 and parameters: {'lambda_l1': 0.0016228255112486223, 'lambda_l2': 8.007222095393149e-05}. Best is trial 50 with value: 0.6520932820946773.
regularization_factors, val_score: 0.652093:  75%|#######5  | 15/20 [00:15<00:05,  1.02s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000382 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667157	valid's binary_logloss: 0.669629
[200]	train's binary_logloss: 0.65666	valid's binary_logloss: 0.660691
[300]	train's binary_logloss: 0.651252	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.648542	valid's binary_logloss: 0.653738
[500]	train's binary_logloss: 0.646904	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.645658	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.644752	valid's binary_logloss: 0.652982
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645415	valid's binary_logloss: 0.652211
regularization_factors, val_score: 0.652093:  80%|########  | 16/20 [00:16<00:04,  1.09s/it][I 2020-09-27 04:41:09,376] Trial 55 finished with value: 0.6522110831327725 and parameters: {'lambda_l1': 8.98931163949704e-08, 'lambda_l2': 0.0002694496140730187}. Best is trial 50 with value: 0.6520932820946773.
regularization_factors, val_score: 0.652093:  80%|########  | 16/20 [00:16<00:04,  1.09s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000250 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.66718	valid's binary_logloss: 0.669612
[200]	train's binary_logloss: 0.656691	valid's binary_logloss: 0.660683
[300]	train's binary_logloss: 0.651278	valid's binary_logloss: 0.656131
[400]	train's binary_logloss: 0.648583	valid's binary_logloss: 0.653603
[500]	train's binary_logloss: 0.646957	valid's binary_logloss: 0.653104
[600]	train's binary_logloss: 0.645691	valid's binary_logloss: 0.652458
[700]	train's binary_logloss: 0.644799	valid's binary_logloss: 0.653165
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645445	valid's binary_logloss: 0.65223
regularization_factors, val_score: 0.652093:  85%|########5 | 17/20 [00:17<00:03,  1.03s/it][I 2020-09-27 04:41:10,268] Trial 56 finished with value: 0.6522296062194786 and parameters: {'lambda_l1': 1.272264047860099e-06, 'lambda_l2': 0.29472834124394065}. Best is trial 50 with value: 0.6520932820946773.
regularization_factors, val_score: 0.652093:  85%|########5 | 17/20 [00:17<00:03,  1.03s/it][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000520 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667157	valid's binary_logloss: 0.669629
[200]	train's binary_logloss: 0.656661	valid's binary_logloss: 0.660691
[300]	train's binary_logloss: 0.651252	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.648542	valid's binary_logloss: 0.653738
[500]	train's binary_logloss: 0.646905	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.645659	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.644752	valid's binary_logloss: 0.652981
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645415	valid's binary_logloss: 0.652211
regularization_factors, val_score: 0.652093:  90%|######### | 18/20 [00:18<00:01,  1.02it/s][I 2020-09-27 04:41:11,137] Trial 57 finished with value: 0.6522109880071963 and parameters: {'lambda_l1': 0.0023192016556508074, 'lambda_l2': 0.0008178552198766434}. Best is trial 50 with value: 0.6520932820946773.
regularization_factors, val_score: 0.652093:  90%|######### | 18/20 [00:18<00:01,  1.02it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000580 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667176	valid's binary_logloss: 0.669608
[200]	train's binary_logloss: 0.656684	valid's binary_logloss: 0.660679
[300]	train's binary_logloss: 0.651272	valid's binary_logloss: 0.656127
[400]	train's binary_logloss: 0.648574	valid's binary_logloss: 0.653598
[500]	train's binary_logloss: 0.646946	valid's binary_logloss: 0.653105
[600]	train's binary_logloss: 0.645649	valid's binary_logloss: 0.652439
[700]	train's binary_logloss: 0.644774	valid's binary_logloss: 0.653205
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.64541	valid's binary_logloss: 0.652214
regularization_factors, val_score: 0.652093:  95%|#########5| 19/20 [00:19<00:00,  1.05it/s][I 2020-09-27 04:41:12,010] Trial 58 finished with value: 0.6522142117867799 and parameters: {'lambda_l1': 0.0007602407297528348, 'lambda_l2': 0.11500147374550874}. Best is trial 50 with value: 0.6520932820946773.
regularization_factors, val_score: 0.652093:  95%|#########5| 19/20 [00:19<00:00,  1.05it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005008 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667156	valid's binary_logloss: 0.669629
[200]	train's binary_logloss: 0.656659	valid's binary_logloss: 0.66069
[300]	train's binary_logloss: 0.651251	valid's binary_logloss: 0.656139
[400]	train's binary_logloss: 0.64854	valid's binary_logloss: 0.653738
[500]	train's binary_logloss: 0.646903	valid's binary_logloss: 0.653287
[600]	train's binary_logloss: 0.645657	valid's binary_logloss: 0.652441
[700]	train's binary_logloss: 0.64475	valid's binary_logloss: 0.652981
Early stopping, best iteration is:
[623]	train's binary_logloss: 0.645413	valid's binary_logloss: 0.652211
regularization_factors, val_score: 0.652093: 100%|##########| 20/20 [00:20<00:00,  1.07s/it][I 2020-09-27 04:41:13,360] Trial 59 finished with value: 0.6522109875781879 and parameters: {'lambda_l1': 1.3713971348787328e-06, 'lambda_l2': 0.0013479533247432005}. Best is trial 50 with value: 0.6520932820946773.
regularization_factors, val_score: 0.652093: 100%|##########| 20/20 [00:20<00:00,  1.03s/it]
min_data_in_leaf, val_score: 0.652093:   0%|          | 0/5 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004509 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667264	valid's binary_logloss: 0.669374
[200]	train's binary_logloss: 0.656765	valid's binary_logloss: 0.660326
[300]	train's binary_logloss: 0.651459	valid's binary_logloss: 0.656068
[400]	train's binary_logloss: 0.648733	valid's binary_logloss: 0.653788
[500]	train's binary_logloss: 0.647111	valid's binary_logloss: 0.653197
[600]	train's binary_logloss: 0.645996	valid's binary_logloss: 0.652398
Early stopping, best iteration is:
[589]	train's binary_logloss: 0.646096	valid's binary_logloss: 0.652267
min_data_in_leaf, val_score: 0.652093:  20%|##        | 1/5 [00:00<00:03,  1.23it/s][I 2020-09-27 04:41:14,186] Trial 60 finished with value: 0.6522671792511908 and parameters: {'min_child_samples': 50}. Best is trial 60 with value: 0.6522671792511908.
min_data_in_leaf, val_score: 0.652093:  20%|##        | 1/5 [00:00<00:03,  1.23it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000404 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667264	valid's binary_logloss: 0.669374
[200]	train's binary_logloss: 0.656765	valid's binary_logloss: 0.660326
[300]	train's binary_logloss: 0.651459	valid's binary_logloss: 0.656068
[400]	train's binary_logloss: 0.648733	valid's binary_logloss: 0.653788
[500]	train's binary_logloss: 0.647134	valid's binary_logloss: 0.653299
[600]	train's binary_logloss: 0.645934	valid's binary_logloss: 0.652389
Early stopping, best iteration is:
[573]	train's binary_logloss: 0.646258	valid's binary_logloss: 0.652282
min_data_in_leaf, val_score: 0.652093:  40%|####      | 2/5 [00:01<00:02,  1.24it/s][I 2020-09-27 04:41:14,982] Trial 61 finished with value: 0.6522824409223347 and parameters: {'min_child_samples': 25}. Best is trial 60 with value: 0.6522671792511908.
min_data_in_leaf, val_score: 0.652093:  40%|####      | 2/5 [00:01<00:02,  1.24it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001187 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667264	valid's binary_logloss: 0.669374
[200]	train's binary_logloss: 0.656765	valid's binary_logloss: 0.660326
[300]	train's binary_logloss: 0.651459	valid's binary_logloss: 0.656068
[400]	train's binary_logloss: 0.648733	valid's binary_logloss: 0.65386
[500]	train's binary_logloss: 0.647145	valid's binary_logloss: 0.653231
[600]	train's binary_logloss: 0.645937	valid's binary_logloss: 0.652184
Early stopping, best iteration is:
[573]	train's binary_logloss: 0.646242	valid's binary_logloss: 0.652
min_data_in_leaf, val_score: 0.652000:  60%|######    | 3/5 [00:02<00:01,  1.25it/s][I 2020-09-27 04:41:15,775] Trial 62 finished with value: 0.6520002540523853 and parameters: {'min_child_samples': 10}. Best is trial 62 with value: 0.6520002540523853.
min_data_in_leaf, val_score: 0.652000:  60%|######    | 3/5 [00:02<00:01,  1.25it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000591 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667264	valid's binary_logloss: 0.669374
[200]	train's binary_logloss: 0.656765	valid's binary_logloss: 0.660326
[300]	train's binary_logloss: 0.651468	valid's binary_logloss: 0.656076
[400]	train's binary_logloss: 0.648844	valid's binary_logloss: 0.65385
[500]	train's binary_logloss: 0.647295	valid's binary_logloss: 0.653583
[600]	train's binary_logloss: 0.646236	valid's binary_logloss: 0.652591
Early stopping, best iteration is:
[572]	train's binary_logloss: 0.64651	valid's binary_logloss: 0.652484
min_data_in_leaf, val_score: 0.652000:  80%|########  | 4/5 [00:03<00:00,  1.07it/s][I 2020-09-27 04:41:17,010] Trial 63 finished with value: 0.6524841574858531 and parameters: {'min_child_samples': 100}. Best is trial 62 with value: 0.6520002540523853.
min_data_in_leaf, val_score: 0.652000:  80%|########  | 4/5 [00:03<00:00,  1.07it/s][LightGBM] [Info] Number of positive: 13149, number of negative: 12850
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000499 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505750 -> initscore=0.023002
[LightGBM] [Info] Start training from score 0.023002
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667264	valid's binary_logloss: 0.669374
[200]	train's binary_logloss: 0.656765	valid's binary_logloss: 0.660326
[300]	train's binary_logloss: 0.651459	valid's binary_logloss: 0.656068
[400]	train's binary_logloss: 0.648733	valid's binary_logloss: 0.65386
[500]	train's binary_logloss: 0.647145	valid's binary_logloss: 0.653231
[600]	train's binary_logloss: 0.645937	valid's binary_logloss: 0.652184
Early stopping, best iteration is:
[573]	train's binary_logloss: 0.646242	valid's binary_logloss: 0.652
min_data_in_leaf, val_score: 0.652000: 100%|##########| 5/5 [00:04<00:00,  1.13it/s][I 2020-09-27 04:41:17,796] Trial 64 finished with value: 0.6520002540523853 and parameters: {'min_child_samples': 5}. Best is trial 62 with value: 0.6520002540523853.
min_data_in_leaf, val_score: 0.652000: 100%|##########| 5/5 [00:04<00:00,  1.13it/s]
Fold : 3
[I 2020-09-27 04:41:17,860] A new study created in memory with name: no-name-2f38a2a7-25fd-480a-a0ca-a8d760f2fdde
feature_fraction, val_score: inf:   0%|          | 0/7 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004894 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.572639	valid's binary_logloss: 0.657549
Early stopping, best iteration is:
[69]	train's binary_logloss: 0.593294	valid's binary_logloss: 0.65643
feature_fraction, val_score: 0.656430:  14%|#4        | 1/7 [00:00<00:03,  1.89it/s][I 2020-09-27 04:41:18,402] Trial 0 finished with value: 0.6564300006090903 and parameters: {'feature_fraction': 0.8}. Best is trial 0 with value: 0.6564300006090903.
feature_fraction, val_score: 0.656430:  14%|#4        | 1/7 [00:00<00:03,  1.89it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000471 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.582961	valid's binary_logloss: 0.657331
[200]	train's binary_logloss: 0.532881	valid's binary_logloss: 0.663509
Early stopping, best iteration is:
[105]	train's binary_logloss: 0.579867	valid's binary_logloss: 0.657057
feature_fraction, val_score: 0.656430:  29%|##8       | 2/7 [00:01<00:02,  1.88it/s][I 2020-09-27 04:41:18,934] Trial 1 finished with value: 0.6570566397890953 and parameters: {'feature_fraction': 0.4}. Best is trial 0 with value: 0.6564300006090903.
feature_fraction, val_score: 0.656430:  29%|##8       | 2/7 [00:01<00:02,  1.88it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000466 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.578663	valid's binary_logloss: 0.658469
Early stopping, best iteration is:
[94]	train's binary_logloss: 0.582397	valid's binary_logloss: 0.657964
feature_fraction, val_score: 0.656430:  43%|####2     | 3/7 [00:01<00:02,  1.85it/s][I 2020-09-27 04:41:19,494] Trial 2 finished with value: 0.6579636929665533 and parameters: {'feature_fraction': 0.5}. Best is trial 0 with value: 0.6564300006090903.
feature_fraction, val_score: 0.656430:  43%|####2     | 3/7 [00:01<00:02,  1.85it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.008636 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.571407	valid's binary_logloss: 0.662124
Early stopping, best iteration is:
[73]	train's binary_logloss: 0.589006	valid's binary_logloss: 0.660046
feature_fraction, val_score: 0.656430:  57%|#####7    | 4/7 [00:02<00:02,  1.47it/s][I 2020-09-27 04:41:20,502] Trial 3 finished with value: 0.6600459491612519 and parameters: {'feature_fraction': 1.0}. Best is trial 0 with value: 0.6564300006090903.
feature_fraction, val_score: 0.656430:  57%|#####7    | 4/7 [00:02<00:02,  1.47it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.007591 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.576825	valid's binary_logloss: 0.659351
Early stopping, best iteration is:
[59]	train's binary_logloss: 0.60446	valid's binary_logloss: 0.656858
feature_fraction, val_score: 0.656430:  71%|#######1  | 5/7 [00:03<00:01,  1.57it/s][I 2020-09-27 04:41:21,037] Trial 4 finished with value: 0.6568575848518106 and parameters: {'feature_fraction': 0.6}. Best is trial 0 with value: 0.6564300006090903.
feature_fraction, val_score: 0.656430:  71%|#######1  | 5/7 [00:03<00:01,  1.57it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.005521 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.574018	valid's binary_logloss: 0.658815
Early stopping, best iteration is:
[58]	train's binary_logloss: 0.603435	valid's binary_logloss: 0.65692
feature_fraction, val_score: 0.656430:  86%|########5 | 6/7 [00:03<00:00,  1.70it/s][I 2020-09-27 04:41:21,513] Trial 5 finished with value: 0.6569204528754933 and parameters: {'feature_fraction': 0.7}. Best is trial 0 with value: 0.6564300006090903.
feature_fraction, val_score: 0.656430:  86%|########5 | 6/7 [00:03<00:00,  1.70it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000901 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.571583	valid's binary_logloss: 0.66348
Early stopping, best iteration is:
[58]	train's binary_logloss: 0.601331	valid's binary_logloss: 0.660277
feature_fraction, val_score: 0.656430: 100%|##########| 7/7 [00:04<00:00,  1.81it/s][I 2020-09-27 04:41:21,986] Trial 6 finished with value: 0.660277034185044 and parameters: {'feature_fraction': 0.8999999999999999}. Best is trial 0 with value: 0.6564300006090903.
feature_fraction, val_score: 0.656430: 100%|##########| 7/7 [00:04<00:00,  1.70it/s]
num_leaves, val_score: 0.656430:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000435 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.316367	valid's binary_logloss: 0.678402
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.477737	valid's binary_logloss: 0.661617
num_leaves, val_score: 0.656430:   5%|5         | 1/20 [00:01<00:22,  1.18s/it][I 2020-09-27 04:41:23,181] Trial 7 finished with value: 0.6616171717247548 and parameters: {'num_leaves': 199}. Best is trial 7 with value: 0.6616171717247548.
num_leaves, val_score: 0.656430:   5%|5         | 1/20 [00:01<00:22,  1.18s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000849 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.294294	valid's binary_logloss: 0.683498
Early stopping, best iteration is:
[30]	train's binary_logloss: 0.496866	valid's binary_logloss: 0.663982
num_leaves, val_score: 0.656430:  10%|#         | 2/20 [00:02<00:24,  1.34s/it][I 2020-09-27 04:41:24,901] Trial 8 finished with value: 0.6639821465726022 and parameters: {'num_leaves': 220}. Best is trial 7 with value: 0.6616171717247548.
num_leaves, val_score: 0.656430:  10%|#         | 2/20 [00:02<00:24,  1.34s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000432 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.317927	valid's binary_logloss: 0.67117
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.478124	valid's binary_logloss: 0.658535
num_leaves, val_score: 0.656430:  15%|#5        | 3/20 [00:04<00:21,  1.28s/it][I 2020-09-27 04:41:26,051] Trial 9 finished with value: 0.6585349270193589 and parameters: {'num_leaves': 197}. Best is trial 9 with value: 0.6585349270193589.
num_leaves, val_score: 0.656430:  15%|#5        | 3/20 [00:04<00:21,  1.28s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000872 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.6214	valid's binary_logloss: 0.657582
[200]	train's binary_logloss: 0.595975	valid's binary_logloss: 0.659137
Early stopping, best iteration is:
[112]	train's binary_logloss: 0.618201	valid's binary_logloss: 0.656843
num_leaves, val_score: 0.656430:  20%|##        | 4/20 [00:04<00:16,  1.04s/it][I 2020-09-27 04:41:26,504] Trial 10 finished with value: 0.6568431632806528 and parameters: {'num_leaves': 12}. Best is trial 10 with value: 0.6568431632806528.
num_leaves, val_score: 0.656430:  20%|##        | 4/20 [00:04<00:16,  1.04s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000825 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.633772	valid's binary_logloss: 0.658331
[200]	train's binary_logloss: 0.615857	valid's binary_logloss: 0.657932
Early stopping, best iteration is:
[151]	train's binary_logloss: 0.623812	valid's binary_logloss: 0.657048
num_leaves, val_score: 0.656430:  25%|##5       | 5/20 [00:04<00:12,  1.16it/s][I 2020-09-27 04:41:26,971] Trial 11 finished with value: 0.6570482164914353 and parameters: {'num_leaves': 8}. Best is trial 10 with value: 0.6568431632806528.
num_leaves, val_score: 0.656430:  25%|##5       | 5/20 [00:04<00:12,  1.16it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008860 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.6214	valid's binary_logloss: 0.657582
[200]	train's binary_logloss: 0.595975	valid's binary_logloss: 0.659137
Early stopping, best iteration is:
[112]	train's binary_logloss: 0.618201	valid's binary_logloss: 0.656843
num_leaves, val_score: 0.656430:  30%|###       | 6/20 [00:05<00:11,  1.18it/s][I 2020-09-27 04:41:27,770] Trial 12 finished with value: 0.6568431632806528 and parameters: {'num_leaves': 12}. Best is trial 10 with value: 0.6568431632806528.
num_leaves, val_score: 0.656430:  30%|###       | 6/20 [00:05<00:11,  1.18it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012578 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.458511	valid's binary_logloss: 0.667979
Early stopping, best iteration is:
[42]	train's binary_logloss: 0.55365	valid's binary_logloss: 0.659615
num_leaves, val_score: 0.656430:  35%|###5      | 7/20 [00:06<00:11,  1.14it/s][I 2020-09-27 04:41:28,720] Trial 13 finished with value: 0.6596145393331994 and parameters: {'num_leaves': 89}. Best is trial 10 with value: 0.6568431632806528.
num_leaves, val_score: 0.656430:  35%|###5      | 7/20 [00:06<00:11,  1.14it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.007578 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.458372	valid's binary_logloss: 0.664586
Early stopping, best iteration is:
[37]	train's binary_logloss: 0.562981	valid's binary_logloss: 0.659102
num_leaves, val_score: 0.656430:  40%|####      | 8/20 [00:07<00:09,  1.23it/s][I 2020-09-27 04:41:29,393] Trial 14 finished with value: 0.6591016497023785 and parameters: {'num_leaves': 90}. Best is trial 10 with value: 0.6568431632806528.
num_leaves, val_score: 0.656430:  40%|####      | 8/20 [00:07<00:09,  1.23it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001036 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.500134	valid's binary_logloss: 0.660323
Early stopping, best iteration is:
[55]	train's binary_logloss: 0.556151	valid's binary_logloss: 0.658021
num_leaves, val_score: 0.656430:  45%|####5     | 9/20 [00:08<00:08,  1.29it/s][I 2020-09-27 04:41:30,070] Trial 15 finished with value: 0.6580206440647851 and parameters: {'num_leaves': 66}. Best is trial 10 with value: 0.6568431632806528.
num_leaves, val_score: 0.656430:  45%|####5     | 9/20 [00:08<00:08,  1.29it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004995 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.37816	valid's binary_logloss: 0.67245
Early stopping, best iteration is:
[33]	train's binary_logloss: 0.531324	valid's binary_logloss: 0.658413
num_leaves, val_score: 0.656430:  50%|#####     | 10/20 [00:08<00:07,  1.27it/s][I 2020-09-27 04:41:30,896] Trial 16 finished with value: 0.6584130655795914 and parameters: {'num_leaves': 144}. Best is trial 10 with value: 0.6568431632806528.
num_leaves, val_score: 0.656430:  50%|#####     | 10/20 [00:08<00:07,  1.27it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000912 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.557347	valid's binary_logloss: 0.658946
Early stopping, best iteration is:
[52]	train's binary_logloss: 0.597034	valid's binary_logloss: 0.657578
num_leaves, val_score: 0.656430:  55%|#####5    | 11/20 [00:09<00:07,  1.15it/s][I 2020-09-27 04:41:31,958] Trial 17 finished with value: 0.6575775290564063 and parameters: {'num_leaves': 38}. Best is trial 10 with value: 0.6568431632806528.
num_leaves, val_score: 0.656430:  55%|#####5    | 11/20 [00:09<00:07,  1.15it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004745 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.380846	valid's binary_logloss: 0.673729
Early stopping, best iteration is:
[29]	train's binary_logloss: 0.545732	valid's binary_logloss: 0.661242
num_leaves, val_score: 0.656430:  60%|######    | 12/20 [00:10<00:06,  1.17it/s][I 2020-09-27 04:41:32,785] Trial 18 finished with value: 0.6612424428440397 and parameters: {'num_leaves': 143}. Best is trial 10 with value: 0.6568431632806528.
num_leaves, val_score: 0.656430:  60%|######    | 12/20 [00:10<00:06,  1.17it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000446 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.270716	valid's binary_logloss: 0.684739
Early stopping, best iteration is:
[22]	train's binary_logloss: 0.522558	valid's binary_logloss: 0.66216
num_leaves, val_score: 0.656430:  65%|######5   | 13/20 [00:12<00:06,  1.03it/s][I 2020-09-27 04:41:34,008] Trial 19 finished with value: 0.6621600912726282 and parameters: {'num_leaves': 248}. Best is trial 10 with value: 0.6568431632806528.
num_leaves, val_score: 0.656430:  65%|######5   | 13/20 [00:12<00:06,  1.03it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000999 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.53747	valid's binary_logloss: 0.659534
Early stopping, best iteration is:
[53]	train's binary_logloss: 0.583686	valid's binary_logloss: 0.65689
num_leaves, val_score: 0.656430:  70%|#######   | 14/20 [00:12<00:05,  1.18it/s][I 2020-09-27 04:41:34,577] Trial 20 finished with value: 0.6568899584326288 and parameters: {'num_leaves': 47}. Best is trial 10 with value: 0.6568431632806528.
num_leaves, val_score: 0.656430:  70%|#######   | 14/20 [00:12<00:05,  1.18it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000858 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.668556	valid's binary_logloss: 0.673758
[200]	train's binary_logloss: 0.658927	valid's binary_logloss: 0.665036
[300]	train's binary_logloss: 0.653688	valid's binary_logloss: 0.660827
[400]	train's binary_logloss: 0.650562	valid's binary_logloss: 0.658072
[500]	train's binary_logloss: 0.648564	valid's binary_logloss: 0.656465
[600]	train's binary_logloss: 0.647233	valid's binary_logloss: 0.655679
[700]	train's binary_logloss: 0.646297	valid's binary_logloss: 0.655125
[800]	train's binary_logloss: 0.64559	valid's binary_logloss: 0.654927
[900]	train's binary_logloss: 0.645037	valid's binary_logloss: 0.654748
[1000]	train's binary_logloss: 0.644582	valid's binary_logloss: 0.654744
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.644582	valid's binary_logloss: 0.654744
num_leaves, val_score: 0.654744:  75%|#######5  | 15/20 [00:14<00:05,  1.10s/it][I 2020-09-27 04:41:36,250] Trial 21 finished with value: 0.6547443275104798 and parameters: {'num_leaves': 2}. Best is trial 21 with value: 0.6547443275104798.
num_leaves, val_score: 0.654744:  75%|#######5  | 15/20 [00:14<00:05,  1.10s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005030 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607455	valid's binary_logloss: 0.658232
Early stopping, best iteration is:
[89]	train's binary_logloss: 0.611867	valid's binary_logloss: 0.657658
num_leaves, val_score: 0.654744:  80%|########  | 16/20 [00:14<00:03,  1.11it/s][I 2020-09-27 04:41:36,691] Trial 22 finished with value: 0.6576583665143456 and parameters: {'num_leaves': 17}. Best is trial 21 with value: 0.6547443275104798.
num_leaves, val_score: 0.654744:  80%|########  | 16/20 [00:14<00:03,  1.11it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000984 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657455	valid's binary_logloss: 0.665049
[200]	train's binary_logloss: 0.647037	valid's binary_logloss: 0.658049
[300]	train's binary_logloss: 0.641503	valid's binary_logloss: 0.655542
[400]	train's binary_logloss: 0.6376	valid's binary_logloss: 0.6548
[500]	train's binary_logloss: 0.634488	valid's binary_logloss: 0.654825
[600]	train's binary_logloss: 0.631676	valid's binary_logloss: 0.654591
Early stopping, best iteration is:
[550]	train's binary_logloss: 0.633043	valid's binary_logloss: 0.654226
num_leaves, val_score: 0.654226:  85%|########5 | 17/20 [00:15<00:02,  1.11it/s][I 2020-09-27 04:41:37,600] Trial 23 finished with value: 0.6542264931984162 and parameters: {'num_leaves': 3}. Best is trial 23 with value: 0.6542264931984162.
num_leaves, val_score: 0.654226:  85%|########5 | 17/20 [00:15<00:02,  1.11it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000878 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.44017	valid's binary_logloss: 0.669582
Early stopping, best iteration is:
[27]	train's binary_logloss: 0.579264	valid's binary_logloss: 0.660699
num_leaves, val_score: 0.654226:  90%|######### | 18/20 [00:16<00:01,  1.17it/s][I 2020-09-27 04:41:38,351] Trial 24 finished with value: 0.6606988142412215 and parameters: {'num_leaves': 101}. Best is trial 23 with value: 0.6542264931984162.
num_leaves, val_score: 0.654226:  90%|######### | 18/20 [00:16<00:01,  1.17it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000476 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.534907	valid's binary_logloss: 0.660247
Early stopping, best iteration is:
[52]	train's binary_logloss: 0.583748	valid's binary_logloss: 0.656588
num_leaves, val_score: 0.654226:  95%|#########5| 19/20 [00:17<00:00,  1.05it/s][I 2020-09-27 04:41:39,536] Trial 25 finished with value: 0.6565875567036938 and parameters: {'num_leaves': 48}. Best is trial 23 with value: 0.6542264931984162.
num_leaves, val_score: 0.654226:  95%|#########5| 19/20 [00:17<00:00,  1.05it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000989 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.412133	valid's binary_logloss: 0.670637
Early stopping, best iteration is:
[35]	train's binary_logloss: 0.542676	valid's binary_logloss: 0.662727
num_leaves, val_score: 0.654226: 100%|##########| 20/20 [00:18<00:00,  1.09it/s][I 2020-09-27 04:41:40,357] Trial 26 finished with value: 0.6627267041691768 and parameters: {'num_leaves': 120}. Best is trial 23 with value: 0.6542264931984162.
num_leaves, val_score: 0.654226: 100%|##########| 20/20 [00:18<00:00,  1.09it/s]
bagging, val_score: 0.654226:   0%|          | 0/10 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004771 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656611	valid's binary_logloss: 0.664545
[200]	train's binary_logloss: 0.64587	valid's binary_logloss: 0.657646
[300]	train's binary_logloss: 0.640645	valid's binary_logloss: 0.655721
[400]	train's binary_logloss: 0.636853	valid's binary_logloss: 0.655725
[500]	train's binary_logloss: 0.633496	valid's binary_logloss: 0.655542
[600]	train's binary_logloss: 0.630016	valid's binary_logloss: 0.655279
Early stopping, best iteration is:
[583]	train's binary_logloss: 0.6307	valid's binary_logloss: 0.655229
bagging, val_score: 0.654226:  10%|#         | 1/10 [00:00<00:08,  1.05it/s][I 2020-09-27 04:41:41,322] Trial 27 finished with value: 0.6552294862634005 and parameters: {'bagging_fraction': 0.7419038111841914, 'bagging_freq': 4}. Best is trial 27 with value: 0.6552294862634005.
bagging, val_score: 0.654226:  10%|#         | 1/10 [00:00<00:08,  1.05it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.011398 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656644	valid's binary_logloss: 0.664211
[200]	train's binary_logloss: 0.646379	valid's binary_logloss: 0.65746
[300]	train's binary_logloss: 0.64107	valid's binary_logloss: 0.655347
[400]	train's binary_logloss: 0.63725	valid's binary_logloss: 0.654521
[500]	train's binary_logloss: 0.633892	valid's binary_logloss: 0.654747
Early stopping, best iteration is:
[418]	train's binary_logloss: 0.636564	valid's binary_logloss: 0.654327
bagging, val_score: 0.654226:  20%|##        | 2/10 [00:01<00:07,  1.10it/s][I 2020-09-27 04:41:42,144] Trial 28 finished with value: 0.6543274423941562 and parameters: {'bagging_fraction': 0.7471333362321411, 'bagging_freq': 4}. Best is trial 28 with value: 0.6543274423941562.
bagging, val_score: 0.654226:  20%|##        | 2/10 [00:01<00:07,  1.10it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000868 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657402	valid's binary_logloss: 0.66564
[200]	train's binary_logloss: 0.646902	valid's binary_logloss: 0.659056
[300]	train's binary_logloss: 0.641447	valid's binary_logloss: 0.656325
[400]	train's binary_logloss: 0.637573	valid's binary_logloss: 0.655624
[500]	train's binary_logloss: 0.634342	valid's binary_logloss: 0.655314
Early stopping, best iteration is:
[451]	train's binary_logloss: 0.635853	valid's binary_logloss: 0.655055
bagging, val_score: 0.654226:  30%|###       | 3/10 [00:03<00:07,  1.05s/it][I 2020-09-27 04:41:43,507] Trial 29 finished with value: 0.6550545130006391 and parameters: {'bagging_fraction': 0.9853823451672025, 'bagging_freq': 7}. Best is trial 28 with value: 0.6543274423941562.
bagging, val_score: 0.654226:  30%|###       | 3/10 [00:03<00:07,  1.05s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005058 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.655834	valid's binary_logloss: 0.664357
[200]	train's binary_logloss: 0.645749	valid's binary_logloss: 0.656264
[300]	train's binary_logloss: 0.641025	valid's binary_logloss: 0.656599
Early stopping, best iteration is:
[244]	train's binary_logloss: 0.643386	valid's binary_logloss: 0.656058
bagging, val_score: 0.654226:  40%|####      | 4/10 [00:03<00:05,  1.08it/s][I 2020-09-27 04:41:44,148] Trial 30 finished with value: 0.6560583873045698 and parameters: {'bagging_fraction': 0.5579781104550179, 'bagging_freq': 3}. Best is trial 28 with value: 0.6543274423941562.
bagging, val_score: 0.654226:  40%|####      | 4/10 [00:03<00:05,  1.08it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000466 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657297	valid's binary_logloss: 0.665521
[200]	train's binary_logloss: 0.64689	valid's binary_logloss: 0.658315
[300]	train's binary_logloss: 0.641484	valid's binary_logloss: 0.655464
[400]	train's binary_logloss: 0.637638	valid's binary_logloss: 0.654412
[500]	train's binary_logloss: 0.634281	valid's binary_logloss: 0.653965
[600]	train's binary_logloss: 0.631536	valid's binary_logloss: 0.653942
Early stopping, best iteration is:
[511]	train's binary_logloss: 0.633982	valid's binary_logloss: 0.65385
bagging, val_score: 0.653850:  50%|#####     | 5/10 [00:04<00:04,  1.06it/s][I 2020-09-27 04:41:45,139] Trial 31 finished with value: 0.6538504332512516 and parameters: {'bagging_fraction': 0.9521342276139986, 'bagging_freq': 7}. Best is trial 31 with value: 0.6538504332512516.
bagging, val_score: 0.653850:  50%|#####     | 5/10 [00:04<00:04,  1.06it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005270 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657127	valid's binary_logloss: 0.665533
[200]	train's binary_logloss: 0.646571	valid's binary_logloss: 0.65777
[300]	train's binary_logloss: 0.641217	valid's binary_logloss: 0.655631
[400]	train's binary_logloss: 0.637412	valid's binary_logloss: 0.654782
[500]	train's binary_logloss: 0.63418	valid's binary_logloss: 0.654189
Early stopping, best iteration is:
[483]	train's binary_logloss: 0.634685	valid's binary_logloss: 0.654112
bagging, val_score: 0.653850:  60%|######    | 6/10 [00:05<00:03,  1.07it/s][I 2020-09-27 04:41:46,049] Trial 32 finished with value: 0.6541122278782455 and parameters: {'bagging_fraction': 0.9284471640035712, 'bagging_freq': 7}. Best is trial 31 with value: 0.6538504332512516.
bagging, val_score: 0.653850:  60%|######    | 6/10 [00:05<00:03,  1.07it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000855 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657282	valid's binary_logloss: 0.665245
[200]	train's binary_logloss: 0.646764	valid's binary_logloss: 0.657757
[300]	train's binary_logloss: 0.641399	valid's binary_logloss: 0.655443
[400]	train's binary_logloss: 0.637639	valid's binary_logloss: 0.654744
[500]	train's binary_logloss: 0.634399	valid's binary_logloss: 0.654233
[600]	train's binary_logloss: 0.63167	valid's binary_logloss: 0.654358
Early stopping, best iteration is:
[506]	train's binary_logloss: 0.634248	valid's binary_logloss: 0.654145
bagging, val_score: 0.653850:  70%|#######   | 7/10 [00:07<00:03,  1.05s/it][I 2020-09-27 04:41:47,378] Trial 33 finished with value: 0.6541452383331995 and parameters: {'bagging_fraction': 0.9500310046900615, 'bagging_freq': 7}. Best is trial 31 with value: 0.6538504332512516.
bagging, val_score: 0.653850:  70%|#######   | 7/10 [00:07<00:03,  1.05s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004962 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657608	valid's binary_logloss: 0.665788
[200]	train's binary_logloss: 0.647027	valid's binary_logloss: 0.6581
[300]	train's binary_logloss: 0.641573	valid's binary_logloss: 0.655658
[400]	train's binary_logloss: 0.637741	valid's binary_logloss: 0.655284
[500]	train's binary_logloss: 0.634569	valid's binary_logloss: 0.654728
[600]	train's binary_logloss: 0.631817	valid's binary_logloss: 0.654394
[700]	train's binary_logloss: 0.629035	valid's binary_logloss: 0.654408
Early stopping, best iteration is:
[653]	train's binary_logloss: 0.630298	valid's binary_logloss: 0.654176
bagging, val_score: 0.653850:  80%|########  | 8/10 [00:08<00:02,  1.11s/it][I 2020-09-27 04:41:48,627] Trial 34 finished with value: 0.654175795227107 and parameters: {'bagging_fraction': 0.9963869875397509, 'bagging_freq': 7}. Best is trial 31 with value: 0.6538504332512516.
bagging, val_score: 0.653850:  80%|########  | 8/10 [00:08<00:02,  1.11s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005318 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657371	valid's binary_logloss: 0.665888
[200]	train's binary_logloss: 0.646844	valid's binary_logloss: 0.658962
[300]	train's binary_logloss: 0.64144	valid's binary_logloss: 0.656351
[400]	train's binary_logloss: 0.637614	valid's binary_logloss: 0.654985
[500]	train's binary_logloss: 0.634424	valid's binary_logloss: 0.654861
[600]	train's binary_logloss: 0.631665	valid's binary_logloss: 0.654356
[700]	train's binary_logloss: 0.628979	valid's binary_logloss: 0.654652
Early stopping, best iteration is:
[625]	train's binary_logloss: 0.631017	valid's binary_logloss: 0.654177
bagging, val_score: 0.653850:  90%|######### | 9/10 [00:09<00:01,  1.19s/it][I 2020-09-27 04:41:50,003] Trial 35 finished with value: 0.6541765913434219 and parameters: {'bagging_fraction': 0.9914234683174175, 'bagging_freq': 7}. Best is trial 31 with value: 0.6538504332512516.
bagging, val_score: 0.653850:  90%|######### | 9/10 [00:09<00:01,  1.19s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.006921 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657129	valid's binary_logloss: 0.666433
[200]	train's binary_logloss: 0.646683	valid's binary_logloss: 0.658903
[300]	train's binary_logloss: 0.641282	valid's binary_logloss: 0.656802
[400]	train's binary_logloss: 0.637336	valid's binary_logloss: 0.655707
[500]	train's binary_logloss: 0.634119	valid's binary_logloss: 0.654705
[600]	train's binary_logloss: 0.631418	valid's binary_logloss: 0.654631
Early stopping, best iteration is:
[510]	train's binary_logloss: 0.633847	valid's binary_logloss: 0.654396
bagging, val_score: 0.653850: 100%|##########| 10/10 [00:11<00:00,  1.28s/it][I 2020-09-27 04:41:51,487] Trial 36 finished with value: 0.6543958028999971 and parameters: {'bagging_fraction': 0.9429983760903378, 'bagging_freq': 7}. Best is trial 31 with value: 0.6538504332512516.
bagging, val_score: 0.653850: 100%|##########| 10/10 [00:11<00:00,  1.11s/it]
feature_fraction_stage2, val_score: 0.653850:   0%|          | 0/6 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000804 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657244	valid's binary_logloss: 0.665676
[200]	train's binary_logloss: 0.646834	valid's binary_logloss: 0.657933
[300]	train's binary_logloss: 0.641472	valid's binary_logloss: 0.655835
[400]	train's binary_logloss: 0.637734	valid's binary_logloss: 0.654847
[500]	train's binary_logloss: 0.634612	valid's binary_logloss: 0.654743
[600]	train's binary_logloss: 0.631834	valid's binary_logloss: 0.654606
Early stopping, best iteration is:
[559]	train's binary_logloss: 0.632955	valid's binary_logloss: 0.654387
feature_fraction_stage2, val_score: 0.653850:  17%|#6        | 1/6 [00:01<00:05,  1.06s/it][I 2020-09-27 04:41:52,568] Trial 37 finished with value: 0.6543873568694546 and parameters: {'feature_fraction': 0.7520000000000001}. Best is trial 37 with value: 0.6543873568694546.
feature_fraction_stage2, val_score: 0.653850:  17%|#6        | 1/6 [00:01<00:05,  1.06s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001058 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637619	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.63474	valid's binary_logloss: 0.653747
feature_fraction_stage2, val_score: 0.653747:  33%|###3      | 2/6 [00:02<00:04,  1.04s/it][I 2020-09-27 04:41:53,565] Trial 38 finished with value: 0.6537470788868942 and parameters: {'feature_fraction': 0.7200000000000001}. Best is trial 38 with value: 0.6537470788868942.
feature_fraction_stage2, val_score: 0.653747:  33%|###3      | 2/6 [00:02<00:04,  1.04s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001682 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657318	valid's binary_logloss: 0.665806
[200]	train's binary_logloss: 0.646625	valid's binary_logloss: 0.658644
[300]	train's binary_logloss: 0.641194	valid's binary_logloss: 0.656049
[400]	train's binary_logloss: 0.63737	valid's binary_logloss: 0.655556
[500]	train's binary_logloss: 0.634009	valid's binary_logloss: 0.654948
[600]	train's binary_logloss: 0.63129	valid's binary_logloss: 0.65525
Early stopping, best iteration is:
[557]	train's binary_logloss: 0.63246	valid's binary_logloss: 0.654813
feature_fraction_stage2, val_score: 0.653747:  50%|#####     | 3/6 [00:03<00:03,  1.21s/it][I 2020-09-27 04:41:55,150] Trial 39 finished with value: 0.6548128899258291 and parameters: {'feature_fraction': 0.8480000000000001}. Best is trial 38 with value: 0.6537470788868942.
feature_fraction_stage2, val_score: 0.653747:  50%|#####     | 3/6 [00:03<00:03,  1.21s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001085 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.665659
[200]	train's binary_logloss: 0.646755	valid's binary_logloss: 0.657975
[300]	train's binary_logloss: 0.641365	valid's binary_logloss: 0.655752
[400]	train's binary_logloss: 0.637538	valid's binary_logloss: 0.654478
[500]	train's binary_logloss: 0.634276	valid's binary_logloss: 0.654404
[600]	train's binary_logloss: 0.63152	valid's binary_logloss: 0.654509
Early stopping, best iteration is:
[533]	train's binary_logloss: 0.633346	valid's binary_logloss: 0.654136
feature_fraction_stage2, val_score: 0.653747:  67%|######6   | 4/6 [00:04<00:02,  1.22s/it][I 2020-09-27 04:41:56,409] Trial 40 finished with value: 0.6541357158243993 and parameters: {'feature_fraction': 0.88}. Best is trial 38 with value: 0.6537470788868942.
feature_fraction_stage2, val_score: 0.653747:  67%|######6   | 4/6 [00:04<00:02,  1.22s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005676 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657297	valid's binary_logloss: 0.665521
[200]	train's binary_logloss: 0.64689	valid's binary_logloss: 0.658315
[300]	train's binary_logloss: 0.641484	valid's binary_logloss: 0.655464
[400]	train's binary_logloss: 0.637638	valid's binary_logloss: 0.654412
[500]	train's binary_logloss: 0.634281	valid's binary_logloss: 0.653965
[600]	train's binary_logloss: 0.631536	valid's binary_logloss: 0.653942
Early stopping, best iteration is:
[511]	train's binary_logloss: 0.633982	valid's binary_logloss: 0.65385
feature_fraction_stage2, val_score: 0.653747:  83%|########3 | 5/6 [00:05<00:01,  1.18s/it][I 2020-09-27 04:41:57,501] Trial 41 finished with value: 0.6538504332512516 and parameters: {'feature_fraction': 0.8160000000000001}. Best is trial 38 with value: 0.6537470788868942.
feature_fraction_stage2, val_score: 0.653747:  83%|########3 | 5/6 [00:06<00:01,  1.18s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010335 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657244	valid's binary_logloss: 0.665676
[200]	train's binary_logloss: 0.646834	valid's binary_logloss: 0.657933
[300]	train's binary_logloss: 0.641472	valid's binary_logloss: 0.655835
[400]	train's binary_logloss: 0.637734	valid's binary_logloss: 0.654847
[500]	train's binary_logloss: 0.634612	valid's binary_logloss: 0.654743
[600]	train's binary_logloss: 0.631834	valid's binary_logloss: 0.654606
Early stopping, best iteration is:
[559]	train's binary_logloss: 0.632955	valid's binary_logloss: 0.654387
feature_fraction_stage2, val_score: 0.653747: 100%|##########| 6/6 [00:07<00:00,  1.29s/it][I 2020-09-27 04:41:59,040] Trial 42 finished with value: 0.6543873568694546 and parameters: {'feature_fraction': 0.784}. Best is trial 38 with value: 0.6537470788868942.
feature_fraction_stage2, val_score: 0.653747: 100%|##########| 6/6 [00:07<00:00,  1.26s/it]
regularization_factors, val_score: 0.653747:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000984 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637619	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.63474	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653747:   5%|5         | 1/20 [00:00<00:17,  1.11it/s][I 2020-09-27 04:41:59,955] Trial 43 finished with value: 0.6537470788623818 and parameters: {'lambda_l1': 1.164724543974479e-08, 'lambda_l2': 6.922005451695355e-06}. Best is trial 43 with value: 0.6537470788623818.
regularization_factors, val_score: 0.653747:   5%|5         | 1/20 [00:00<00:17,  1.11it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000880 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637619	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.63474	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653747:  10%|#         | 2/20 [00:01<00:16,  1.11it/s][I 2020-09-27 04:42:00,859] Trial 44 finished with value: 0.6537470788688746 and parameters: {'lambda_l1': 1.2101619670742207e-08, 'lambda_l2': 3.569458567905685e-06}. Best is trial 43 with value: 0.6537470788623818.
regularization_factors, val_score: 0.653747:  10%|#         | 2/20 [00:01<00:16,  1.11it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000804 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637619	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.63474	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653747:  15%|#5        | 3/20 [00:02<00:15,  1.08it/s][I 2020-09-27 04:42:01,825] Trial 45 finished with value: 0.653747078870228 and parameters: {'lambda_l1': 1.0924605897873326e-08, 'lambda_l2': 3.7438094337593657e-06}. Best is trial 43 with value: 0.6537470788623818.
regularization_factors, val_score: 0.653747:  15%|#5        | 3/20 [00:02<00:15,  1.08it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.009826 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637619	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.63474	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653747:  20%|##        | 4/20 [00:04<00:20,  1.28s/it][I 2020-09-27 04:42:03,949] Trial 46 finished with value: 0.6537470788718757 and parameters: {'lambda_l1': 1.479258071574677e-08, 'lambda_l2': 3.0264187607450267e-06}. Best is trial 43 with value: 0.6537470788623818.
regularization_factors, val_score: 0.653747:  20%|##        | 4/20 [00:04<00:20,  1.28s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004927 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637619	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.63474	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653747:  25%|##5       | 5/20 [00:05<00:18,  1.20s/it][I 2020-09-27 04:42:04,966] Trial 47 finished with value: 0.6537470788672778 and parameters: {'lambda_l1': 1.2446439098514144e-08, 'lambda_l2': 4.099446197874459e-06}. Best is trial 43 with value: 0.6537470788623818.
regularization_factors, val_score: 0.653747:  25%|##5       | 5/20 [00:05<00:18,  1.20s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004710 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637619	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.63474	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653747:  30%|###       | 6/20 [00:06<00:15,  1.11s/it][I 2020-09-27 04:42:05,863] Trial 48 finished with value: 0.6537470788771779 and parameters: {'lambda_l1': 1.2455751430953037e-08, 'lambda_l2': 2.1195700656837565e-06}. Best is trial 43 with value: 0.6537470788623818.
regularization_factors, val_score: 0.653747:  30%|###       | 6/20 [00:06<00:15,  1.11s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000751 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637619	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.63474	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653747:  35%|###5      | 7/20 [00:08<00:15,  1.18s/it][I 2020-09-27 04:42:07,198] Trial 49 finished with value: 0.6537470788676445 and parameters: {'lambda_l1': 1.3371676440125825e-08, 'lambda_l2': 4.133315482813227e-06}. Best is trial 43 with value: 0.6537470788623818.
regularization_factors, val_score: 0.653747:  35%|###5      | 7/20 [00:08<00:15,  1.18s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000868 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637619	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.63474	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653747:  40%|####      | 8/20 [00:09<00:13,  1.14s/it][I 2020-09-27 04:42:08,255] Trial 50 finished with value: 0.653747078859269 and parameters: {'lambda_l1': 1.1210362012517112e-08, 'lambda_l2': 6.439200774621681e-06}. Best is trial 50 with value: 0.653747078859269.
regularization_factors, val_score: 0.653747:  40%|####      | 8/20 [00:09<00:13,  1.14s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000976 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637619	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.63474	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653747:  45%|####5     | 9/20 [00:10<00:11,  1.07s/it][I 2020-09-27 04:42:09,153] Trial 51 finished with value: 0.6537470788593895 and parameters: {'lambda_l1': 1.3454643930345611e-08, 'lambda_l2': 6.467158854009067e-06}. Best is trial 50 with value: 0.653747078859269.
regularization_factors, val_score: 0.653747:  45%|####5     | 9/20 [00:10<00:11,  1.07s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.006696 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637619	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.63474	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653747:  50%|#####     | 10/20 [00:11<00:10,  1.02s/it][I 2020-09-27 04:42:10,075] Trial 52 finished with value: 0.6537470788591189 and parameters: {'lambda_l1': 1.1646094395679663e-08, 'lambda_l2': 6.358551933697898e-06}. Best is trial 52 with value: 0.6537470788591189.
regularization_factors, val_score: 0.653747:  50%|#####     | 10/20 [00:11<00:10,  1.02s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.011178 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637619	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.63474	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653747:  55%|#####5    | 11/20 [00:12<00:09,  1.11s/it][I 2020-09-27 04:42:11,387] Trial 53 finished with value: 0.6537470788474831 and parameters: {'lambda_l1': 1.0924146750147604e-08, 'lambda_l2': 1.2976298424529484e-05}. Best is trial 53 with value: 0.6537470788474831.
regularization_factors, val_score: 0.653747:  55%|#####5    | 11/20 [00:12<00:09,  1.11s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005401 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637619	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.63474	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653747:  60%|######    | 12/20 [00:13<00:08,  1.05s/it][I 2020-09-27 04:42:12,300] Trial 54 finished with value: 0.65374707865709 and parameters: {'lambda_l1': 1.1771590372611887e-08, 'lambda_l2': 6.880400470984478e-05}. Best is trial 54 with value: 0.65374707865709.
regularization_factors, val_score: 0.653747:  60%|######    | 12/20 [00:13<00:08,  1.05s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000845 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657307	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.63762	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.63434	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.634741	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653747:  65%|######5   | 13/20 [00:14<00:07,  1.02s/it][I 2020-09-27 04:42:13,235] Trial 55 finished with value: 0.6537470737830029 and parameters: {'lambda_l1': 1.2904541521956938e-08, 'lambda_l2': 0.0015325466166573717}. Best is trial 55 with value: 0.6537470737830029.
regularization_factors, val_score: 0.653747:  65%|######5   | 13/20 [00:14<00:07,  1.02s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004533 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657308	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.64145	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637678	valid's binary_logloss: 0.654439
[500]	train's binary_logloss: 0.634307	valid's binary_logloss: 0.653862
[600]	train's binary_logloss: 0.631527	valid's binary_logloss: 0.653578
[700]	train's binary_logloss: 0.628591	valid's binary_logloss: 0.653558
[800]	train's binary_logloss: 0.625872	valid's binary_logloss: 0.653736
Early stopping, best iteration is:
[750]	train's binary_logloss: 0.627225	valid's binary_logloss: 0.653307
regularization_factors, val_score: 0.653307:  70%|#######   | 14/20 [00:15<00:07,  1.22s/it][I 2020-09-27 04:42:14,917] Trial 56 finished with value: 0.653306938112237 and parameters: {'lambda_l1': 1.1130628894602006e-05, 'lambda_l2': 0.007158975373652711}. Best is trial 56 with value: 0.653306938112237.
regularization_factors, val_score: 0.653307:  70%|#######   | 14/20 [00:15<00:07,  1.22s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.019645 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657308	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646792	valid's binary_logloss: 0.657796
[300]	train's binary_logloss: 0.641385	valid's binary_logloss: 0.655452
[400]	train's binary_logloss: 0.637507	valid's binary_logloss: 0.654514
[500]	train's binary_logloss: 0.634295	valid's binary_logloss: 0.653752
Early stopping, best iteration is:
[490]	train's binary_logloss: 0.63465	valid's binary_logloss: 0.653564
regularization_factors, val_score: 0.653307:  75%|#######5  | 15/20 [00:16<00:05,  1.15s/it][I 2020-09-27 04:42:15,922] Trial 57 finished with value: 0.6535639999340161 and parameters: {'lambda_l1': 0.00026873318136680526, 'lambda_l2': 0.011949459080105706}. Best is trial 56 with value: 0.653306938112237.
regularization_factors, val_score: 0.653307:  75%|#######5  | 15/20 [00:16<00:05,  1.15s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000821 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657308	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646792	valid's binary_logloss: 0.657796
[300]	train's binary_logloss: 0.641358	valid's binary_logloss: 0.655395
[400]	train's binary_logloss: 0.637525	valid's binary_logloss: 0.654849
[500]	train's binary_logloss: 0.63422	valid's binary_logloss: 0.654174
[600]	train's binary_logloss: 0.631488	valid's binary_logloss: 0.654245
Early stopping, best iteration is:
[588]	train's binary_logloss: 0.631831	valid's binary_logloss: 0.654014
regularization_factors, val_score: 0.653307:  80%|########  | 16/20 [00:17<00:04,  1.12s/it][I 2020-09-27 04:42:16,965] Trial 58 finished with value: 0.6540142566432265 and parameters: {'lambda_l1': 0.0002391919617921031, 'lambda_l2': 0.020184808500303845}. Best is trial 56 with value: 0.653306938112237.
regularization_factors, val_score: 0.653307:  80%|########  | 16/20 [00:17<00:04,  1.12s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004897 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657308	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.63762	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.634341	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.634741	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653307:  85%|########5 | 17/20 [00:18<00:03,  1.06s/it][I 2020-09-27 04:42:17,888] Trial 59 finished with value: 0.6537470700910835 and parameters: {'lambda_l1': 0.00026680911329737025, 'lambda_l2': 0.0033110276633739268}. Best is trial 56 with value: 0.653306938112237.
regularization_factors, val_score: 0.653307:  85%|########5 | 17/20 [00:18<00:03,  1.06s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004895 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657308	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637621	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.634341	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.634742	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653307:  90%|######### | 18/20 [00:20<00:02,  1.17s/it][I 2020-09-27 04:42:19,323] Trial 60 finished with value: 0.6537470677716328 and parameters: {'lambda_l1': 0.0003014398725480866, 'lambda_l2': 0.004090143660649428}. Best is trial 56 with value: 0.653306938112237.
regularization_factors, val_score: 0.653307:  90%|######### | 18/20 [00:20<00:02,  1.17s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000797 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657308	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637621	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.634342	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.634742	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653307:  95%|#########5| 19/20 [00:21<00:01,  1.10s/it][I 2020-09-27 04:42:20,258] Trial 61 finished with value: 0.6537470687771331 and parameters: {'lambda_l1': 0.0005632008256227812, 'lambda_l2': 0.004446124070486416}. Best is trial 56 with value: 0.653306938112237.
regularization_factors, val_score: 0.653307:  95%|#########5| 19/20 [00:21<00:01,  1.10s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000827 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657308	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641449	valid's binary_logloss: 0.65506
[400]	train's binary_logloss: 0.637621	valid's binary_logloss: 0.654377
[500]	train's binary_logloss: 0.634341	valid's binary_logloss: 0.653884
Early stopping, best iteration is:
[488]	train's binary_logloss: 0.634742	valid's binary_logloss: 0.653747
regularization_factors, val_score: 0.653307: 100%|##########| 20/20 [00:22<00:00,  1.04s/it][I 2020-09-27 04:42:21,158] Trial 62 finished with value: 0.6537470665554331 and parameters: {'lambda_l1': 0.0003132915633232416, 'lambda_l2': 0.004489563552243147}. Best is trial 56 with value: 0.653306938112237.
regularization_factors, val_score: 0.653307: 100%|##########| 20/20 [00:22<00:00,  1.11s/it]
min_data_in_leaf, val_score: 0.653307:   0%|          | 0/5 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004952 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657308	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641484	valid's binary_logloss: 0.655571
[400]	train's binary_logloss: 0.63758	valid's binary_logloss: 0.654925
Early stopping, best iteration is:
[368]	train's binary_logloss: 0.638713	valid's binary_logloss: 0.654841
min_data_in_leaf, val_score: 0.653307:  20%|##        | 1/5 [00:00<00:03,  1.21it/s][I 2020-09-27 04:42:21,996] Trial 63 finished with value: 0.6548408998104529 and parameters: {'min_child_samples': 5}. Best is trial 63 with value: 0.6548408998104529.
min_data_in_leaf, val_score: 0.653307:  20%|##        | 1/5 [00:00<00:03,  1.21it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005026 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657308	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646871	valid's binary_logloss: 0.657816
[300]	train's binary_logloss: 0.641553	valid's binary_logloss: 0.655747
[400]	train's binary_logloss: 0.637618	valid's binary_logloss: 0.655166
[500]	train's binary_logloss: 0.63433	valid's binary_logloss: 0.654821
Early stopping, best iteration is:
[483]	train's binary_logloss: 0.634862	valid's binary_logloss: 0.654649
min_data_in_leaf, val_score: 0.653307:  40%|####      | 2/5 [00:02<00:02,  1.04it/s][I 2020-09-27 04:42:23,270] Trial 64 finished with value: 0.6546487352338517 and parameters: {'min_child_samples': 25}. Best is trial 64 with value: 0.6546487352338517.
min_data_in_leaf, val_score: 0.653307:  40%|####      | 2/5 [00:02<00:02,  1.04it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004979 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657308	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646783	valid's binary_logloss: 0.657725
[300]	train's binary_logloss: 0.641452	valid's binary_logloss: 0.655218
[400]	train's binary_logloss: 0.637601	valid's binary_logloss: 0.654481
[500]	train's binary_logloss: 0.634142	valid's binary_logloss: 0.653967
[600]	train's binary_logloss: 0.631213	valid's binary_logloss: 0.654248
Early stopping, best iteration is:
[539]	train's binary_logloss: 0.632957	valid's binary_logloss: 0.653768
min_data_in_leaf, val_score: 0.653307:  60%|######    | 3/5 [00:03<00:01,  1.02it/s][I 2020-09-27 04:42:24,298] Trial 65 finished with value: 0.653767937805501 and parameters: {'min_child_samples': 10}. Best is trial 65 with value: 0.653767937805501.
min_data_in_leaf, val_score: 0.653307:  60%|######    | 3/5 [00:03<00:01,  1.02it/s][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000881 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657308	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646871	valid's binary_logloss: 0.657816
[300]	train's binary_logloss: 0.641452	valid's binary_logloss: 0.655693
[400]	train's binary_logloss: 0.637718	valid's binary_logloss: 0.654576
[500]	train's binary_logloss: 0.634473	valid's binary_logloss: 0.654212
[600]	train's binary_logloss: 0.631693	valid's binary_logloss: 0.654007
Early stopping, best iteration is:
[580]	train's binary_logloss: 0.63221	valid's binary_logloss: 0.653668
min_data_in_leaf, val_score: 0.653307:  80%|########  | 4/5 [00:04<00:01,  1.01s/it][I 2020-09-27 04:42:25,378] Trial 66 finished with value: 0.6536677236980378 and parameters: {'min_child_samples': 50}. Best is trial 66 with value: 0.6536677236980378.
min_data_in_leaf, val_score: 0.653307:  80%|########  | 4/5 [00:04<00:01,  1.01s/it][LightGBM] [Info] Number of positive: 13145, number of negative: 12854
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004998 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505596 -> initscore=0.022386
[LightGBM] [Info] Start training from score 0.022386
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657308	valid's binary_logloss: 0.665041
[200]	train's binary_logloss: 0.646872	valid's binary_logloss: 0.657799
[300]	train's binary_logloss: 0.641589	valid's binary_logloss: 0.655365
[400]	train's binary_logloss: 0.638214	valid's binary_logloss: 0.65449
[500]	train's binary_logloss: 0.635191	valid's binary_logloss: 0.653777
[600]	train's binary_logloss: 0.632659	valid's binary_logloss: 0.653578
Early stopping, best iteration is:
[576]	train's binary_logloss: 0.633154	valid's binary_logloss: 0.653401
min_data_in_leaf, val_score: 0.653307: 100%|##########| 5/5 [00:05<00:00,  1.17s/it][I 2020-09-27 04:42:26,912] Trial 67 finished with value: 0.653400524735626 and parameters: {'min_child_samples': 100}. Best is trial 67 with value: 0.653400524735626.
min_data_in_leaf, val_score: 0.653307: 100%|##########| 5/5 [00:05<00:00,  1.15s/it]
Fold : 4
[I 2020-09-27 04:42:26,996] A new study created in memory with name: no-name-95b73563-c9f0-498a-9add-fe0ff51242f2
feature_fraction, val_score: inf:   0%|          | 0/7 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000967 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.569629	valid's binary_logloss: 0.666525
Early stopping, best iteration is:
[63]	train's binary_logloss: 0.595461	valid's binary_logloss: 0.665516
feature_fraction, val_score: 0.665516:  14%|#4        | 1/7 [00:00<00:03,  1.75it/s][I 2020-09-27 04:42:27,577] Trial 0 finished with value: 0.6655156235172205 and parameters: {'feature_fraction': 1.0}. Best is trial 0 with value: 0.6655156235172205.
feature_fraction, val_score: 0.665516:  14%|#4        | 1/7 [00:00<00:03,  1.75it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000965 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.572137	valid's binary_logloss: 0.663019
Early stopping, best iteration is:
[48]	train's binary_logloss: 0.610212	valid's binary_logloss: 0.661166
feature_fraction, val_score: 0.661166:  29%|##8       | 2/7 [00:01<00:02,  1.85it/s][I 2020-09-27 04:42:28,047] Trial 1 finished with value: 0.6611663859231157 and parameters: {'feature_fraction': 0.8}. Best is trial 1 with value: 0.6611663859231157.
feature_fraction, val_score: 0.661166:  29%|##8       | 2/7 [00:01<00:02,  1.85it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004690 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.57535	valid's binary_logloss: 0.662198
Early stopping, best iteration is:
[58]	train's binary_logloss: 0.604866	valid's binary_logloss: 0.660561
feature_fraction, val_score: 0.660561:  43%|####2     | 3/7 [00:01<00:02,  1.92it/s][I 2020-09-27 04:42:28,523] Trial 2 finished with value: 0.660560693860469 and parameters: {'feature_fraction': 0.6}. Best is trial 2 with value: 0.660560693860469.
feature_fraction, val_score: 0.660561:  43%|####2     | 3/7 [00:01<00:02,  1.92it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000938 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.571052	valid's binary_logloss: 0.6616
Early stopping, best iteration is:
[50]	train's binary_logloss: 0.607427	valid's binary_logloss: 0.660867
feature_fraction, val_score: 0.660561:  57%|#####7    | 4/7 [00:02<00:01,  1.94it/s][I 2020-09-27 04:42:29,026] Trial 3 finished with value: 0.6608672257275293 and parameters: {'feature_fraction': 0.8999999999999999}. Best is trial 2 with value: 0.660560693860469.
feature_fraction, val_score: 0.660561:  57%|#####7    | 4/7 [00:02<00:01,  1.94it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000868 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.574313	valid's binary_logloss: 0.662136
Early stopping, best iteration is:
[97]	train's binary_logloss: 0.576201	valid's binary_logloss: 0.661927
feature_fraction, val_score: 0.660561:  71%|#######1  | 5/7 [00:02<00:01,  1.79it/s][I 2020-09-27 04:42:29,690] Trial 4 finished with value: 0.661927284753647 and parameters: {'feature_fraction': 0.7}. Best is trial 2 with value: 0.660560693860469.
feature_fraction, val_score: 0.660561:  71%|#######1  | 5/7 [00:02<00:01,  1.79it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012827 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.578468	valid's binary_logloss: 0.661585
Early stopping, best iteration is:
[54]	train's binary_logloss: 0.610009	valid's binary_logloss: 0.661106
feature_fraction, val_score: 0.660561:  86%|########5 | 6/7 [00:03<00:00,  1.58it/s][I 2020-09-27 04:42:30,492] Trial 5 finished with value: 0.6611057071152201 and parameters: {'feature_fraction': 0.5}. Best is trial 2 with value: 0.660560693860469.
feature_fraction, val_score: 0.660561:  86%|########5 | 6/7 [00:03<00:00,  1.58it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000508 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.581603	valid's binary_logloss: 0.660601
Early stopping, best iteration is:
[81]	train's binary_logloss: 0.593654	valid's binary_logloss: 0.660443
feature_fraction, val_score: 0.660443: 100%|##########| 7/7 [00:03<00:00,  1.68it/s][I 2020-09-27 04:42:30,996] Trial 6 finished with value: 0.6604425614247917 and parameters: {'feature_fraction': 0.4}. Best is trial 6 with value: 0.6604425614247917.
feature_fraction, val_score: 0.660443: 100%|##########| 7/7 [00:04<00:00,  1.75it/s]
num_leaves, val_score: 0.660443:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004658 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.356495	valid's binary_logloss: 0.671539
Early stopping, best iteration is:
[41]	train's binary_logloss: 0.499664	valid's binary_logloss: 0.664498
num_leaves, val_score: 0.660443:   5%|5         | 1/20 [00:00<00:16,  1.14it/s][I 2020-09-27 04:42:31,893] Trial 7 finished with value: 0.6644984899613434 and parameters: {'num_leaves': 190}. Best is trial 7 with value: 0.6644984899613434.
num_leaves, val_score: 0.660443:   5%|5         | 1/20 [00:00<00:16,  1.14it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000609 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.526375	valid's binary_logloss: 0.665053
Early stopping, best iteration is:
[67]	train's binary_logloss: 0.561219	valid's binary_logloss: 0.662832
num_leaves, val_score: 0.660443:  10%|#         | 2/20 [00:01<00:14,  1.28it/s][I 2020-09-27 04:42:32,449] Trial 8 finished with value: 0.6628317798276503 and parameters: {'num_leaves': 60}. Best is trial 8 with value: 0.6628317798276503.
num_leaves, val_score: 0.660443:  10%|#         | 2/20 [00:01<00:14,  1.28it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000379 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.608789	valid's binary_logloss: 0.659931
[200]	train's binary_logloss: 0.573948	valid's binary_logloss: 0.661564
Early stopping, best iteration is:
[106]	train's binary_logloss: 0.606284	valid's binary_logloss: 0.659493
num_leaves, val_score: 0.659493:  15%|#5        | 3/20 [00:01<00:11,  1.48it/s][I 2020-09-27 04:42:32,881] Trial 9 finished with value: 0.6594930122196023 and parameters: {'num_leaves': 19}. Best is trial 9 with value: 0.6594930122196023.
num_leaves, val_score: 0.659493:  15%|#5        | 3/20 [00:01<00:11,  1.48it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000374 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.62851	valid's binary_logloss: 0.661694
[200]	train's binary_logloss: 0.606786	valid's binary_logloss: 0.660582
Early stopping, best iteration is:
[177]	train's binary_logloss: 0.611571	valid's binary_logloss: 0.660135
num_leaves, val_score: 0.659493:  20%|##        | 4/20 [00:02<00:10,  1.56it/s][I 2020-09-27 04:42:33,444] Trial 10 finished with value: 0.6601347890202464 and parameters: {'num_leaves': 11}. Best is trial 9 with value: 0.6594930122196023.
num_leaves, val_score: 0.659493:  20%|##        | 4/20 [00:02<00:10,  1.56it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.002220 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.623153	valid's binary_logloss: 0.659612
[200]	train's binary_logloss: 0.598247	valid's binary_logloss: 0.660773
Early stopping, best iteration is:
[106]	train's binary_logloss: 0.621362	valid's binary_logloss: 0.659491
num_leaves, val_score: 0.659491:  25%|##5       | 5/20 [00:03<00:10,  1.41it/s][I 2020-09-27 04:42:34,315] Trial 11 finished with value: 0.6594905167648981 and parameters: {'num_leaves': 13}. Best is trial 11 with value: 0.6594905167648981.
num_leaves, val_score: 0.659491:  25%|##5       | 5/20 [00:03<00:10,  1.41it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000240 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.633624	valid's binary_logloss: 0.661805
[200]	train's binary_logloss: 0.6154	valid's binary_logloss: 0.661626
Early stopping, best iteration is:
[118]	train's binary_logloss: 0.629667	valid's binary_logloss: 0.661053
num_leaves, val_score: 0.659491:  30%|###       | 6/20 [00:03<00:08,  1.61it/s][I 2020-09-27 04:42:34,733] Trial 12 finished with value: 0.6610531437027433 and parameters: {'num_leaves': 9}. Best is trial 11 with value: 0.6594905167648981.
num_leaves, val_score: 0.659491:  30%|###       | 6/20 [00:03<00:08,  1.61it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000443 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.489932	valid's binary_logloss: 0.664356
Early stopping, best iteration is:
[42]	train's binary_logloss: 0.574496	valid's binary_logloss: 0.661322
num_leaves, val_score: 0.659491:  35%|###5      | 7/20 [00:04<00:07,  1.63it/s][I 2020-09-27 04:42:35,325] Trial 13 finished with value: 0.6613221980695407 and parameters: {'num_leaves': 83}. Best is trial 11 with value: 0.6594905167648981.
num_leaves, val_score: 0.659491:  35%|###5      | 7/20 [00:04<00:07,  1.63it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000456 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.668466	valid's binary_logloss: 0.674773
[200]	train's binary_logloss: 0.658603	valid's binary_logloss: 0.667149
[300]	train's binary_logloss: 0.653224	valid's binary_logloss: 0.663743
[400]	train's binary_logloss: 0.649988	valid's binary_logloss: 0.661832
[500]	train's binary_logloss: 0.647929	valid's binary_logloss: 0.661043
[600]	train's binary_logloss: 0.646563	valid's binary_logloss: 0.660633
[700]	train's binary_logloss: 0.645608	valid's binary_logloss: 0.660633
[800]	train's binary_logloss: 0.644889	valid's binary_logloss: 0.660438
[900]	train's binary_logloss: 0.644316	valid's binary_logloss: 0.66034
[1000]	train's binary_logloss: 0.64384	valid's binary_logloss: 0.660316
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.64384	valid's binary_logloss: 0.660316
num_leaves, val_score: 0.659491:  40%|####      | 8/20 [00:05<00:09,  1.28it/s][I 2020-09-27 04:42:36,492] Trial 14 finished with value: 0.6603164467173235 and parameters: {'num_leaves': 2}. Best is trial 11 with value: 0.6594905167648981.
num_leaves, val_score: 0.659491:  40%|####      | 8/20 [00:05<00:09,  1.28it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000395 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.481541	valid's binary_logloss: 0.666981
Early stopping, best iteration is:
[44]	train's binary_logloss: 0.566666	valid's binary_logloss: 0.662735
num_leaves, val_score: 0.659491:  45%|####5     | 9/20 [00:06<00:08,  1.35it/s][I 2020-09-27 04:42:37,143] Trial 15 finished with value: 0.6627351233501201 and parameters: {'num_leaves': 88}. Best is trial 11 with value: 0.6594905167648981.
num_leaves, val_score: 0.659491:  45%|####5     | 9/20 [00:06<00:08,  1.35it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000349 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.3973	valid's binary_logloss: 0.672802
Early stopping, best iteration is:
[29]	train's binary_logloss: 0.560753	valid's binary_logloss: 0.66553
num_leaves, val_score: 0.659491:  50%|#####     | 10/20 [00:07<00:09,  1.08it/s][I 2020-09-27 04:42:38,498] Trial 16 finished with value: 0.6655299653769075 and parameters: {'num_leaves': 152}. Best is trial 11 with value: 0.6594905167648981.
num_leaves, val_score: 0.659491:  50%|#####     | 10/20 [00:07<00:09,  1.08it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000373 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.299677	valid's binary_logloss: 0.679053
Early stopping, best iteration is:
[28]	train's binary_logloss: 0.514613	valid's binary_logloss: 0.667796
num_leaves, val_score: 0.659491:  55%|#####5    | 11/20 [00:08<00:08,  1.03it/s][I 2020-09-27 04:42:39,559] Trial 17 finished with value: 0.6677957520143049 and parameters: {'num_leaves': 256}. Best is trial 11 with value: 0.6594905167648981.
num_leaves, val_score: 0.659491:  55%|#####5    | 11/20 [00:08<00:08,  1.03it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000315 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.5506	valid's binary_logloss: 0.661771
Early stopping, best iteration is:
[46]	train's binary_logloss: 0.602349	valid's binary_logloss: 0.660582
num_leaves, val_score: 0.659491:  60%|######    | 12/20 [00:08<00:06,  1.23it/s][I 2020-09-27 04:42:40,010] Trial 18 finished with value: 0.6605823886498043 and parameters: {'num_leaves': 47}. Best is trial 11 with value: 0.6594905167648981.
num_leaves, val_score: 0.659491:  60%|######    | 12/20 [00:09<00:06,  1.23it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000360 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.573596	valid's binary_logloss: 0.659418
Early stopping, best iteration is:
[95]	train's binary_logloss: 0.576854	valid's binary_logloss: 0.658933
num_leaves, val_score: 0.658933:  65%|######5   | 13/20 [00:09<00:04,  1.40it/s][I 2020-09-27 04:42:40,490] Trial 19 finished with value: 0.6589331658992789 and parameters: {'num_leaves': 35}. Best is trial 19 with value: 0.6589331658992789.
num_leaves, val_score: 0.658933:  65%|######5   | 13/20 [00:09<00:04,  1.40it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000370 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.433641	valid's binary_logloss: 0.669498
Early stopping, best iteration is:
[30]	train's binary_logloss: 0.5737	valid's binary_logloss: 0.666057
num_leaves, val_score: 0.658933:  70%|#######   | 14/20 [00:10<00:04,  1.31it/s][I 2020-09-27 04:42:41,370] Trial 20 finished with value: 0.666056567776029 and parameters: {'num_leaves': 123}. Best is trial 19 with value: 0.6589331658992789.
num_leaves, val_score: 0.658933:  70%|#######   | 14/20 [00:10<00:04,  1.31it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.011203 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.578104	valid's binary_logloss: 0.662017
Early stopping, best iteration is:
[72]	train's binary_logloss: 0.596911	valid's binary_logloss: 0.661574
num_leaves, val_score: 0.658933:  75%|#######5  | 15/20 [00:10<00:03,  1.41it/s][I 2020-09-27 04:42:41,958] Trial 21 finished with value: 0.6615737018014829 and parameters: {'num_leaves': 33}. Best is trial 19 with value: 0.6589331658992789.
num_leaves, val_score: 0.658933:  75%|#######5  | 15/20 [00:10<00:03,  1.41it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000381 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.668466	valid's binary_logloss: 0.674773
[200]	train's binary_logloss: 0.658603	valid's binary_logloss: 0.667149
[300]	train's binary_logloss: 0.653224	valid's binary_logloss: 0.663743
[400]	train's binary_logloss: 0.649988	valid's binary_logloss: 0.661832
[500]	train's binary_logloss: 0.647929	valid's binary_logloss: 0.661043
[600]	train's binary_logloss: 0.646563	valid's binary_logloss: 0.660633
[700]	train's binary_logloss: 0.645608	valid's binary_logloss: 0.660633
[800]	train's binary_logloss: 0.644889	valid's binary_logloss: 0.660438
[900]	train's binary_logloss: 0.644316	valid's binary_logloss: 0.66034
[1000]	train's binary_logloss: 0.64384	valid's binary_logloss: 0.660316
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.64384	valid's binary_logloss: 0.660316
num_leaves, val_score: 0.658933:  80%|########  | 16/20 [00:12<00:03,  1.18it/s][I 2020-09-27 04:42:43,132] Trial 22 finished with value: 0.6603164467173235 and parameters: {'num_leaves': 2}. Best is trial 19 with value: 0.6589331658992789.
num_leaves, val_score: 0.658933:  80%|########  | 16/20 [00:12<00:03,  1.18it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000388 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.581603	valid's binary_logloss: 0.660601
Early stopping, best iteration is:
[81]	train's binary_logloss: 0.593654	valid's binary_logloss: 0.660443
num_leaves, val_score: 0.658933:  85%|########5 | 17/20 [00:12<00:02,  1.34it/s][I 2020-09-27 04:42:43,636] Trial 23 finished with value: 0.6604425614247916 and parameters: {'num_leaves': 31}. Best is trial 19 with value: 0.6589331658992789.
num_leaves, val_score: 0.658933:  85%|########5 | 17/20 [00:12<00:02,  1.34it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000430 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.517038	valid's binary_logloss: 0.664708
Early stopping, best iteration is:
[42]	train's binary_logloss: 0.589858	valid's binary_logloss: 0.660965
num_leaves, val_score: 0.658933:  90%|######### | 18/20 [00:13<00:01,  1.47it/s][I 2020-09-27 04:42:44,171] Trial 24 finished with value: 0.660965322744683 and parameters: {'num_leaves': 66}. Best is trial 19 with value: 0.6589331658992789.
num_leaves, val_score: 0.658933:  90%|######### | 18/20 [00:13<00:01,  1.47it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000404 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.590111	valid's binary_logloss: 0.659776
Early stopping, best iteration is:
[96]	train's binary_logloss: 0.592038	valid's binary_logloss: 0.65953
num_leaves, val_score: 0.658933:  95%|#########5| 19/20 [00:13<00:00,  1.61it/s][I 2020-09-27 04:42:44,655] Trial 25 finished with value: 0.6595301896319116 and parameters: {'num_leaves': 27}. Best is trial 19 with value: 0.6589331658992789.
num_leaves, val_score: 0.658933:  95%|#########5| 19/20 [00:13<00:00,  1.61it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000349 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.454148	valid's binary_logloss: 0.667807
Early stopping, best iteration is:
[45]	train's binary_logloss: 0.54925	valid's binary_logloss: 0.664334
num_leaves, val_score: 0.658933: 100%|##########| 20/20 [00:14<00:00,  1.28it/s][I 2020-09-27 04:42:45,813] Trial 26 finished with value: 0.6643342584470427 and parameters: {'num_leaves': 106}. Best is trial 19 with value: 0.6589331658992789.
num_leaves, val_score: 0.658933: 100%|##########| 20/20 [00:14<00:00,  1.35it/s]
bagging, val_score: 0.658933:   0%|          | 0/10 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000378 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.57621	valid's binary_logloss: 0.663844
Early stopping, best iteration is:
[82]	train's binary_logloss: 0.587864	valid's binary_logloss: 0.662968
bagging, val_score: 0.658933:  10%|#         | 1/10 [00:00<00:04,  1.88it/s][I 2020-09-27 04:42:46,361] Trial 27 finished with value: 0.6629683704520333 and parameters: {'bagging_fraction': 0.6524827037560431, 'bagging_freq': 6}. Best is trial 27 with value: 0.6629683704520333.
bagging, val_score: 0.658933:  10%|#         | 1/10 [00:00<00:04,  1.88it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000416 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.573429	valid's binary_logloss: 0.66344
Early stopping, best iteration is:
[70]	train's binary_logloss: 0.595186	valid's binary_logloss: 0.661013
bagging, val_score: 0.658933:  20%|##        | 2/10 [00:00<00:04,  1.96it/s][I 2020-09-27 04:42:46,824] Trial 28 finished with value: 0.6610133814310396 and parameters: {'bagging_fraction': 0.9909619380758035, 'bagging_freq': 1}. Best is trial 28 with value: 0.6610133814310396.
bagging, val_score: 0.658933:  20%|##        | 2/10 [00:01<00:04,  1.96it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000474 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.580089	valid's binary_logloss: 0.668617
Early stopping, best iteration is:
[39]	train's binary_logloss: 0.625728	valid's binary_logloss: 0.665767
bagging, val_score: 0.658933:  30%|###       | 3/10 [00:01<00:03,  2.09it/s][I 2020-09-27 04:42:47,229] Trial 29 finished with value: 0.6657670269268177 and parameters: {'bagging_fraction': 0.42763463150211456, 'bagging_freq': 1}. Best is trial 28 with value: 0.6610133814310396.
bagging, val_score: 0.658933:  30%|###       | 3/10 [00:01<00:03,  2.09it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000410 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.573775	valid's binary_logloss: 0.662206
Early stopping, best iteration is:
[84]	train's binary_logloss: 0.585244	valid's binary_logloss: 0.661816
bagging, val_score: 0.658933:  40%|####      | 4/10 [00:01<00:02,  2.03it/s][I 2020-09-27 04:42:47,755] Trial 30 finished with value: 0.6618156054088046 and parameters: {'bagging_fraction': 0.9647411027897097, 'bagging_freq': 7}. Best is trial 28 with value: 0.6610133814310396.
bagging, val_score: 0.658933:  40%|####      | 4/10 [00:01<00:02,  2.03it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004478 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579254	valid's binary_logloss: 0.665041
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607077	valid's binary_logloss: 0.658928
bagging, val_score: 0.658928:  50%|#####     | 5/10 [00:02<00:02,  2.12it/s][I 2020-09-27 04:42:48,176] Trial 31 finished with value: 0.6589276849000186 and parameters: {'bagging_fraction': 0.4607699724160956, 'bagging_freq': 4}. Best is trial 31 with value: 0.6589276849000186.
bagging, val_score: 0.658928:  50%|#####     | 5/10 [00:02<00:02,  2.12it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000381 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.580916	valid's binary_logloss: 0.671067
Early stopping, best iteration is:
[36]	train's binary_logloss: 0.628471	valid's binary_logloss: 0.662306
bagging, val_score: 0.658928:  60%|######    | 6/10 [00:03<00:02,  1.82it/s][I 2020-09-27 04:42:48,914] Trial 32 finished with value: 0.6623061121182913 and parameters: {'bagging_fraction': 0.4089026397973168, 'bagging_freq': 4}. Best is trial 31 with value: 0.6589276849000186.
bagging, val_score: 0.658928:  60%|######    | 6/10 [00:03<00:02,  1.82it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001551 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.57696	valid's binary_logloss: 0.664792
Early stopping, best iteration is:
[44]	train's binary_logloss: 0.620171	valid's binary_logloss: 0.661311
bagging, val_score: 0.658928:  70%|#######   | 7/10 [00:03<00:01,  1.75it/s][I 2020-09-27 04:42:49,530] Trial 33 finished with value: 0.6613113079015099 and parameters: {'bagging_fraction': 0.5861812306541639, 'bagging_freq': 4}. Best is trial 31 with value: 0.6589276849000186.
bagging, val_score: 0.658928:  70%|#######   | 7/10 [00:03<00:01,  1.75it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000437 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.575286	valid's binary_logloss: 0.663569
Early stopping, best iteration is:
[62]	train's binary_logloss: 0.60231	valid's binary_logloss: 0.661403
bagging, val_score: 0.658928:  80%|########  | 8/10 [00:04<00:01,  1.80it/s][I 2020-09-27 04:42:50,048] Trial 34 finished with value: 0.6614028699158252 and parameters: {'bagging_fraction': 0.8070642452561286, 'bagging_freq': 3}. Best is trial 31 with value: 0.6589276849000186.
bagging, val_score: 0.658928:  80%|########  | 8/10 [00:04<00:01,  1.80it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000238 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.578059	valid's binary_logloss: 0.66561
Early stopping, best iteration is:
[57]	train's binary_logloss: 0.60883	valid's binary_logloss: 0.660744
bagging, val_score: 0.658928:  90%|######### | 9/10 [00:04<00:00,  1.85it/s][I 2020-09-27 04:42:50,558] Trial 35 finished with value: 0.6607440151387087 and parameters: {'bagging_fraction': 0.5229694779889069, 'bagging_freq': 5}. Best is trial 31 with value: 0.6589276849000186.
bagging, val_score: 0.658928:  90%|######### | 9/10 [00:04<00:00,  1.85it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000982 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.574809	valid's binary_logloss: 0.665725
Early stopping, best iteration is:
[73]	train's binary_logloss: 0.593617	valid's binary_logloss: 0.664242
bagging, val_score: 0.658928: 100%|##########| 10/10 [00:05<00:00,  1.76it/s][I 2020-09-27 04:42:51,184] Trial 36 finished with value: 0.6642418801252681 and parameters: {'bagging_fraction': 0.7857998781191295, 'bagging_freq': 2}. Best is trial 31 with value: 0.6589276849000186.
bagging, val_score: 0.658928: 100%|##########| 10/10 [00:05<00:00,  1.86it/s]
feature_fraction_stage2, val_score: 0.658928:   0%|          | 0/3 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004751 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.576158	valid's binary_logloss: 0.667365
Early stopping, best iteration is:
[40]	train's binary_logloss: 0.622262	valid's binary_logloss: 0.662845
feature_fraction_stage2, val_score: 0.658928:  33%|###3      | 1/3 [00:00<00:00,  2.32it/s][I 2020-09-27 04:42:51,633] Trial 37 finished with value: 0.6628451796786387 and parameters: {'feature_fraction': 0.44800000000000006}. Best is trial 37 with value: 0.6628451796786387.
feature_fraction_stage2, val_score: 0.658928:  33%|###3      | 1/3 [00:00<00:00,  2.32it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000343 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.577968	valid's binary_logloss: 0.668202
Early stopping, best iteration is:
[34]	train's binary_logloss: 0.629452	valid's binary_logloss: 0.662397
feature_fraction_stage2, val_score: 0.658928:  67%|######6   | 2/3 [00:00<00:00,  2.34it/s][I 2020-09-27 04:42:52,051] Trial 38 finished with value: 0.6623967977826815 and parameters: {'feature_fraction': 0.41600000000000004}. Best is trial 38 with value: 0.6623967977826815.
feature_fraction_stage2, val_score: 0.658928:  67%|######6   | 2/3 [00:00<00:00,  2.34it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.007814 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.576158	valid's binary_logloss: 0.667365
Early stopping, best iteration is:
[40]	train's binary_logloss: 0.622262	valid's binary_logloss: 0.662845
feature_fraction_stage2, val_score: 0.658928: 100%|##########| 3/3 [00:01<00:00,  2.00it/s][I 2020-09-27 04:42:52,719] Trial 39 finished with value: 0.6628451796786387 and parameters: {'feature_fraction': 0.48000000000000004}. Best is trial 38 with value: 0.6623967977826815.
feature_fraction_stage2, val_score: 0.658928: 100%|##########| 3/3 [00:01<00:00,  1.97it/s]
regularization_factors, val_score: 0.658928:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016489 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.578756	valid's binary_logloss: 0.666904
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.606361	valid's binary_logloss: 0.662873
regularization_factors, val_score: 0.658928:   5%|5         | 1/20 [00:00<00:12,  1.52it/s][I 2020-09-27 04:42:53,400] Trial 40 finished with value: 0.6628730879769316 and parameters: {'lambda_l1': 0.00027860521966329625, 'lambda_l2': 0.23171527606664835}. Best is trial 40 with value: 0.6628730879769316.
regularization_factors, val_score: 0.658928:   5%|5         | 1/20 [00:00<00:12,  1.52it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000377 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.57899	valid's binary_logloss: 0.664607
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607074	valid's binary_logloss: 0.658905
regularization_factors, val_score: 0.658905:  10%|#         | 2/20 [00:01<00:11,  1.62it/s][I 2020-09-27 04:42:53,923] Trial 41 finished with value: 0.6589045464246605 and parameters: {'lambda_l1': 1.0562160571745345e-08, 'lambda_l2': 1.2335861617943539e-08}. Best is trial 41 with value: 0.6589045464246605.
regularization_factors, val_score: 0.658905:  10%|#         | 2/20 [00:01<00:11,  1.62it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000451 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579401	valid's binary_logloss: 0.664539
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607081	valid's binary_logloss: 0.658927
regularization_factors, val_score: 0.658905:  15%|#5        | 3/20 [00:01<00:09,  1.70it/s][I 2020-09-27 04:42:54,440] Trial 42 finished with value: 0.6589271408352281 and parameters: {'lambda_l1': 1.4846795215967304e-08, 'lambda_l2': 2.1890701373620828e-08}. Best is trial 41 with value: 0.6589045464246605.
regularization_factors, val_score: 0.658905:  15%|#5        | 3/20 [00:01<00:09,  1.70it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000474 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579252	valid's binary_logloss: 0.665028
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607075	valid's binary_logloss: 0.658927
regularization_factors, val_score: 0.658905:  20%|##        | 4/20 [00:02<00:09,  1.71it/s][I 2020-09-27 04:42:55,018] Trial 43 finished with value: 0.6589273467253626 and parameters: {'lambda_l1': 1.1433697445980986e-08, 'lambda_l2': 1.3023515499807511e-08}. Best is trial 41 with value: 0.6589045464246605.
regularization_factors, val_score: 0.658905:  20%|##        | 4/20 [00:02<00:09,  1.71it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000451 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579216	valid's binary_logloss: 0.664431
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607079	valid's binary_logloss: 0.658949
regularization_factors, val_score: 0.658905:  25%|##5       | 5/20 [00:02<00:08,  1.75it/s][I 2020-09-27 04:42:55,554] Trial 44 finished with value: 0.6589487630970773 and parameters: {'lambda_l1': 1.0140056425681878e-08, 'lambda_l2': 1.2300058074804937e-08}. Best is trial 41 with value: 0.6589045464246605.
regularization_factors, val_score: 0.658905:  25%|##5       | 5/20 [00:02<00:08,  1.75it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000479 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.578759	valid's binary_logloss: 0.665688
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607073	valid's binary_logloss: 0.658904
regularization_factors, val_score: 0.658904:  30%|###       | 6/20 [00:03<00:07,  1.84it/s][I 2020-09-27 04:42:56,039] Trial 45 finished with value: 0.658904002539462 and parameters: {'lambda_l1': 1.0953495812808097e-08, 'lambda_l2': 1.828459432273254e-08}. Best is trial 45 with value: 0.658904002539462.
regularization_factors, val_score: 0.658904:  30%|###       | 6/20 [00:03<00:07,  1.84it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000378 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579237	valid's binary_logloss: 0.665024
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.60708	valid's binary_logloss: 0.658928
regularization_factors, val_score: 0.658904:  35%|###5      | 7/20 [00:04<00:08,  1.49it/s][I 2020-09-27 04:42:57,006] Trial 46 finished with value: 0.6589280313288517 and parameters: {'lambda_l1': 1.0619074780064361e-08, 'lambda_l2': 1.470079266120941e-08}. Best is trial 45 with value: 0.658904002539462.
regularization_factors, val_score: 0.658904:  35%|###5      | 7/20 [00:04<00:08,  1.49it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000339 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.5794	valid's binary_logloss: 0.66454
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607078	valid's binary_logloss: 0.658904
regularization_factors, val_score: 0.658904:  40%|####      | 8/20 [00:04<00:07,  1.63it/s][I 2020-09-27 04:42:57,486] Trial 47 finished with value: 0.6589040025411772 and parameters: {'lambda_l1': 1.3254321926524086e-08, 'lambda_l2': 2.193885866761881e-08}. Best is trial 45 with value: 0.658904002539462.
regularization_factors, val_score: 0.658904:  40%|####      | 8/20 [00:04<00:07,  1.63it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000424 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579223	valid's binary_logloss: 0.664396
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607078	valid's binary_logloss: 0.658927
regularization_factors, val_score: 0.658904:  45%|####5     | 9/20 [00:05<00:06,  1.76it/s][I 2020-09-27 04:42:57,951] Trial 48 finished with value: 0.6589267944230356 and parameters: {'lambda_l1': 2.4619075323767072e-08, 'lambda_l2': 2.017164260358036e-08}. Best is trial 45 with value: 0.658904002539462.
regularization_factors, val_score: 0.658904:  45%|####5     | 9/20 [00:05<00:06,  1.76it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000536 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.57921	valid's binary_logloss: 0.664407
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607077	valid's binary_logloss: 0.658925
regularization_factors, val_score: 0.658904:  50%|#####     | 10/20 [00:05<00:05,  1.81it/s][I 2020-09-27 04:42:58,460] Trial 49 finished with value: 0.6589253929659314 and parameters: {'lambda_l1': 1.2653815920250245e-08, 'lambda_l2': 2.3139792938926494e-08}. Best is trial 45 with value: 0.658904002539462.
regularization_factors, val_score: 0.658904:  50%|#####     | 10/20 [00:05<00:05,  1.81it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004484 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579211	valid's binary_logloss: 0.664425
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607081	valid's binary_logloss: 0.658927
regularization_factors, val_score: 0.658904:  55%|#####5    | 11/20 [00:06<00:04,  1.90it/s][I 2020-09-27 04:42:58,929] Trial 50 finished with value: 0.6589271408099914 and parameters: {'lambda_l1': 1.3776160562135245e-08, 'lambda_l2': 2.6228878428660015e-07}. Best is trial 45 with value: 0.658904002539462.
regularization_factors, val_score: 0.658904:  55%|#####5    | 11/20 [00:06<00:04,  1.90it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000373 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.57921	valid's binary_logloss: 0.664409
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607081	valid's binary_logloss: 0.658927
regularization_factors, val_score: 0.658904:  60%|######    | 12/20 [00:06<00:04,  1.94it/s][I 2020-09-27 04:42:59,423] Trial 51 finished with value: 0.658927140833919 and parameters: {'lambda_l1': 1.664666646319637e-08, 'lambda_l2': 4.250351022086274e-08}. Best is trial 45 with value: 0.658904002539462.
regularization_factors, val_score: 0.658904:  60%|######    | 12/20 [00:06<00:04,  1.94it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000390 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.57924	valid's binary_logloss: 0.665044
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607082	valid's binary_logloss: 0.65895
regularization_factors, val_score: 0.658904:  65%|######5   | 13/20 [00:07<00:03,  1.94it/s][I 2020-09-27 04:42:59,937] Trial 52 finished with value: 0.6589497680473425 and parameters: {'lambda_l1': 1.439041833170667e-08, 'lambda_l2': 3.7007408240165166e-07}. Best is trial 45 with value: 0.658904002539462.
regularization_factors, val_score: 0.658904:  65%|######5   | 13/20 [00:07<00:03,  1.94it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000634 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579255	valid's binary_logloss: 0.665048
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607079	valid's binary_logloss: 0.658949
regularization_factors, val_score: 0.658904:  70%|#######   | 14/20 [00:08<00:04,  1.47it/s][I 2020-09-27 04:43:01,006] Trial 53 finished with value: 0.6589494209805504 and parameters: {'lambda_l1': 1.7572183872374385e-07, 'lambda_l2': 9.82651046880018e-07}. Best is trial 45 with value: 0.658904002539462.
regularization_factors, val_score: 0.658904:  70%|#######   | 14/20 [00:08<00:04,  1.47it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000481 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579214	valid's binary_logloss: 0.664386
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607075	valid's binary_logloss: 0.658904
regularization_factors, val_score: 0.658904:  75%|#######5  | 15/20 [00:08<00:03,  1.60it/s][I 2020-09-27 04:43:01,494] Trial 54 finished with value: 0.6589036559607917 and parameters: {'lambda_l1': 8.538730576389241e-07, 'lambda_l2': 1.1721978821541268e-08}. Best is trial 54 with value: 0.6589036559607917.
regularization_factors, val_score: 0.658904:  75%|#######5  | 15/20 [00:08<00:03,  1.60it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000414 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579216	valid's binary_logloss: 0.664476
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607081	valid's binary_logloss: 0.658927
regularization_factors, val_score: 0.658904:  80%|########  | 16/20 [00:09<00:02,  1.72it/s][I 2020-09-27 04:43:01,976] Trial 55 finished with value: 0.6589271405882069 and parameters: {'lambda_l1': 2.1323929878136295e-06, 'lambda_l2': 1.2914444598331435e-08}. Best is trial 54 with value: 0.6589036559607917.
regularization_factors, val_score: 0.658904:  80%|########  | 16/20 [00:09<00:02,  1.72it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004379 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579258	valid's binary_logloss: 0.665048
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607079	valid's binary_logloss: 0.658949
regularization_factors, val_score: 0.658904:  85%|########5 | 17/20 [00:09<00:01,  1.82it/s][I 2020-09-27 04:43:02,455] Trial 56 finished with value: 0.6589494210443969 and parameters: {'lambda_l1': 7.412222130813083e-07, 'lambda_l2': 1.1296738167689051e-08}. Best is trial 54 with value: 0.6589036559607917.
regularization_factors, val_score: 0.658904:  85%|########5 | 17/20 [00:09<00:01,  1.82it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000368 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579256	valid's binary_logloss: 0.665048
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607077	valid's binary_logloss: 0.658949
regularization_factors, val_score: 0.658904:  90%|######### | 18/20 [00:10<00:01,  1.87it/s][I 2020-09-27 04:43:02,951] Trial 57 finished with value: 0.6589493135745349 and parameters: {'lambda_l1': 2.9906109268867005e-07, 'lambda_l2': 1.0924154779547353e-05}. Best is trial 54 with value: 0.6589036559607917.
regularization_factors, val_score: 0.658904:  90%|######### | 18/20 [00:10<00:01,  1.87it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000413 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579254	valid's binary_logloss: 0.665027
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607078	valid's binary_logloss: 0.658927
regularization_factors, val_score: 0.658904:  95%|#########5| 19/20 [00:10<00:00,  1.95it/s][I 2020-09-27 04:43:03,413] Trial 58 finished with value: 0.658926859627446 and parameters: {'lambda_l1': 1.208558071943979e-07, 'lambda_l2': 1.4153597871067393e-07}. Best is trial 54 with value: 0.6589036559607917.
regularization_factors, val_score: 0.658904:  95%|#########5| 19/20 [00:10<00:00,  1.95it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004727 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579258	valid's binary_logloss: 0.665048
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607079	valid's binary_logloss: 0.658949
regularization_factors, val_score: 0.658904: 100%|##########| 20/20 [00:11<00:00,  1.94it/s][I 2020-09-27 04:43:03,937] Trial 59 finished with value: 0.6589494193426407 and parameters: {'lambda_l1': 1.386319739479547e-05, 'lambda_l2': 1.070237025489994e-08}. Best is trial 54 with value: 0.6589036559607917.
regularization_factors, val_score: 0.658904: 100%|##########| 20/20 [00:11<00:00,  1.78it/s]
min_data_in_leaf, val_score: 0.658904:   0%|          | 0/5 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000392 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.576809	valid's binary_logloss: 0.668517
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.605306	valid's binary_logloss: 0.663449
min_data_in_leaf, val_score: 0.658904:  20%|##        | 1/5 [00:00<00:03,  1.01it/s][I 2020-09-27 04:43:04,955] Trial 60 finished with value: 0.6634487376945314 and parameters: {'min_child_samples': 5}. Best is trial 60 with value: 0.6634487376945314.
min_data_in_leaf, val_score: 0.658904:  20%|##        | 1/5 [00:01<00:03,  1.01it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000379 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579781	valid's binary_logloss: 0.664717
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.607283	valid's binary_logloss: 0.661178
min_data_in_leaf, val_score: 0.658904:  40%|####      | 2/5 [00:01<00:02,  1.17it/s][I 2020-09-27 04:43:05,479] Trial 61 finished with value: 0.6611778602130459 and parameters: {'min_child_samples': 25}. Best is trial 61 with value: 0.6611778602130459.
min_data_in_leaf, val_score: 0.658904:  40%|####      | 2/5 [00:01<00:02,  1.17it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004590 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.587755	valid's binary_logloss: 0.665012
Early stopping, best iteration is:
[66]	train's binary_logloss: 0.607178	valid's binary_logloss: 0.662389
min_data_in_leaf, val_score: 0.658904:  60%|######    | 3/5 [00:02<00:01,  1.34it/s][I 2020-09-27 04:43:05,977] Trial 62 finished with value: 0.6623889382406033 and parameters: {'min_child_samples': 100}. Best is trial 61 with value: 0.6611778602130459.
min_data_in_leaf, val_score: 0.658904:  60%|######    | 3/5 [00:02<00:01,  1.34it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000482 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.577819	valid's binary_logloss: 0.667336
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.605667	valid's binary_logloss: 0.663345
min_data_in_leaf, val_score: 0.658904:  80%|########  | 4/5 [00:02<00:00,  1.50it/s][I 2020-09-27 04:43:06,461] Trial 63 finished with value: 0.6633454558808694 and parameters: {'min_child_samples': 10}. Best is trial 61 with value: 0.6611778602130459.
min_data_in_leaf, val_score: 0.658904:  80%|########  | 4/5 [00:02<00:00,  1.50it/s][LightGBM] [Info] Number of positive: 13155, number of negative: 12844
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000449 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.505981 -> initscore=0.023925
[LightGBM] [Info] Start training from score 0.023925
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.583153	valid's binary_logloss: 0.66795
Early stopping, best iteration is:
[59]	train's binary_logloss: 0.610418	valid's binary_logloss: 0.663074
min_data_in_leaf, val_score: 0.658904: 100%|##########| 5/5 [00:02<00:00,  1.63it/s][I 2020-09-27 04:43:06,953] Trial 64 finished with value: 0.6630735000484437 and parameters: {'min_child_samples': 50}. Best is trial 61 with value: 0.6611778602130459.
min_data_in_leaf, val_score: 0.658904: 100%|##########| 5/5 [00:02<00:00,  1.67it/s]
Fold : 5
[I 2020-09-27 04:43:06,989] A new study created in memory with name: no-name-b151d15a-5a47-42dc-ace0-4fe96bb3c92c
feature_fraction, val_score: inf:   0%|          | 0/7 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004614 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.577114	valid's binary_logloss: 0.654088
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.604064	valid's binary_logloss: 0.652612
feature_fraction, val_score: 0.652612:  14%|#4        | 1/7 [00:00<00:02,  2.09it/s][I 2020-09-27 04:43:07,480] Trial 0 finished with value: 0.6526118300374438 and parameters: {'feature_fraction': 0.6}. Best is trial 0 with value: 0.6526118300374438.
feature_fraction, val_score: 0.652612:  14%|#4        | 1/7 [00:00<00:02,  2.09it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001021 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.570555	valid's binary_logloss: 0.652172
[200]	train's binary_logloss: 0.514477	valid's binary_logloss: 0.655545
Early stopping, best iteration is:
[114]	train's binary_logloss: 0.561489	valid's binary_logloss: 0.651258
feature_fraction, val_score: 0.651258:  29%|##8       | 2/7 [00:01<00:03,  1.52it/s][I 2020-09-27 04:43:08,551] Trial 1 finished with value: 0.6512579255125732 and parameters: {'feature_fraction': 1.0}. Best is trial 1 with value: 0.6512579255125732.
feature_fraction, val_score: 0.651258:  29%|##8       | 2/7 [00:01<00:03,  1.52it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004468 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.583108	valid's binary_logloss: 0.653379
Early stopping, best iteration is:
[83]	train's binary_logloss: 0.593455	valid's binary_logloss: 0.652673
feature_fraction, val_score: 0.651258:  43%|####2     | 3/7 [00:01<00:02,  1.69it/s][I 2020-09-27 04:43:08,993] Trial 2 finished with value: 0.652673016771765 and parameters: {'feature_fraction': 0.4}. Best is trial 1 with value: 0.6512579255125732.
feature_fraction, val_score: 0.651258:  43%|####2     | 3/7 [00:01<00:02,  1.69it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000894 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.573161	valid's binary_logloss: 0.653887
[200]	train's binary_logloss: 0.518397	valid's binary_logloss: 0.656355
Early stopping, best iteration is:
[117]	train's binary_logloss: 0.562873	valid's binary_logloss: 0.652985
feature_fraction, val_score: 0.651258:  57%|#####7    | 4/7 [00:02<00:01,  1.66it/s][I 2020-09-27 04:43:09,624] Trial 3 finished with value: 0.6529847756787174 and parameters: {'feature_fraction': 0.8}. Best is trial 1 with value: 0.6512579255125732.
feature_fraction, val_score: 0.651258:  57%|#####7    | 4/7 [00:02<00:01,  1.66it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000779 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.574837	valid's binary_logloss: 0.653256
Early stopping, best iteration is:
[78]	train's binary_logloss: 0.589091	valid's binary_logloss: 0.651848
feature_fraction, val_score: 0.651258:  71%|#######1  | 5/7 [00:03<00:01,  1.75it/s][I 2020-09-27 04:43:10,125] Trial 4 finished with value: 0.6518483772036818 and parameters: {'feature_fraction': 0.7}. Best is trial 1 with value: 0.6512579255125732.
feature_fraction, val_score: 0.651258:  71%|#######1  | 5/7 [00:03<00:01,  1.75it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000916 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.572031	valid's binary_logloss: 0.652234
Early stopping, best iteration is:
[76]	train's binary_logloss: 0.588368	valid's binary_logloss: 0.651017
feature_fraction, val_score: 0.651017:  86%|########5 | 6/7 [00:03<00:00,  1.79it/s][I 2020-09-27 04:43:10,646] Trial 5 finished with value: 0.6510174434829193 and parameters: {'feature_fraction': 0.8999999999999999}. Best is trial 5 with value: 0.6510174434829193.
feature_fraction, val_score: 0.651017:  86%|########5 | 6/7 [00:03<00:00,  1.79it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000533 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.578051	valid's binary_logloss: 0.652005
[200]	train's binary_logloss: 0.525344	valid's binary_logloss: 0.654096
Early stopping, best iteration is:
[137]	train's binary_logloss: 0.557437	valid's binary_logloss: 0.651283
feature_fraction, val_score: 0.651017: 100%|##########| 7/7 [00:04<00:00,  1.76it/s][I 2020-09-27 04:43:11,237] Trial 6 finished with value: 0.6512826266964433 and parameters: {'feature_fraction': 0.5}. Best is trial 5 with value: 0.6510174434829193.
feature_fraction, val_score: 0.651017: 100%|##########| 7/7 [00:04<00:00,  1.65it/s]
num_leaves, val_score: 0.651017:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000858 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.519602	valid's binary_logloss: 0.657482
Early stopping, best iteration is:
[48]	train's binary_logloss: 0.578782	valid's binary_logloss: 0.655742
num_leaves, val_score: 0.651017:   5%|5         | 1/20 [00:01<00:23,  1.23s/it][I 2020-09-27 04:43:12,484] Trial 7 finished with value: 0.655741889936831 and parameters: {'num_leaves': 55}. Best is trial 7 with value: 0.655741889936831.
num_leaves, val_score: 0.651017:   5%|5         | 1/20 [00:01<00:23,  1.23s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000474 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.352805	valid's binary_logloss: 0.663625
Early stopping, best iteration is:
[40]	train's binary_logloss: 0.492845	valid's binary_logloss: 0.6534
num_leaves, val_score: 0.651017:  10%|#         | 2/20 [00:02<00:21,  1.18s/it][I 2020-09-27 04:43:13,542] Trial 8 finished with value: 0.653399633240566 and parameters: {'num_leaves': 164}. Best is trial 8 with value: 0.653399633240566.
num_leaves, val_score: 0.651017:  10%|#         | 2/20 [00:02<00:21,  1.18s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005116 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.645865	valid's binary_logloss: 0.655865
[200]	train's binary_logloss: 0.633609	valid's binary_logloss: 0.652415
[300]	train's binary_logloss: 0.624725	valid's binary_logloss: 0.651323
Early stopping, best iteration is:
[292]	train's binary_logloss: 0.625397	valid's binary_logloss: 0.65122
num_leaves, val_score: 0.651017:  15%|#5        | 3/20 [00:02<00:17,  1.02s/it][I 2020-09-27 04:43:14,213] Trial 9 finished with value: 0.6512203104712043 and parameters: {'num_leaves': 5}. Best is trial 9 with value: 0.6512203104712043.
num_leaves, val_score: 0.651017:  15%|#5        | 3/20 [00:02<00:17,  1.02s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000891 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.297872	valid's binary_logloss: 0.671789
Early stopping, best iteration is:
[27]	train's binary_logloss: 0.510065	valid's binary_logloss: 0.658243
num_leaves, val_score: 0.651017:  20%|##        | 4/20 [00:04<00:20,  1.30s/it][I 2020-09-27 04:43:16,157] Trial 10 finished with value: 0.6582425249764476 and parameters: {'num_leaves': 215}. Best is trial 9 with value: 0.6512203104712043.
num_leaves, val_score: 0.651017:  20%|##        | 4/20 [00:04<00:20,  1.30s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000907 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.669067	valid's binary_logloss: 0.671099
[200]	train's binary_logloss: 0.659436	valid's binary_logloss: 0.661559
[300]	train's binary_logloss: 0.654191	valid's binary_logloss: 0.656836
[400]	train's binary_logloss: 0.651007	valid's binary_logloss: 0.65419
[500]	train's binary_logloss: 0.64896	valid's binary_logloss: 0.652841
[600]	train's binary_logloss: 0.647587	valid's binary_logloss: 0.652048
[700]	train's binary_logloss: 0.646612	valid's binary_logloss: 0.651613
[800]	train's binary_logloss: 0.645876	valid's binary_logloss: 0.651451
[900]	train's binary_logloss: 0.645296	valid's binary_logloss: 0.651421
Early stopping, best iteration is:
[887]	train's binary_logloss: 0.645365	valid's binary_logloss: 0.651313
num_leaves, val_score: 0.651017:  25%|##5       | 5/20 [00:06<00:19,  1.29s/it][I 2020-09-27 04:43:17,415] Trial 11 finished with value: 0.6513127631312726 and parameters: {'num_leaves': 2}. Best is trial 9 with value: 0.6512203104712043.
num_leaves, val_score: 0.651017:  25%|##5       | 5/20 [00:06<00:19,  1.29s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001145 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.487358	valid's binary_logloss: 0.659662
Early stopping, best iteration is:
[35]	train's binary_logloss: 0.581991	valid's binary_logloss: 0.657768
num_leaves, val_score: 0.651017:  30%|###       | 6/20 [00:06<00:15,  1.10s/it][I 2020-09-27 04:43:18,063] Trial 12 finished with value: 0.6577682570812166 and parameters: {'num_leaves': 73}. Best is trial 9 with value: 0.6512203104712043.
num_leaves, val_score: 0.651017:  30%|###       | 6/20 [00:06<00:15,  1.10s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001036 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607358	valid's binary_logloss: 0.650129
[200]	train's binary_logloss: 0.572939	valid's binary_logloss: 0.650833
Early stopping, best iteration is:
[137]	train's binary_logloss: 0.593957	valid's binary_logloss: 0.649769
num_leaves, val_score: 0.649769:  35%|###5      | 7/20 [00:07<00:12,  1.07it/s][I 2020-09-27 04:43:18,630] Trial 13 finished with value: 0.649768864866138 and parameters: {'num_leaves': 17}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769:  35%|###5      | 7/20 [00:07<00:12,  1.07it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001092 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.41967	valid's binary_logloss: 0.663995
Early stopping, best iteration is:
[33]	train's binary_logloss: 0.552391	valid's binary_logloss: 0.657367
num_leaves, val_score: 0.649769:  40%|####      | 8/20 [00:08<00:12,  1.07s/it][I 2020-09-27 04:43:20,007] Trial 14 finished with value: 0.6573672765256382 and parameters: {'num_leaves': 114}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769:  40%|####      | 8/20 [00:08<00:12,  1.07s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000964 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.262043	valid's binary_logloss: 0.668022
Early stopping, best iteration is:
[31]	train's binary_logloss: 0.470728	valid's binary_logloss: 0.656929
num_leaves, val_score: 0.649769:  45%|####5     | 9/20 [00:10<00:13,  1.26s/it][I 2020-09-27 04:43:21,707] Trial 15 finished with value: 0.6569287641382702 and parameters: {'num_leaves': 256}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769:  45%|####5     | 9/20 [00:10<00:13,  1.26s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000921 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.534752	valid's binary_logloss: 0.65496
Early stopping, best iteration is:
[63]	train's binary_logloss: 0.569855	valid's binary_logloss: 0.652951
num_leaves, val_score: 0.649769:  50%|#####     | 10/20 [00:11<00:10,  1.07s/it][I 2020-09-27 04:43:22,329] Trial 16 finished with value: 0.6529514815127009 and parameters: {'num_leaves': 48}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769:  50%|#####     | 10/20 [00:11<00:10,  1.07s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004780 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.382444	valid's binary_logloss: 0.663277
Early stopping, best iteration is:
[36]	train's binary_logloss: 0.523542	valid's binary_logloss: 0.654707
num_leaves, val_score: 0.649769:  55%|#####5    | 11/20 [00:12<00:10,  1.12s/it][I 2020-09-27 04:43:23,589] Trial 17 finished with value: 0.6547072669652428 and parameters: {'num_leaves': 140}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769:  55%|#####5    | 11/20 [00:12<00:10,  1.12s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.003413 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.448357	valid's binary_logloss: 0.661818
Early stopping, best iteration is:
[46]	train's binary_logloss: 0.538243	valid's binary_logloss: 0.657096
num_leaves, val_score: 0.649769:  60%|######    | 12/20 [00:13<00:08,  1.08s/it][I 2020-09-27 04:43:24,567] Trial 18 finished with value: 0.6570959414584372 and parameters: {'num_leaves': 95}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769:  60%|######    | 12/20 [00:13<00:08,  1.08s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010468 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.589787	valid's binary_logloss: 0.652218
Early stopping, best iteration is:
[97]	train's binary_logloss: 0.591353	valid's binary_logloss: 0.652064
num_leaves, val_score: 0.649769:  65%|######5   | 13/20 [00:13<00:06,  1.09it/s][I 2020-09-27 04:43:25,099] Trial 19 finished with value: 0.6520641906266819 and parameters: {'num_leaves': 24}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769:  65%|######5   | 13/20 [00:13<00:06,  1.09it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000894 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.332931	valid's binary_logloss: 0.669943
Early stopping, best iteration is:
[24]	train's binary_logloss: 0.541056	valid's binary_logloss: 0.658153
num_leaves, val_score: 0.649769:  70%|#######   | 14/20 [00:14<00:05,  1.02it/s][I 2020-09-27 04:43:26,213] Trial 20 finished with value: 0.6581527972889681 and parameters: {'num_leaves': 181}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769:  70%|#######   | 14/20 [00:14<00:05,  1.02it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010464 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.645865	valid's binary_logloss: 0.655865
[200]	train's binary_logloss: 0.633609	valid's binary_logloss: 0.652415
[300]	train's binary_logloss: 0.624725	valid's binary_logloss: 0.651323
Early stopping, best iteration is:
[292]	train's binary_logloss: 0.625397	valid's binary_logloss: 0.65122
num_leaves, val_score: 0.649769:  75%|#######5  | 15/20 [00:15<00:04,  1.13it/s][I 2020-09-27 04:43:26,886] Trial 21 finished with value: 0.6512203104712043 and parameters: {'num_leaves': 5}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769:  75%|#######5  | 15/20 [00:15<00:04,  1.13it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000872 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579189	valid's binary_logloss: 0.654451
[200]	train's binary_logloss: 0.527779	valid's binary_logloss: 0.653698
Early stopping, best iteration is:
[151]	train's binary_logloss: 0.551963	valid's binary_logloss: 0.653462
num_leaves, val_score: 0.649769:  80%|########  | 16/20 [00:16<00:04,  1.00s/it][I 2020-09-27 04:43:28,167] Trial 22 finished with value: 0.6534616704278683 and parameters: {'num_leaves': 28}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769:  80%|########  | 16/20 [00:16<00:04,  1.00s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.011614 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.472661	valid's binary_logloss: 0.658505
Early stopping, best iteration is:
[28]	train's binary_logloss: 0.589272	valid's binary_logloss: 0.658288
num_leaves, val_score: 0.649769:  85%|########5 | 17/20 [00:17<00:02,  1.13it/s][I 2020-09-27 04:43:28,784] Trial 23 finished with value: 0.658288148145066 and parameters: {'num_leaves': 81}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769:  85%|########5 | 17/20 [00:17<00:02,  1.13it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000925 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.645865	valid's binary_logloss: 0.655865
[200]	train's binary_logloss: 0.633609	valid's binary_logloss: 0.652415
[300]	train's binary_logloss: 0.624725	valid's binary_logloss: 0.651323
Early stopping, best iteration is:
[292]	train's binary_logloss: 0.625397	valid's binary_logloss: 0.65122
num_leaves, val_score: 0.649769:  90%|######### | 18/20 [00:18<00:01,  1.23it/s][I 2020-09-27 04:43:29,433] Trial 24 finished with value: 0.6512203104712043 and parameters: {'num_leaves': 5}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769:  90%|######### | 18/20 [00:18<00:01,  1.23it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004987 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.565987	valid's binary_logloss: 0.655697
Early stopping, best iteration is:
[59]	train's binary_logloss: 0.596142	valid's binary_logloss: 0.65418
num_leaves, val_score: 0.649769:  95%|#########5| 19/20 [00:18<00:00,  1.34it/s][I 2020-09-27 04:43:30,009] Trial 25 finished with value: 0.6541801455639907 and parameters: {'num_leaves': 34}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769:  95%|#########5| 19/20 [00:18<00:00,  1.34it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010666 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.512496	valid's binary_logloss: 0.657066
Early stopping, best iteration is:
[78]	train's binary_logloss: 0.536548	valid's binary_logloss: 0.654935
num_leaves, val_score: 0.649769: 100%|##########| 20/20 [00:19<00:00,  1.38it/s][I 2020-09-27 04:43:30,691] Trial 26 finished with value: 0.6549352289587036 and parameters: {'num_leaves': 59}. Best is trial 13 with value: 0.649768864866138.
num_leaves, val_score: 0.649769: 100%|##########| 20/20 [00:19<00:00,  1.03it/s]
bagging, val_score: 0.649769:   0%|          | 0/10 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000913 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.612583	valid's binary_logloss: 0.656379
Early stopping, best iteration is:
[86]	train's binary_logloss: 0.617678	valid's binary_logloss: 0.655144
bagging, val_score: 0.649769:  10%|#         | 1/10 [00:00<00:07,  1.17it/s][I 2020-09-27 04:43:31,563] Trial 27 finished with value: 0.6551437821744738 and parameters: {'bagging_fraction': 0.40320944016041527, 'bagging_freq': 7}. Best is trial 27 with value: 0.6551437821744738.
bagging, val_score: 0.649769:  10%|#         | 1/10 [00:00<00:07,  1.17it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.014340 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607438	valid's binary_logloss: 0.653472
[200]	train's binary_logloss: 0.573071	valid's binary_logloss: 0.653346
Early stopping, best iteration is:
[139]	train's binary_logloss: 0.593171	valid's binary_logloss: 0.652081
bagging, val_score: 0.649769:  20%|##        | 2/10 [00:01<00:06,  1.21it/s][I 2020-09-27 04:43:32,320] Trial 28 finished with value: 0.6520811290472097 and parameters: {'bagging_fraction': 0.9561159020668242, 'bagging_freq': 1}. Best is trial 28 with value: 0.6520811290472097.
bagging, val_score: 0.649769:  20%|##        | 2/10 [00:01<00:06,  1.21it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001122 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.610187	valid's binary_logloss: 0.652923
Early stopping, best iteration is:
[81]	train's binary_logloss: 0.617563	valid's binary_logloss: 0.651804
bagging, val_score: 0.649769:  30%|###       | 3/10 [00:02<00:05,  1.39it/s][I 2020-09-27 04:43:32,796] Trial 29 finished with value: 0.6518044921241909 and parameters: {'bagging_fraction': 0.5465375900437621, 'bagging_freq': 4}. Best is trial 29 with value: 0.6518044921241909.
bagging, val_score: 0.649769:  30%|###       | 3/10 [00:02<00:05,  1.39it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.002328 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607479	valid's binary_logloss: 0.651182
[200]	train's binary_logloss: 0.571824	valid's binary_logloss: 0.653082
Early stopping, best iteration is:
[118]	train's binary_logloss: 0.600602	valid's binary_logloss: 0.650679
bagging, val_score: 0.649769:  40%|####      | 4/10 [00:02<00:04,  1.46it/s][I 2020-09-27 04:43:33,403] Trial 30 finished with value: 0.6506791175850941 and parameters: {'bagging_fraction': 0.9711225430872865, 'bagging_freq': 7}. Best is trial 30 with value: 0.6506791175850941.
bagging, val_score: 0.649769:  40%|####      | 4/10 [00:02<00:04,  1.46it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005891 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607508	valid's binary_logloss: 0.652005
[200]	train's binary_logloss: 0.572706	valid's binary_logloss: 0.650621
Early stopping, best iteration is:
[129]	train's binary_logloss: 0.596726	valid's binary_logloss: 0.650382
bagging, val_score: 0.649769:  50%|#####     | 5/10 [00:03<00:03,  1.48it/s][I 2020-09-27 04:43:34,052] Trial 31 finished with value: 0.6503824259794613 and parameters: {'bagging_fraction': 0.9985128973254092, 'bagging_freq': 7}. Best is trial 31 with value: 0.6503824259794613.
bagging, val_score: 0.649769:  50%|#####     | 5/10 [00:03<00:03,  1.48it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005483 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.60784	valid's binary_logloss: 0.652715
[200]	train's binary_logloss: 0.572748	valid's binary_logloss: 0.653478
Early stopping, best iteration is:
[124]	train's binary_logloss: 0.598892	valid's binary_logloss: 0.651677
bagging, val_score: 0.649769:  60%|######    | 6/10 [00:04<00:02,  1.48it/s][I 2020-09-27 04:43:34,731] Trial 32 finished with value: 0.6516771078927124 and parameters: {'bagging_fraction': 0.9937517875557063, 'bagging_freq': 7}. Best is trial 31 with value: 0.6503824259794613.
bagging, val_score: 0.649769:  60%|######    | 6/10 [00:04<00:02,  1.48it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.003559 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607577	valid's binary_logloss: 0.652062
[200]	train's binary_logloss: 0.572523	valid's binary_logloss: 0.651285
[300]	train's binary_logloss: 0.541985	valid's binary_logloss: 0.651463
Early stopping, best iteration is:
[260]	train's binary_logloss: 0.554233	valid's binary_logloss: 0.65052
bagging, val_score: 0.649769:  70%|#######   | 7/10 [00:05<00:02,  1.13it/s][I 2020-09-27 04:43:36,093] Trial 33 finished with value: 0.6505203576283998 and parameters: {'bagging_fraction': 0.8468420031688184, 'bagging_freq': 6}. Best is trial 31 with value: 0.6503824259794613.
bagging, val_score: 0.649769:  70%|#######   | 7/10 [00:05<00:02,  1.13it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001043 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.60809	valid's binary_logloss: 0.653721
Early stopping, best iteration is:
[83]	train's binary_logloss: 0.614817	valid's binary_logloss: 0.653287
bagging, val_score: 0.649769:  80%|########  | 8/10 [00:05<00:01,  1.30it/s][I 2020-09-27 04:43:36,593] Trial 34 finished with value: 0.6532867425106659 and parameters: {'bagging_fraction': 0.8464729466696903, 'bagging_freq': 7}. Best is trial 31 with value: 0.6503824259794613.
bagging, val_score: 0.649769:  80%|########  | 8/10 [00:05<00:01,  1.30it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000488 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.608096	valid's binary_logloss: 0.652476
[200]	train's binary_logloss: 0.572069	valid's binary_logloss: 0.654848
Early stopping, best iteration is:
[105]	train's binary_logloss: 0.606068	valid's binary_logloss: 0.652367
bagging, val_score: 0.649769:  90%|######### | 9/10 [00:06<00:00,  1.41it/s][I 2020-09-27 04:43:37,161] Trial 35 finished with value: 0.6523668340051799 and parameters: {'bagging_fraction': 0.830591192364971, 'bagging_freq': 5}. Best is trial 31 with value: 0.6503824259794613.
bagging, val_score: 0.649769:  90%|######### | 9/10 [00:06<00:00,  1.41it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000998 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607246	valid's binary_logloss: 0.652045
Early stopping, best iteration is:
[95]	train's binary_logloss: 0.609255	valid's binary_logloss: 0.651848
bagging, val_score: 0.649769: 100%|##########| 10/10 [00:07<00:00,  1.50it/s][I 2020-09-27 04:43:37,726] Trial 36 finished with value: 0.6518478119160271 and parameters: {'bagging_fraction': 0.9992112947235333, 'bagging_freq': 6}. Best is trial 31 with value: 0.6503824259794613.
bagging, val_score: 0.649769: 100%|##########| 10/10 [00:07<00:00,  1.42it/s]
feature_fraction_stage2, val_score: 0.649769:   0%|          | 0/6 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000951 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.606705	valid's binary_logloss: 0.652233
[200]	train's binary_logloss: 0.572023	valid's binary_logloss: 0.650662
[300]	train's binary_logloss: 0.542509	valid's binary_logloss: 0.653346
Early stopping, best iteration is:
[200]	train's binary_logloss: 0.572023	valid's binary_logloss: 0.650662
feature_fraction_stage2, val_score: 0.649769:  17%|#6        | 1/6 [00:00<00:03,  1.42it/s][I 2020-09-27 04:43:38,443] Trial 37 finished with value: 0.6506622068727096 and parameters: {'feature_fraction': 0.9799999999999999}. Best is trial 37 with value: 0.6506622068727096.
feature_fraction_stage2, val_score: 0.649769:  17%|#6        | 1/6 [00:00<00:03,  1.42it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000943 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607972	valid's binary_logloss: 0.651443
[200]	train's binary_logloss: 0.573729	valid's binary_logloss: 0.651681
Early stopping, best iteration is:
[174]	train's binary_logloss: 0.58195	valid's binary_logloss: 0.650991
feature_fraction_stage2, val_score: 0.649769:  33%|###3      | 2/6 [00:01<00:03,  1.17it/s][I 2020-09-27 04:43:39,648] Trial 38 finished with value: 0.6509908340902059 and parameters: {'feature_fraction': 0.852}. Best is trial 37 with value: 0.6506622068727096.
feature_fraction_stage2, val_score: 0.649769:  33%|###3      | 2/6 [00:01<00:03,  1.17it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000929 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.606705	valid's binary_logloss: 0.652233
[200]	train's binary_logloss: 0.572023	valid's binary_logloss: 0.650662
[300]	train's binary_logloss: 0.542509	valid's binary_logloss: 0.653346
Early stopping, best iteration is:
[200]	train's binary_logloss: 0.572023	valid's binary_logloss: 0.650662
feature_fraction_stage2, val_score: 0.649769:  50%|#####     | 3/6 [00:02<00:02,  1.22it/s][I 2020-09-27 04:43:40,396] Trial 39 finished with value: 0.6506622068727096 and parameters: {'feature_fraction': 0.948}. Best is trial 37 with value: 0.6506622068727096.
feature_fraction_stage2, val_score: 0.649769:  50%|#####     | 3/6 [00:02<00:02,  1.22it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000871 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.60867	valid's binary_logloss: 0.651571
Early stopping, best iteration is:
[95]	train's binary_logloss: 0.610608	valid's binary_logloss: 0.651188
feature_fraction_stage2, val_score: 0.649769:  67%|######6   | 4/6 [00:03<00:01,  1.37it/s][I 2020-09-27 04:43:40,904] Trial 40 finished with value: 0.6511879268615787 and parameters: {'feature_fraction': 0.82}. Best is trial 37 with value: 0.6506622068727096.
feature_fraction_stage2, val_score: 0.649769:  67%|######6   | 4/6 [00:03<00:01,  1.37it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005317 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.60735	valid's binary_logloss: 0.652555
Early stopping, best iteration is:
[72]	train's binary_logloss: 0.618944	valid's binary_logloss: 0.651968
feature_fraction_stage2, val_score: 0.649769:  83%|########3 | 5/6 [00:03<00:00,  1.53it/s][I 2020-09-27 04:43:41,380] Trial 41 finished with value: 0.6519681937770695 and parameters: {'feature_fraction': 0.9159999999999999}. Best is trial 37 with value: 0.6506622068727096.
feature_fraction_stage2, val_score: 0.649769:  83%|########3 | 5/6 [00:03<00:00,  1.53it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005114 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607358	valid's binary_logloss: 0.650129
[200]	train's binary_logloss: 0.572939	valid's binary_logloss: 0.650833
Early stopping, best iteration is:
[137]	train's binary_logloss: 0.593957	valid's binary_logloss: 0.649769
feature_fraction_stage2, val_score: 0.649769: 100%|##########| 6/6 [00:04<00:00,  1.54it/s][I 2020-09-27 04:43:42,023] Trial 42 finished with value: 0.649768864866138 and parameters: {'feature_fraction': 0.8839999999999999}. Best is trial 42 with value: 0.649768864866138.
feature_fraction_stage2, val_score: 0.649769: 100%|##########| 6/6 [00:04<00:00,  1.40it/s]
regularization_factors, val_score: 0.649769:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000939 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607973	valid's binary_logloss: 0.65115
[200]	train's binary_logloss: 0.573107	valid's binary_logloss: 0.649872
[300]	train's binary_logloss: 0.543329	valid's binary_logloss: 0.651302
Early stopping, best iteration is:
[229]	train's binary_logloss: 0.564107	valid's binary_logloss: 0.649364
regularization_factors, val_score: 0.649364:   5%|5         | 1/20 [00:01<00:24,  1.27s/it][I 2020-09-27 04:43:43,315] Trial 43 finished with value: 0.6493642189216182 and parameters: {'lambda_l1': 0.16783910794593826, 'lambda_l2': 6.907642804430548e-07}. Best is trial 43 with value: 0.6493642189216182.
regularization_factors, val_score: 0.649364:   5%|5         | 1/20 [00:01<00:24,  1.27s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000994 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.608181	valid's binary_logloss: 0.65244
[200]	train's binary_logloss: 0.574199	valid's binary_logloss: 0.652483
Early stopping, best iteration is:
[119]	train's binary_logloss: 0.601088	valid's binary_logloss: 0.651426
regularization_factors, val_score: 0.649364:  10%|#         | 2/20 [00:01<00:19,  1.08s/it][I 2020-09-27 04:43:43,949] Trial 44 finished with value: 0.6514259112904369 and parameters: {'lambda_l1': 0.3129099140173397, 'lambda_l2': 3.3514640861329857e-07}. Best is trial 43 with value: 0.6493642189216182.
regularization_factors, val_score: 0.649364:  10%|#         | 2/20 [00:01<00:19,  1.08s/it][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.009984 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607688	valid's binary_logloss: 0.651501
[200]	train's binary_logloss: 0.57304	valid's binary_logloss: 0.648731
[300]	train's binary_logloss: 0.543627	valid's binary_logloss: 0.650347
Early stopping, best iteration is:
[212]	train's binary_logloss: 0.569524	valid's binary_logloss: 0.64867
regularization_factors, val_score: 0.648670:  15%|#5        | 3/20 [00:02<00:16,  1.01it/s][I 2020-09-27 04:43:44,720] Trial 45 finished with value: 0.6486701923102753 and parameters: {'lambda_l1': 7.774361825486131e-06, 'lambda_l2': 0.006013598728332988}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  15%|#5        | 3/20 [00:02<00:16,  1.01it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000855 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607956	valid's binary_logloss: 0.650205
[200]	train's binary_logloss: 0.573144	valid's binary_logloss: 0.651349
Early stopping, best iteration is:
[109]	train's binary_logloss: 0.604467	valid's binary_logloss: 0.649709
regularization_factors, val_score: 0.648670:  20%|##        | 4/20 [00:03<00:13,  1.17it/s][I 2020-09-27 04:43:45,259] Trial 46 finished with value: 0.6497086221332595 and parameters: {'lambda_l1': 4.0595395206550105e-07, 'lambda_l2': 0.16859061913783596}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  20%|##        | 4/20 [00:03<00:13,  1.17it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000898 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607896	valid's binary_logloss: 0.649423
[200]	train's binary_logloss: 0.574175	valid's binary_logloss: 0.64969
Early stopping, best iteration is:
[121]	train's binary_logloss: 0.600172	valid's binary_logloss: 0.648913
regularization_factors, val_score: 0.648670:  25%|##5       | 5/20 [00:03<00:11,  1.29it/s][I 2020-09-27 04:43:45,841] Trial 47 finished with value: 0.6489130270973635 and parameters: {'lambda_l1': 1.9089796516560767e-07, 'lambda_l2': 0.2199091551428564}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  25%|##5       | 5/20 [00:03<00:11,  1.29it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000906 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607889	valid's binary_logloss: 0.650079
[200]	train's binary_logloss: 0.574212	valid's binary_logloss: 0.65031
Early stopping, best iteration is:
[131]	train's binary_logloss: 0.59635	valid's binary_logloss: 0.648762
regularization_factors, val_score: 0.648670:  30%|###       | 6/20 [00:04<00:11,  1.23it/s][I 2020-09-27 04:43:46,751] Trial 48 finished with value: 0.6487624267173515 and parameters: {'lambda_l1': 4.2319368851060655e-07, 'lambda_l2': 0.34004310864421733}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  30%|###       | 6/20 [00:04<00:11,  1.23it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.014681 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.608275	valid's binary_logloss: 0.650944
[200]	train's binary_logloss: 0.573416	valid's binary_logloss: 0.651579
Early stopping, best iteration is:
[113]	train's binary_logloss: 0.603189	valid's binary_logloss: 0.650489
regularization_factors, val_score: 0.648670:  35%|###5      | 7/20 [00:05<00:10,  1.22it/s][I 2020-09-27 04:43:47,591] Trial 49 finished with value: 0.6504888821768885 and parameters: {'lambda_l1': 1.875716331797824e-07, 'lambda_l2': 0.3304208617133868}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  35%|###5      | 7/20 [00:05<00:10,  1.22it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001188 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.608185	valid's binary_logloss: 0.650237
Early stopping, best iteration is:
[98]	train's binary_logloss: 0.608923	valid's binary_logloss: 0.650108
regularization_factors, val_score: 0.648670:  40%|####      | 8/20 [00:06<00:08,  1.35it/s][I 2020-09-27 04:43:48,143] Trial 50 finished with value: 0.6501084626471252 and parameters: {'lambda_l1': 4.671913690463379e-07, 'lambda_l2': 0.1995460595674108}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  40%|####      | 8/20 [00:06<00:08,  1.35it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000823 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.60758	valid's binary_logloss: 0.651374
[200]	train's binary_logloss: 0.57306	valid's binary_logloss: 0.651712
Early stopping, best iteration is:
[154]	train's binary_logloss: 0.588205	valid's binary_logloss: 0.650518
regularization_factors, val_score: 0.648670:  45%|####5     | 9/20 [00:06<00:07,  1.39it/s][I 2020-09-27 04:43:48,805] Trial 51 finished with value: 0.6505179255791086 and parameters: {'lambda_l1': 7.519476866197216e-07, 'lambda_l2': 0.08987068172093761}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  45%|####5     | 9/20 [00:06<00:07,  1.39it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000917 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607423	valid's binary_logloss: 0.652085
[200]	train's binary_logloss: 0.573345	valid's binary_logloss: 0.651502
Early stopping, best iteration is:
[183]	train's binary_logloss: 0.578885	valid's binary_logloss: 0.651003
regularization_factors, val_score: 0.648670:  50%|#####     | 10/20 [00:07<00:07,  1.40it/s][I 2020-09-27 04:43:49,511] Trial 52 finished with value: 0.6510031934822363 and parameters: {'lambda_l1': 7.307330460252661e-07, 'lambda_l2': 0.0186645031244208}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  50%|#####     | 10/20 [00:07<00:07,  1.40it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.002545 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607358	valid's binary_logloss: 0.650129
[200]	train's binary_logloss: 0.572941	valid's binary_logloss: 0.650864
Early stopping, best iteration is:
[137]	train's binary_logloss: 0.593957	valid's binary_logloss: 0.649769
regularization_factors, val_score: 0.648670:  55%|#####5    | 11/20 [00:08<00:06,  1.46it/s][I 2020-09-27 04:43:50,135] Trial 53 finished with value: 0.6497687989983096 and parameters: {'lambda_l1': 0.0002292384736520708, 'lambda_l2': 8.689375650085261e-05}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  55%|#####5    | 11/20 [00:08<00:06,  1.46it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000469 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607679	valid's binary_logloss: 0.651246
[200]	train's binary_logloss: 0.57339	valid's binary_logloss: 0.653045
Early stopping, best iteration is:
[127]	train's binary_logloss: 0.597662	valid's binary_logloss: 0.650697
regularization_factors, val_score: 0.648670:  60%|######    | 12/20 [00:09<00:06,  1.23it/s][I 2020-09-27 04:43:51,237] Trial 54 finished with value: 0.6506972734005597 and parameters: {'lambda_l1': 0.001430301077955183, 'lambda_l2': 5.329476633929677e-06}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  60%|######    | 12/20 [00:09<00:06,  1.23it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000893 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607359	valid's binary_logloss: 0.650129
[200]	train's binary_logloss: 0.572942	valid's binary_logloss: 0.650833
Early stopping, best iteration is:
[137]	train's binary_logloss: 0.593959	valid's binary_logloss: 0.649769
regularization_factors, val_score: 0.648670:  65%|######5   | 13/20 [00:09<00:05,  1.31it/s][I 2020-09-27 04:43:51,881] Trial 55 finished with value: 0.6497686874558546 and parameters: {'lambda_l1': 0.00011580371818856396, 'lambda_l2': 0.0005902239811498276}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  65%|######5   | 13/20 [00:09<00:05,  1.31it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.007667 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607682	valid's binary_logloss: 0.651246
[200]	train's binary_logloss: 0.57384	valid's binary_logloss: 0.650954
[300]	train's binary_logloss: 0.543099	valid's binary_logloss: 0.650065
Early stopping, best iteration is:
[273]	train's binary_logloss: 0.551078	valid's binary_logloss: 0.649674
regularization_factors, val_score: 0.648670:  70%|#######   | 14/20 [00:10<00:04,  1.25it/s][I 2020-09-27 04:43:52,762] Trial 56 finished with value: 0.649674471194536 and parameters: {'lambda_l1': 3.268557145355658e-05, 'lambda_l2': 0.003069238534147287}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  70%|#######   | 14/20 [00:10<00:04,  1.25it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000897 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.612768	valid's binary_logloss: 0.651575
[200]	train's binary_logloss: 0.586233	valid's binary_logloss: 0.651305
Early stopping, best iteration is:
[160]	train's binary_logloss: 0.595921	valid's binary_logloss: 0.650515
regularization_factors, val_score: 0.648670:  75%|#######5  | 15/20 [00:11<00:03,  1.31it/s][I 2020-09-27 04:43:53,439] Trial 57 finished with value: 0.6505145838820116 and parameters: {'lambda_l1': 5.615730195745978e-06, 'lambda_l2': 7.365468833430753}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  75%|#######5  | 15/20 [00:11<00:03,  1.31it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.007481 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.60768	valid's binary_logloss: 0.651246
[200]	train's binary_logloss: 0.573394	valid's binary_logloss: 0.653045
Early stopping, best iteration is:
[127]	train's binary_logloss: 0.597664	valid's binary_logloss: 0.650697
regularization_factors, val_score: 0.648670:  80%|########  | 16/20 [00:12<00:02,  1.38it/s][I 2020-09-27 04:43:54,076] Trial 58 finished with value: 0.6506971138019576 and parameters: {'lambda_l1': 2.3099984324631693e-08, 'lambda_l2': 0.0023773905893283293}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  80%|########  | 16/20 [00:12<00:02,  1.38it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001747 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.61046	valid's binary_logloss: 0.65107
[200]	train's binary_logloss: 0.581203	valid's binary_logloss: 0.651453
Early stopping, best iteration is:
[171]	train's binary_logloss: 0.588919	valid's binary_logloss: 0.650336
regularization_factors, val_score: 0.648670:  85%|########5 | 17/20 [00:13<00:02,  1.15it/s][I 2020-09-27 04:43:55,278] Trial 59 finished with value: 0.6503360249834589 and parameters: {'lambda_l1': 1.2365048737402212e-05, 'lambda_l2': 3.1664647322579405}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  85%|########5 | 17/20 [00:13<00:02,  1.15it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000874 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607694	valid's binary_logloss: 0.6515
[200]	train's binary_logloss: 0.573475	valid's binary_logloss: 0.651154
Early stopping, best iteration is:
[133]	train's binary_logloss: 0.595863	valid's binary_logloss: 0.650792
regularization_factors, val_score: 0.648670:  90%|######### | 18/20 [00:13<00:01,  1.27it/s][I 2020-09-27 04:43:55,883] Trial 60 finished with value: 0.6507917448366202 and parameters: {'lambda_l1': 1.4432590771728913e-08, 'lambda_l2': 0.01029217837602668}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  90%|######### | 18/20 [00:13<00:01,  1.27it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000902 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607359	valid's binary_logloss: 0.650129
[200]	train's binary_logloss: 0.572942	valid's binary_logloss: 0.650864
Early stopping, best iteration is:
[137]	train's binary_logloss: 0.593958	valid's binary_logloss: 0.649769
regularization_factors, val_score: 0.648670:  95%|#########5| 19/20 [00:14<00:00,  1.34it/s][I 2020-09-27 04:43:56,523] Trial 61 finished with value: 0.6497687138706673 and parameters: {'lambda_l1': 0.00010150982895717758, 'lambda_l2': 0.0005002352741242428}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670:  95%|#########5| 19/20 [00:14<00:00,  1.34it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000964 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607358	valid's binary_logloss: 0.650129
[200]	train's binary_logloss: 0.57294	valid's binary_logloss: 0.650849
Early stopping, best iteration is:
[137]	train's binary_logloss: 0.593957	valid's binary_logloss: 0.649769
regularization_factors, val_score: 0.648670: 100%|##########| 20/20 [00:15<00:00,  1.42it/s][I 2020-09-27 04:43:57,130] Trial 62 finished with value: 0.6497688040658035 and parameters: {'lambda_l1': 1.4113098720716957e-05, 'lambda_l2': 0.00022029341576689507}. Best is trial 45 with value: 0.6486701923102753.
regularization_factors, val_score: 0.648670: 100%|##########| 20/20 [00:15<00:00,  1.32it/s]
min_data_in_leaf, val_score: 0.648670:   0%|          | 0/5 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000908 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607864	valid's binary_logloss: 0.650842
[200]	train's binary_logloss: 0.573843	valid's binary_logloss: 0.649839
[300]	train's binary_logloss: 0.544181	valid's binary_logloss: 0.65183
Early stopping, best iteration is:
[206]	train's binary_logloss: 0.572189	valid's binary_logloss: 0.649438
min_data_in_leaf, val_score: 0.648670:  20%|##        | 1/5 [00:00<00:03,  1.26it/s][I 2020-09-27 04:43:57,935] Trial 63 finished with value: 0.6494379643673076 and parameters: {'min_child_samples': 25}. Best is trial 63 with value: 0.6494379643673076.
min_data_in_leaf, val_score: 0.648670:  20%|##        | 1/5 [00:00<00:03,  1.26it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012759 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.610343	valid's binary_logloss: 0.652543
Early stopping, best iteration is:
[69]	train's binary_logloss: 0.622017	valid's binary_logloss: 0.652141
min_data_in_leaf, val_score: 0.648670:  40%|####      | 2/5 [00:01<00:02,  1.13it/s][I 2020-09-27 04:43:59,052] Trial 64 finished with value: 0.6521414167861022 and parameters: {'min_child_samples': 100}. Best is trial 63 with value: 0.6494379643673076.
min_data_in_leaf, val_score: 0.648670:  40%|####      | 2/5 [00:01<00:02,  1.13it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000904 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.60911	valid's binary_logloss: 0.649506
[200]	train's binary_logloss: 0.576093	valid's binary_logloss: 0.651061
Early stopping, best iteration is:
[123]	train's binary_logloss: 0.601092	valid's binary_logloss: 0.648901
min_data_in_leaf, val_score: 0.648670:  60%|######    | 3/5 [00:02<00:01,  1.23it/s][I 2020-09-27 04:43:59,681] Trial 65 finished with value: 0.6489009066838253 and parameters: {'min_child_samples': 50}. Best is trial 65 with value: 0.6489009066838253.
min_data_in_leaf, val_score: 0.648670:  60%|######    | 3/5 [00:02<00:01,  1.23it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000900 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607172	valid's binary_logloss: 0.65
[200]	train's binary_logloss: 0.572515	valid's binary_logloss: 0.651594
Early stopping, best iteration is:
[100]	train's binary_logloss: 0.607172	valid's binary_logloss: 0.65
min_data_in_leaf, val_score: 0.648670:  80%|########  | 4/5 [00:03<00:00,  1.33it/s][I 2020-09-27 04:44:00,293] Trial 66 finished with value: 0.6500004234119392 and parameters: {'min_child_samples': 10}. Best is trial 65 with value: 0.6489009066838253.
min_data_in_leaf, val_score: 0.648670:  80%|########  | 4/5 [00:03<00:00,  1.33it/s][LightGBM] [Info] Number of positive: 12813, number of negative: 13186
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000902 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4239
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.492827 -> initscore=-0.028695
[LightGBM] [Info] Start training from score -0.028695
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607283	valid's binary_logloss: 0.652029
[200]	train's binary_logloss: 0.572898	valid's binary_logloss: 0.653554
Early stopping, best iteration is:
[137]	train's binary_logloss: 0.593598	valid's binary_logloss: 0.650833
min_data_in_leaf, val_score: 0.648670: 100%|##########| 5/5 [00:03<00:00,  1.38it/s][I 2020-09-27 04:44:00,964] Trial 67 finished with value: 0.6508328689277415 and parameters: {'min_child_samples': 5}. Best is trial 65 with value: 0.6489009066838253.
min_data_in_leaf, val_score: 0.648670: 100%|##########| 5/5 [00:03<00:00,  1.31it/s]
Fold : 6
[I 2020-09-27 04:44:01,016] A new study created in memory with name: no-name-d7eb95f7-845c-40a3-9583-af27505411ce
feature_fraction, val_score: inf:   0%|          | 0/7 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.003088 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.574382	valid's binary_logloss: 0.658832
Early stopping, best iteration is:
[54]	train's binary_logloss: 0.607296	valid's binary_logloss: 0.657769
feature_fraction, val_score: 0.657769:  14%|#4        | 1/7 [00:00<00:03,  1.69it/s][I 2020-09-27 04:44:01,616] Trial 0 finished with value: 0.6577689712084953 and parameters: {'feature_fraction': 0.7}. Best is trial 0 with value: 0.6577689712084953.
feature_fraction, val_score: 0.657769:  14%|#4        | 1/7 [00:00<00:03,  1.69it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004999 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.573426	valid's binary_logloss: 0.659195
Early stopping, best iteration is:
[52]	train's binary_logloss: 0.607667	valid's binary_logloss: 0.65734
feature_fraction, val_score: 0.657340:  29%|##8       | 2/7 [00:01<00:03,  1.50it/s][I 2020-09-27 04:44:02,465] Trial 1 finished with value: 0.6573399663499826 and parameters: {'feature_fraction': 0.8}. Best is trial 1 with value: 0.6573399663499826.
feature_fraction, val_score: 0.657340:  29%|##8       | 2/7 [00:01<00:03,  1.50it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012514 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.578959	valid's binary_logloss: 0.658522
Early stopping, best iteration is:
[92]	train's binary_logloss: 0.583977	valid's binary_logloss: 0.657645
feature_fraction, val_score: 0.657340:  43%|####2     | 3/7 [00:02<00:02,  1.48it/s][I 2020-09-27 04:44:03,161] Trial 2 finished with value: 0.6576453641626878 and parameters: {'feature_fraction': 0.5}. Best is trial 1 with value: 0.6573399663499826.
feature_fraction, val_score: 0.657340:  43%|####2     | 3/7 [00:02<00:02,  1.48it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000380 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.582749	valid's binary_logloss: 0.655485
[200]	train's binary_logloss: 0.531378	valid's binary_logloss: 0.660147
Early stopping, best iteration is:
[107]	train's binary_logloss: 0.57849	valid's binary_logloss: 0.654569
feature_fraction, val_score: 0.654569:  57%|#####7    | 4/7 [00:03<00:02,  1.04it/s][I 2020-09-27 04:44:04,801] Trial 3 finished with value: 0.654569007042944 and parameters: {'feature_fraction': 0.4}. Best is trial 3 with value: 0.654569007042944.
feature_fraction, val_score: 0.654569:  57%|#####7    | 4/7 [00:03<00:02,  1.04it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000972 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.572042	valid's binary_logloss: 0.658846
Early stopping, best iteration is:
[79]	train's binary_logloss: 0.586097	valid's binary_logloss: 0.65804
feature_fraction, val_score: 0.654569:  71%|#######1  | 5/7 [00:04<00:01,  1.16it/s][I 2020-09-27 04:44:05,420] Trial 4 finished with value: 0.6580401508293939 and parameters: {'feature_fraction': 0.8999999999999999}. Best is trial 3 with value: 0.654569007042944.
feature_fraction, val_score: 0.654569:  71%|#######1  | 5/7 [00:04<00:01,  1.16it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000929 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.571857	valid's binary_logloss: 0.659029
Early stopping, best iteration is:
[44]	train's binary_logloss: 0.612853	valid's binary_logloss: 0.657339
feature_fraction, val_score: 0.654569:  86%|########5 | 6/7 [00:04<00:00,  1.30it/s][I 2020-09-27 04:44:05,979] Trial 5 finished with value: 0.6573389310156069 and parameters: {'feature_fraction': 1.0}. Best is trial 3 with value: 0.654569007042944.
feature_fraction, val_score: 0.654569:  86%|########5 | 6/7 [00:04<00:00,  1.30it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016710 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.577175	valid's binary_logloss: 0.657844
Early stopping, best iteration is:
[54]	train's binary_logloss: 0.608418	valid's binary_logloss: 0.656152
feature_fraction, val_score: 0.654569: 100%|##########| 7/7 [00:05<00:00,  1.24it/s][I 2020-09-27 04:44:06,878] Trial 6 finished with value: 0.6561524021651702 and parameters: {'feature_fraction': 0.6}. Best is trial 3 with value: 0.654569007042944.
feature_fraction, val_score: 0.654569: 100%|##########| 7/7 [00:05<00:00,  1.19it/s]
num_leaves, val_score: 0.654569:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004746 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.384415	valid's binary_logloss: 0.6703
Early stopping, best iteration is:
[44]	train's binary_logloss: 0.508564	valid's binary_logloss: 0.662989
num_leaves, val_score: 0.654569:   5%|5         | 1/20 [00:00<00:14,  1.28it/s][I 2020-09-27 04:44:07,670] Trial 7 finished with value: 0.6629889780799463 and parameters: {'num_leaves': 164}. Best is trial 7 with value: 0.6629889780799463.
num_leaves, val_score: 0.654569:   5%|5         | 1/20 [00:00<00:14,  1.28it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000332 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.410682	valid's binary_logloss: 0.66176
Early stopping, best iteration is:
[62]	train's binary_logloss: 0.482134	valid's binary_logloss: 0.658251
num_leaves, val_score: 0.654569:  10%|#         | 2/20 [00:01<00:14,  1.22it/s][I 2020-09-27 04:44:08,569] Trial 8 finished with value: 0.6582514287091177 and parameters: {'num_leaves': 142}. Best is trial 8 with value: 0.6582514287091177.
num_leaves, val_score: 0.654569:  10%|#         | 2/20 [00:01<00:14,  1.22it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000603 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.412047	valid's binary_logloss: 0.659767
Early stopping, best iteration is:
[58]	train's binary_logloss: 0.492004	valid's binary_logloss: 0.656213
num_leaves, val_score: 0.654569:  15%|#5        | 3/20 [00:02<00:14,  1.20it/s][I 2020-09-27 04:44:09,432] Trial 9 finished with value: 0.6562133173559287 and parameters: {'num_leaves': 140}. Best is trial 9 with value: 0.6562133173559287.
num_leaves, val_score: 0.654569:  15%|#5        | 3/20 [00:02<00:14,  1.20it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004277 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.608851	valid's binary_logloss: 0.656989
Early stopping, best iteration is:
[87]	train's binary_logloss: 0.614406	valid's binary_logloss: 0.656446
num_leaves, val_score: 0.654569:  20%|##        | 4/20 [00:03<00:11,  1.37it/s][I 2020-09-27 04:44:09,931] Trial 10 finished with value: 0.6564457093023696 and parameters: {'num_leaves': 19}. Best is trial 9 with value: 0.6562133173559287.
num_leaves, val_score: 0.654569:  20%|##        | 4/20 [00:03<00:11,  1.37it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004598 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.301686	valid's binary_logloss: 0.673949
Early stopping, best iteration is:
[31]	train's binary_logloss: 0.504321	valid's binary_logloss: 0.666042
num_leaves, val_score: 0.654569:  25%|##5       | 5/20 [00:04<00:13,  1.09it/s][I 2020-09-27 04:44:11,277] Trial 11 finished with value: 0.66604185972251 and parameters: {'num_leaves': 254}. Best is trial 9 with value: 0.6562133173559287.
num_leaves, val_score: 0.654569:  25%|##5       | 5/20 [00:04<00:13,  1.09it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000377 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.618907	valid's binary_logloss: 0.654583
[200]	train's binary_logloss: 0.590148	valid's binary_logloss: 0.653268
Early stopping, best iteration is:
[171]	train's binary_logloss: 0.597901	valid's binary_logloss: 0.652352
num_leaves, val_score: 0.652352:  30%|###       | 6/20 [00:04<00:11,  1.24it/s][I 2020-09-27 04:44:11,818] Trial 12 finished with value: 0.6523523176355726 and parameters: {'num_leaves': 15}. Best is trial 12 with value: 0.6523523176355726.
num_leaves, val_score: 0.652352:  30%|###       | 6/20 [00:04<00:11,  1.24it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000404 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.658358	valid's binary_logloss: 0.663406
[200]	train's binary_logloss: 0.647972	valid's binary_logloss: 0.655898
[300]	train's binary_logloss: 0.642797	valid's binary_logloss: 0.652806
[400]	train's binary_logloss: 0.639134	valid's binary_logloss: 0.651593
[500]	train's binary_logloss: 0.636175	valid's binary_logloss: 0.651137
[600]	train's binary_logloss: 0.633356	valid's binary_logloss: 0.65085
[700]	train's binary_logloss: 0.630829	valid's binary_logloss: 0.650912
Early stopping, best iteration is:
[631]	train's binary_logloss: 0.63254	valid's binary_logloss: 0.650592
num_leaves, val_score: 0.650592:  35%|###5      | 7/20 [00:05<00:11,  1.17it/s][I 2020-09-27 04:44:12,795] Trial 13 finished with value: 0.6505921558008332 and parameters: {'num_leaves': 3}. Best is trial 13 with value: 0.6505921558008332.
num_leaves, val_score: 0.650592:  35%|###5      | 7/20 [00:05<00:11,  1.17it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004382 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.643838	valid's binary_logloss: 0.655968
[200]	train's binary_logloss: 0.630606	valid's binary_logloss: 0.651656
[300]	train's binary_logloss: 0.62092	valid's binary_logloss: 0.651649
Early stopping, best iteration is:
[275]	train's binary_logloss: 0.623153	valid's binary_logloss: 0.651151
num_leaves, val_score: 0.650592:  40%|####      | 8/20 [00:06<00:09,  1.29it/s][I 2020-09-27 04:44:13,393] Trial 14 finished with value: 0.6511512029998484 and parameters: {'num_leaves': 6}. Best is trial 13 with value: 0.6505921558008332.
num_leaves, val_score: 0.650592:  40%|####      | 8/20 [00:06<00:09,  1.29it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005223 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.536652	valid's binary_logloss: 0.654767
Early stopping, best iteration is:
[88]	train's binary_logloss: 0.548536	valid's binary_logloss: 0.654093
num_leaves, val_score: 0.650592:  45%|####5     | 9/20 [00:07<00:09,  1.12it/s][I 2020-09-27 04:44:14,546] Trial 15 finished with value: 0.6540930395271628 and parameters: {'num_leaves': 55}. Best is trial 13 with value: 0.6505921558008332.
num_leaves, val_score: 0.650592:  45%|####5     | 9/20 [00:07<00:09,  1.12it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000391 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.522391	valid's binary_logloss: 0.660187
Early stopping, best iteration is:
[56]	train's binary_logloss: 0.573263	valid's binary_logloss: 0.659042
num_leaves, val_score: 0.650592:  50%|#####     | 10/20 [00:08<00:07,  1.25it/s][I 2020-09-27 04:44:15,132] Trial 16 finished with value: 0.6590416618563378 and parameters: {'num_leaves': 63}. Best is trial 13 with value: 0.6505921558008332.
num_leaves, val_score: 0.650592:  50%|#####     | 10/20 [00:08<00:07,  1.25it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000241 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.500962	valid's binary_logloss: 0.659121
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.552156	valid's binary_logloss: 0.658112
num_leaves, val_score: 0.650592:  55%|#####5    | 11/20 [00:08<00:06,  1.33it/s][I 2020-09-27 04:44:15,768] Trial 17 finished with value: 0.6581124834744749 and parameters: {'num_leaves': 77}. Best is trial 13 with value: 0.6505921558008332.
num_leaves, val_score: 0.650592:  55%|#####5    | 11/20 [00:08<00:06,  1.33it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000455 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.626446	valid's binary_logloss: 0.654342
[200]	train's binary_logloss: 0.602897	valid's binary_logloss: 0.654305
Early stopping, best iteration is:
[128]	train's binary_logloss: 0.619208	valid's binary_logloss: 0.653208
num_leaves, val_score: 0.650592:  60%|######    | 12/20 [00:09<00:05,  1.50it/s][I 2020-09-27 04:44:16,240] Trial 18 finished with value: 0.6532083579968055 and parameters: {'num_leaves': 12}. Best is trial 13 with value: 0.6505921558008332.
num_leaves, val_score: 0.650592:  60%|######    | 12/20 [00:09<00:05,  1.50it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000331 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.340384	valid's binary_logloss: 0.669675
Early stopping, best iteration is:
[43]	train's binary_logloss: 0.483746	valid's binary_logloss: 0.660937
num_leaves, val_score: 0.650592:  65%|######5   | 13/20 [00:10<00:05,  1.31it/s][I 2020-09-27 04:44:17,237] Trial 19 finished with value: 0.6609367707387375 and parameters: {'num_leaves': 209}. Best is trial 13 with value: 0.6505921558008332.
num_leaves, val_score: 0.650592:  65%|######5   | 13/20 [00:10<00:05,  1.31it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000416 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.47136	valid's binary_logloss: 0.661857
Early stopping, best iteration is:
[66]	train's binary_logloss: 0.520512	valid's binary_logloss: 0.659368
num_leaves, val_score: 0.650592:  70%|#######   | 14/20 [00:11<00:05,  1.08it/s][I 2020-09-27 04:44:18,525] Trial 20 finished with value: 0.6593681790082808 and parameters: {'num_leaves': 95}. Best is trial 13 with value: 0.6505921558008332.
num_leaves, val_score: 0.650592:  70%|#######   | 14/20 [00:11<00:05,  1.08it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000434 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.634417	valid's binary_logloss: 0.655116
[200]	train's binary_logloss: 0.616494	valid's binary_logloss: 0.653099
[300]	train's binary_logloss: 0.601536	valid's binary_logloss: 0.65265
Early stopping, best iteration is:
[250]	train's binary_logloss: 0.608783	valid's binary_logloss: 0.651395
num_leaves, val_score: 0.650592:  75%|#######5  | 15/20 [00:12<00:04,  1.20it/s][I 2020-09-27 04:44:19,142] Trial 21 finished with value: 0.6513954160672126 and parameters: {'num_leaves': 9}. Best is trial 13 with value: 0.6505921558008332.
num_leaves, val_score: 0.650592:  75%|#######5  | 15/20 [00:12<00:04,  1.20it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000366 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.668969	valid's binary_logloss: 0.671303
[200]	train's binary_logloss: 0.659328	valid's binary_logloss: 0.662192
[300]	train's binary_logloss: 0.65408	valid's binary_logloss: 0.657049
[400]	train's binary_logloss: 0.650963	valid's binary_logloss: 0.654645
[500]	train's binary_logloss: 0.648956	valid's binary_logloss: 0.652696
[600]	train's binary_logloss: 0.647608	valid's binary_logloss: 0.651783
[700]	train's binary_logloss: 0.646665	valid's binary_logloss: 0.651143
[800]	train's binary_logloss: 0.645958	valid's binary_logloss: 0.650652
[900]	train's binary_logloss: 0.645405	valid's binary_logloss: 0.6504
[1000]	train's binary_logloss: 0.644946	valid's binary_logloss: 0.650075
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.644946	valid's binary_logloss: 0.650075
num_leaves, val_score: 0.650075:  80%|########  | 16/20 [00:13<00:03,  1.05it/s][I 2020-09-27 04:44:20,393] Trial 22 finished with value: 0.6500754523975805 and parameters: {'num_leaves': 2}. Best is trial 22 with value: 0.6500754523975805.
num_leaves, val_score: 0.650075:  80%|########  | 16/20 [00:13<00:03,  1.05it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.005069 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.565299	valid's binary_logloss: 0.658702
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.597675	valid's binary_logloss: 0.657428
num_leaves, val_score: 0.650075:  85%|########5 | 17/20 [00:13<00:02,  1.23it/s][I 2020-09-27 04:44:20,871] Trial 23 finished with value: 0.6574278070760912 and parameters: {'num_leaves': 39}. Best is trial 22 with value: 0.6500754523975805.
num_leaves, val_score: 0.650075:  85%|########5 | 17/20 [00:13<00:02,  1.23it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000464 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.652156	valid's binary_logloss: 0.66018
[200]	train's binary_logloss: 0.641398	valid's binary_logloss: 0.653863
[300]	train's binary_logloss: 0.635014	valid's binary_logloss: 0.652565
[400]	train's binary_logloss: 0.629702	valid's binary_logloss: 0.65154
[500]	train's binary_logloss: 0.624786	valid's binary_logloss: 0.652002
Early stopping, best iteration is:
[409]	train's binary_logloss: 0.629299	valid's binary_logloss: 0.651463
num_leaves, val_score: 0.650075:  90%|######### | 18/20 [00:15<00:01,  1.05it/s][I 2020-09-27 04:44:22,155] Trial 24 finished with value: 0.6514630998860335 and parameters: {'num_leaves': 4}. Best is trial 22 with value: 0.6500754523975805.
num_leaves, val_score: 0.650075:  90%|######### | 18/20 [00:15<00:01,  1.05it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000333 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.570374	valid's binary_logloss: 0.656731
Early stopping, best iteration is:
[72]	train's binary_logloss: 0.590744	valid's binary_logloss: 0.655417
num_leaves, val_score: 0.650075:  95%|#########5| 19/20 [00:15<00:00,  1.23it/s][I 2020-09-27 04:44:22,639] Trial 25 finished with value: 0.655416641463825 and parameters: {'num_leaves': 37}. Best is trial 22 with value: 0.6500754523975805.
num_leaves, val_score: 0.650075:  95%|#########5| 19/20 [00:15<00:00,  1.23it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000693 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.455281	valid's binary_logloss: 0.66208
Early stopping, best iteration is:
[62]	train's binary_logloss: 0.51632	valid's binary_logloss: 0.658037
num_leaves, val_score: 0.650075: 100%|##########| 20/20 [00:16<00:00,  1.26it/s][I 2020-09-27 04:44:23,396] Trial 26 finished with value: 0.6580367008399806 and parameters: {'num_leaves': 107}. Best is trial 22 with value: 0.6500754523975805.
num_leaves, val_score: 0.650075: 100%|##########| 20/20 [00:16<00:00,  1.21it/s]
bagging, val_score: 0.650075:   0%|          | 0/10 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000433 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667722	valid's binary_logloss: 0.669888
[200]	train's binary_logloss: 0.657469	valid's binary_logloss: 0.660998
[300]	train's binary_logloss: 0.652189	valid's binary_logloss: 0.656002
[400]	train's binary_logloss: 0.649008	valid's binary_logloss: 0.65329
[500]	train's binary_logloss: 0.64724	valid's binary_logloss: 0.651649
[600]	train's binary_logloss: 0.646196	valid's binary_logloss: 0.651202
[700]	train's binary_logloss: 0.645282	valid's binary_logloss: 0.650767
[800]	train's binary_logloss: 0.644572	valid's binary_logloss: 0.650399
[900]	train's binary_logloss: 0.643859	valid's binary_logloss: 0.650238
[1000]	train's binary_logloss: 0.643295	valid's binary_logloss: 0.649949
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643295	valid's binary_logloss: 0.649949
bagging, val_score: 0.649949:  10%|#         | 1/10 [00:01<00:12,  1.38s/it][I 2020-09-27 04:44:24,791] Trial 27 finished with value: 0.6499492453087289 and parameters: {'bagging_fraction': 0.6374753391797192, 'bagging_freq': 6}. Best is trial 27 with value: 0.6499492453087289.
bagging, val_score: 0.649949:  10%|#         | 1/10 [00:01<00:12,  1.38s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000379 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667741	valid's binary_logloss: 0.670126
[200]	train's binary_logloss: 0.657401	valid's binary_logloss: 0.660759
[300]	train's binary_logloss: 0.652134	valid's binary_logloss: 0.655834
[400]	train's binary_logloss: 0.649009	valid's binary_logloss: 0.653296
[500]	train's binary_logloss: 0.647228	valid's binary_logloss: 0.651516
[600]	train's binary_logloss: 0.646169	valid's binary_logloss: 0.65095
[700]	train's binary_logloss: 0.645307	valid's binary_logloss: 0.650552
[800]	train's binary_logloss: 0.644552	valid's binary_logloss: 0.650135
[900]	train's binary_logloss: 0.643876	valid's binary_logloss: 0.650283
Early stopping, best iteration is:
[811]	train's binary_logloss: 0.64445	valid's binary_logloss: 0.649941
bagging, val_score: 0.649941:  20%|##        | 2/10 [00:03<00:11,  1.48s/it][I 2020-09-27 04:44:26,503] Trial 28 finished with value: 0.6499412034252128 and parameters: {'bagging_fraction': 0.6431039528688518, 'bagging_freq': 6}. Best is trial 28 with value: 0.6499412034252128.
bagging, val_score: 0.649941:  20%|##        | 2/10 [00:03<00:11,  1.48s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000402 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644511	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643837	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
bagging, val_score: 0.649721:  30%|###       | 3/10 [00:04<00:10,  1.44s/it][I 2020-09-27 04:44:27,842] Trial 29 finished with value: 0.6497207980729593 and parameters: {'bagging_fraction': 0.6147135266937416, 'bagging_freq': 6}. Best is trial 29 with value: 0.6497207980729593.
bagging, val_score: 0.649721:  30%|###       | 3/10 [00:04<00:10,  1.44s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000403 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.66762	valid's binary_logloss: 0.670222
[200]	train's binary_logloss: 0.657378	valid's binary_logloss: 0.660905
[300]	train's binary_logloss: 0.652057	valid's binary_logloss: 0.655995
[400]	train's binary_logloss: 0.648925	valid's binary_logloss: 0.653365
[500]	train's binary_logloss: 0.647153	valid's binary_logloss: 0.65154
[600]	train's binary_logloss: 0.646118	valid's binary_logloss: 0.651221
[700]	train's binary_logloss: 0.64522	valid's binary_logloss: 0.650848
[800]	train's binary_logloss: 0.644499	valid's binary_logloss: 0.650374
[900]	train's binary_logloss: 0.643818	valid's binary_logloss: 0.650522
Early stopping, best iteration is:
[811]	train's binary_logloss: 0.644399	valid's binary_logloss: 0.650183
bagging, val_score: 0.649721:  40%|####      | 4/10 [00:05<00:08,  1.40s/it][I 2020-09-27 04:44:29,154] Trial 30 finished with value: 0.6501826485329506 and parameters: {'bagging_fraction': 0.6282514585296812, 'bagging_freq': 6}. Best is trial 29 with value: 0.6497207980729593.
bagging, val_score: 0.649721:  40%|####      | 4/10 [00:05<00:08,  1.40s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000425 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667636	valid's binary_logloss: 0.670402
[200]	train's binary_logloss: 0.65741	valid's binary_logloss: 0.660721
[300]	train's binary_logloss: 0.652135	valid's binary_logloss: 0.655888
[400]	train's binary_logloss: 0.649005	valid's binary_logloss: 0.653394
[500]	train's binary_logloss: 0.647265	valid's binary_logloss: 0.651817
[600]	train's binary_logloss: 0.646169	valid's binary_logloss: 0.651429
[700]	train's binary_logloss: 0.645289	valid's binary_logloss: 0.651273
Early stopping, best iteration is:
[625]	train's binary_logloss: 0.645884	valid's binary_logloss: 0.650972
bagging, val_score: 0.649721:  50%|#####     | 5/10 [00:07<00:06,  1.39s/it][I 2020-09-27 04:44:30,538] Trial 31 finished with value: 0.6509722145299361 and parameters: {'bagging_fraction': 0.625788067973422, 'bagging_freq': 6}. Best is trial 29 with value: 0.6497207980729593.
bagging, val_score: 0.649721:  50%|#####     | 5/10 [00:07<00:06,  1.39s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000386 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.66771	valid's binary_logloss: 0.670672
[200]	train's binary_logloss: 0.657453	valid's binary_logloss: 0.660978
[300]	train's binary_logloss: 0.652108	valid's binary_logloss: 0.655884
[400]	train's binary_logloss: 0.648979	valid's binary_logloss: 0.653214
[500]	train's binary_logloss: 0.647206	valid's binary_logloss: 0.651492
[600]	train's binary_logloss: 0.646131	valid's binary_logloss: 0.650889
[700]	train's binary_logloss: 0.645274	valid's binary_logloss: 0.650764
[800]	train's binary_logloss: 0.644537	valid's binary_logloss: 0.650318
[900]	train's binary_logloss: 0.643867	valid's binary_logloss: 0.650313
Early stopping, best iteration is:
[811]	train's binary_logloss: 0.644435	valid's binary_logloss: 0.650123
bagging, val_score: 0.649721:  60%|######    | 6/10 [00:08<00:05,  1.34s/it][I 2020-09-27 04:44:31,752] Trial 32 finished with value: 0.650122761267748 and parameters: {'bagging_fraction': 0.6308230415432386, 'bagging_freq': 6}. Best is trial 29 with value: 0.6497207980729593.
bagging, val_score: 0.649721:  60%|######    | 6/10 [00:08<00:05,  1.34s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000509 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.668475	valid's binary_logloss: 0.670977
[200]	train's binary_logloss: 0.658545	valid's binary_logloss: 0.661641
[300]	train's binary_logloss: 0.653145	valid's binary_logloss: 0.656461
[400]	train's binary_logloss: 0.649945	valid's binary_logloss: 0.653659
[500]	train's binary_logloss: 0.647957	valid's binary_logloss: 0.652033
[600]	train's binary_logloss: 0.646663	valid's binary_logloss: 0.651209
[700]	train's binary_logloss: 0.645743	valid's binary_logloss: 0.650593
[800]	train's binary_logloss: 0.644998	valid's binary_logloss: 0.649891
[900]	train's binary_logloss: 0.644366	valid's binary_logloss: 0.649745
[1000]	train's binary_logloss: 0.643856	valid's binary_logloss: 0.649763
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643856	valid's binary_logloss: 0.649763
bagging, val_score: 0.649721:  70%|#######   | 7/10 [00:09<00:04,  1.41s/it][I 2020-09-27 04:44:33,330] Trial 33 finished with value: 0.649762914198675 and parameters: {'bagging_fraction': 0.881712034252823, 'bagging_freq': 6}. Best is trial 29 with value: 0.6497207980729593.
bagging, val_score: 0.649721:  70%|#######   | 7/10 [00:09<00:04,  1.41s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001411 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.668676	valid's binary_logloss: 0.671455
[200]	train's binary_logloss: 0.658729	valid's binary_logloss: 0.66187
[300]	train's binary_logloss: 0.653368	valid's binary_logloss: 0.656513
[400]	train's binary_logloss: 0.650201	valid's binary_logloss: 0.654227
[500]	train's binary_logloss: 0.648173	valid's binary_logloss: 0.652343
[600]	train's binary_logloss: 0.646868	valid's binary_logloss: 0.651436
[700]	train's binary_logloss: 0.645933	valid's binary_logloss: 0.650838
[800]	train's binary_logloss: 0.645188	valid's binary_logloss: 0.650137
[900]	train's binary_logloss: 0.644607	valid's binary_logloss: 0.64995
[1000]	train's binary_logloss: 0.644106	valid's binary_logloss: 0.649864
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.644106	valid's binary_logloss: 0.649864
bagging, val_score: 0.649721:  80%|########  | 8/10 [00:11<00:03,  1.51s/it][I 2020-09-27 04:44:35,075] Trial 34 finished with value: 0.6498641722707447 and parameters: {'bagging_fraction': 0.9222411833757059, 'bagging_freq': 6}. Best is trial 29 with value: 0.6497207980729593.
bagging, val_score: 0.649721:  80%|########  | 8/10 [00:11<00:03,  1.51s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000737 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.668902	valid's binary_logloss: 0.671648
[200]	train's binary_logloss: 0.65921	valid's binary_logloss: 0.661972
[300]	train's binary_logloss: 0.653923	valid's binary_logloss: 0.65701
[400]	train's binary_logloss: 0.650779	valid's binary_logloss: 0.654505
[500]	train's binary_logloss: 0.648744	valid's binary_logloss: 0.652567
[600]	train's binary_logloss: 0.647377	valid's binary_logloss: 0.651548
[700]	train's binary_logloss: 0.646422	valid's binary_logloss: 0.651081
[800]	train's binary_logloss: 0.645708	valid's binary_logloss: 0.650642
[900]	train's binary_logloss: 0.64514	valid's binary_logloss: 0.650392
[1000]	train's binary_logloss: 0.644665	valid's binary_logloss: 0.650089
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.644665	valid's binary_logloss: 0.650089
bagging, val_score: 0.649721:  90%|######### | 9/10 [00:13<00:01,  1.50s/it][I 2020-09-27 04:44:36,546] Trial 35 finished with value: 0.6500890749178282 and parameters: {'bagging_fraction': 0.9922180140725776, 'bagging_freq': 2}. Best is trial 29 with value: 0.6497207980729593.
bagging, val_score: 0.649721:  90%|######### | 9/10 [00:13<00:01,  1.50s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000379 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.66868	valid's binary_logloss: 0.671198
[200]	train's binary_logloss: 0.658732	valid's binary_logloss: 0.661411
[300]	train's binary_logloss: 0.653393	valid's binary_logloss: 0.656638
[400]	train's binary_logloss: 0.650183	valid's binary_logloss: 0.653991
[500]	train's binary_logloss: 0.648188	valid's binary_logloss: 0.652694
[600]	train's binary_logloss: 0.646874	valid's binary_logloss: 0.651422
[700]	train's binary_logloss: 0.645949	valid's binary_logloss: 0.650978
[800]	train's binary_logloss: 0.645235	valid's binary_logloss: 0.650342
[900]	train's binary_logloss: 0.644629	valid's binary_logloss: 0.649845
[1000]	train's binary_logloss: 0.644118	valid's binary_logloss: 0.649971
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.644118	valid's binary_logloss: 0.649971
bagging, val_score: 0.649721: 100%|##########| 10/10 [00:15<00:00,  1.61s/it][I 2020-09-27 04:44:38,424] Trial 36 finished with value: 0.6499714458134913 and parameters: {'bagging_fraction': 0.9261093121309201, 'bagging_freq': 7}. Best is trial 29 with value: 0.6497207980729593.
bagging, val_score: 0.649721: 100%|##########| 10/10 [00:15<00:00,  1.50s/it]
feature_fraction_stage2, val_score: 0.649721:   0%|          | 0/3 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000513 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667513	valid's binary_logloss: 0.670285
[200]	train's binary_logloss: 0.65728	valid's binary_logloss: 0.661217
[300]	train's binary_logloss: 0.652006	valid's binary_logloss: 0.656051
[400]	train's binary_logloss: 0.648934	valid's binary_logloss: 0.653472
[500]	train's binary_logloss: 0.647192	valid's binary_logloss: 0.651779
[600]	train's binary_logloss: 0.646099	valid's binary_logloss: 0.651503
[700]	train's binary_logloss: 0.645255	valid's binary_logloss: 0.650916
[800]	train's binary_logloss: 0.644508	valid's binary_logloss: 0.650651
[900]	train's binary_logloss: 0.643868	valid's binary_logloss: 0.650537
Early stopping, best iteration is:
[811]	train's binary_logloss: 0.644389	valid's binary_logloss: 0.650382
feature_fraction_stage2, val_score: 0.649721:  33%|###3      | 1/3 [00:01<00:02,  1.23s/it][I 2020-09-27 04:44:39,673] Trial 37 finished with value: 0.6503821331647123 and parameters: {'feature_fraction': 0.48000000000000004}. Best is trial 37 with value: 0.6503821331647123.
feature_fraction_stage2, val_score: 0.649721:  33%|###3      | 1/3 [00:01<00:02,  1.23s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000502 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667466	valid's binary_logloss: 0.670008
[200]	train's binary_logloss: 0.657197	valid's binary_logloss: 0.660612
[300]	train's binary_logloss: 0.651972	valid's binary_logloss: 0.655778
[400]	train's binary_logloss: 0.648867	valid's binary_logloss: 0.653147
[500]	train's binary_logloss: 0.647122	valid's binary_logloss: 0.651573
[600]	train's binary_logloss: 0.646035	valid's binary_logloss: 0.650926
[700]	train's binary_logloss: 0.645194	valid's binary_logloss: 0.650578
[800]	train's binary_logloss: 0.644427	valid's binary_logloss: 0.650523
[900]	train's binary_logloss: 0.643736	valid's binary_logloss: 0.650556
Early stopping, best iteration is:
[811]	train's binary_logloss: 0.644312	valid's binary_logloss: 0.650291
feature_fraction_stage2, val_score: 0.649721:  67%|######6   | 2/3 [00:02<00:01,  1.29s/it][I 2020-09-27 04:44:41,097] Trial 38 finished with value: 0.6502911144681769 and parameters: {'feature_fraction': 0.41600000000000004}. Best is trial 38 with value: 0.6502911144681769.
feature_fraction_stage2, val_score: 0.649721:  67%|######6   | 2/3 [00:02<00:01,  1.29s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.002477 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667513	valid's binary_logloss: 0.670285
[200]	train's binary_logloss: 0.65728	valid's binary_logloss: 0.661217
[300]	train's binary_logloss: 0.652006	valid's binary_logloss: 0.656051
[400]	train's binary_logloss: 0.648934	valid's binary_logloss: 0.653472
[500]	train's binary_logloss: 0.647192	valid's binary_logloss: 0.651779
[600]	train's binary_logloss: 0.646099	valid's binary_logloss: 0.651503
[700]	train's binary_logloss: 0.645255	valid's binary_logloss: 0.650916
[800]	train's binary_logloss: 0.644508	valid's binary_logloss: 0.650651
[900]	train's binary_logloss: 0.643868	valid's binary_logloss: 0.650537
Early stopping, best iteration is:
[811]	train's binary_logloss: 0.644389	valid's binary_logloss: 0.650382
feature_fraction_stage2, val_score: 0.649721: 100%|##########| 3/3 [00:04<00:00,  1.33s/it][I 2020-09-27 04:44:42,520] Trial 39 finished with value: 0.6503821331647123 and parameters: {'feature_fraction': 0.44800000000000006}. Best is trial 38 with value: 0.6502911144681769.
feature_fraction_stage2, val_score: 0.649721: 100%|##########| 3/3 [00:04<00:00,  1.36s/it]
regularization_factors, val_score: 0.649721:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000379 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644511	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643837	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649721:   5%|5         | 1/20 [00:01<00:25,  1.35s/it][I 2020-09-27 04:44:43,887] Trial 40 finished with value: 0.6497208080532643 and parameters: {'lambda_l1': 4.606489682248236e-06, 'lambda_l2': 0.0004490091062700392}. Best is trial 40 with value: 0.6497208080532643.
regularization_factors, val_score: 0.649721:   5%|5         | 1/20 [00:01<00:25,  1.35s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000342 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644511	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643837	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649721:  10%|#         | 2/20 [00:03<00:26,  1.50s/it][I 2020-09-27 04:44:45,731] Trial 41 finished with value: 0.6497208177282439 and parameters: {'lambda_l1': 3.367646783477069e-06, 'lambda_l2': 0.0006824013831978261}. Best is trial 40 with value: 0.6497208080532643.
regularization_factors, val_score: 0.649721:  10%|#         | 2/20 [00:03<00:26,  1.50s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000463 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644511	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643837	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649721:  15%|#5        | 3/20 [00:04<00:24,  1.44s/it][I 2020-09-27 04:44:47,018] Trial 42 finished with value: 0.6497208110181181 and parameters: {'lambda_l1': 1.2594003820953281e-06, 'lambda_l2': 0.0005901706717435754}. Best is trial 40 with value: 0.6497208080532643.
regularization_factors, val_score: 0.649721:  15%|#5        | 3/20 [00:04<00:24,  1.44s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000389 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644511	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643838	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643266	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643266	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649721:  20%|##        | 4/20 [00:05<00:22,  1.38s/it][I 2020-09-27 04:44:48,282] Trial 43 finished with value: 0.6497208467241731 and parameters: {'lambda_l1': 1.7272310002902818e-06, 'lambda_l2': 0.0005516895364430477}. Best is trial 40 with value: 0.6497208080532643.
regularization_factors, val_score: 0.649721:  20%|##        | 4/20 [00:05<00:22,  1.38s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000424 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644512	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643838	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643266	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643266	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649721:  25%|##5       | 5/20 [00:07<00:22,  1.49s/it][I 2020-09-27 04:44:50,019] Trial 44 finished with value: 0.6497209472919824 and parameters: {'lambda_l1': 1.66122924086294e-06, 'lambda_l2': 0.0004859937346429329}. Best is trial 40 with value: 0.6497208080532643.
regularization_factors, val_score: 0.649721:  25%|##5       | 5/20 [00:07<00:22,  1.49s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000580 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.64451	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643837	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643264	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643264	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649721:  30%|###       | 6/20 [00:08<00:20,  1.45s/it][I 2020-09-27 04:44:51,378] Trial 45 finished with value: 0.6497207058626523 and parameters: {'lambda_l1': 3.2445067392487893e-06, 'lambda_l2': 0.0004261866691401461}. Best is trial 45 with value: 0.6497207058626523.
regularization_factors, val_score: 0.649721:  30%|###       | 6/20 [00:08<00:20,  1.45s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000393 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644511	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643837	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643264	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643264	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649721:  35%|###5      | 7/20 [00:10<00:19,  1.53s/it][I 2020-09-27 04:44:53,078] Trial 46 finished with value: 0.6497207182286581 and parameters: {'lambda_l1': 3.850005827927253e-06, 'lambda_l2': 0.0005152754958015667}. Best is trial 45 with value: 0.6497207058626523.
regularization_factors, val_score: 0.649721:  35%|###5      | 7/20 [00:10<00:19,  1.53s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000406 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644511	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643837	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649721:  40%|####      | 8/20 [00:11<00:17,  1.49s/it][I 2020-09-27 04:44:54,488] Trial 47 finished with value: 0.6497208048936114 and parameters: {'lambda_l1': 6.216594964021316e-06, 'lambda_l2': 0.0003007341715934148}. Best is trial 45 with value: 0.6497207058626523.
regularization_factors, val_score: 0.649721:  40%|####      | 8/20 [00:11<00:17,  1.49s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000538 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647226	valid's binary_logloss: 0.65118
[600]	train's binary_logloss: 0.64616	valid's binary_logloss: 0.650744
[700]	train's binary_logloss: 0.645279	valid's binary_logloss: 0.650352
[800]	train's binary_logloss: 0.644543	valid's binary_logloss: 0.650069
[900]	train's binary_logloss: 0.643859	valid's binary_logloss: 0.649948
[1000]	train's binary_logloss: 0.643284	valid's binary_logloss: 0.649627
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643284	valid's binary_logloss: 0.649627
regularization_factors, val_score: 0.649627:  45%|####5     | 9/20 [00:13<00:15,  1.45s/it][I 2020-09-27 04:44:55,828] Trial 48 finished with value: 0.6496273811877121 and parameters: {'lambda_l1': 0.0008017384871250849, 'lambda_l2': 2.2446245169734426e-07}. Best is trial 48 with value: 0.6496273811877121.
regularization_factors, val_score: 0.649627:  45%|####5     | 9/20 [00:13<00:15,  1.45s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000443 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667565	valid's binary_logloss: 0.669845
[200]	train's binary_logloss: 0.657208	valid's binary_logloss: 0.660603
[300]	train's binary_logloss: 0.651961	valid's binary_logloss: 0.65582
[400]	train's binary_logloss: 0.648962	valid's binary_logloss: 0.65306
[500]	train's binary_logloss: 0.647204	valid's binary_logloss: 0.651318
[600]	train's binary_logloss: 0.646126	valid's binary_logloss: 0.650793
[700]	train's binary_logloss: 0.645262	valid's binary_logloss: 0.650639
[800]	train's binary_logloss: 0.644495	valid's binary_logloss: 0.650417
[900]	train's binary_logloss: 0.643822	valid's binary_logloss: 0.650127
[1000]	train's binary_logloss: 0.643242	valid's binary_logloss: 0.649774
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643242	valid's binary_logloss: 0.649774
regularization_factors, val_score: 0.649627:  50%|#####     | 10/20 [00:15<00:15,  1.53s/it][I 2020-09-27 04:44:57,545] Trial 49 finished with value: 0.6497744578765494 and parameters: {'lambda_l1': 0.012286863044744575, 'lambda_l2': 1.2600943736007061e-08}. Best is trial 48 with value: 0.6496273811877121.
regularization_factors, val_score: 0.649627:  50%|#####     | 10/20 [00:15<00:15,  1.53s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000454 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667583	valid's binary_logloss: 0.6698
[200]	train's binary_logloss: 0.657274	valid's binary_logloss: 0.66074
[300]	train's binary_logloss: 0.652046	valid's binary_logloss: 0.655766
[400]	train's binary_logloss: 0.649051	valid's binary_logloss: 0.65315
[500]	train's binary_logloss: 0.647312	valid's binary_logloss: 0.651523
[600]	train's binary_logloss: 0.646232	valid's binary_logloss: 0.651124
[700]	train's binary_logloss: 0.645397	valid's binary_logloss: 0.650768
[800]	train's binary_logloss: 0.644668	valid's binary_logloss: 0.650466
[900]	train's binary_logloss: 0.643987	valid's binary_logloss: 0.650261
[1000]	train's binary_logloss: 0.643431	valid's binary_logloss: 0.649835
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643431	valid's binary_logloss: 0.649835
regularization_factors, val_score: 0.649627:  55%|#####5    | 11/20 [00:16<00:13,  1.47s/it][I 2020-09-27 04:44:58,890] Trial 50 finished with value: 0.6498345397919513 and parameters: {'lambda_l1': 0.0005572386578667407, 'lambda_l2': 2.4208863399594898}. Best is trial 48 with value: 0.6496273811877121.
regularization_factors, val_score: 0.649627:  55%|#####5    | 11/20 [00:16<00:13,  1.47s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000464 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644511	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643837	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649627:  60%|######    | 12/20 [00:18<00:12,  1.52s/it][I 2020-09-27 04:45:00,537] Trial 51 finished with value: 0.6497208015259045 and parameters: {'lambda_l1': 8.699166692204462e-05, 'lambda_l2': 3.087574877344173e-06}. Best is trial 48 with value: 0.6496273811877121.
regularization_factors, val_score: 0.649627:  60%|######    | 12/20 [00:18<00:12,  1.52s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.002115 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644511	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643837	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649627:  65%|######5   | 13/20 [00:19<00:10,  1.50s/it][I 2020-09-27 04:45:01,964] Trial 52 finished with value: 0.6497208051737818 and parameters: {'lambda_l1': 0.00018230426317109586, 'lambda_l2': 4.548397960595403e-07}. Best is trial 48 with value: 0.6496273811877121.
regularization_factors, val_score: 0.649627:  65%|######5   | 13/20 [00:19<00:10,  1.50s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000506 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644511	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643837	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649627:  70%|#######   | 14/20 [00:20<00:08,  1.43s/it][I 2020-09-27 04:45:03,256] Trial 53 finished with value: 0.6497208043193139 and parameters: {'lambda_l1': 0.00015987949714084808, 'lambda_l2': 1.645060058121504e-06}. Best is trial 48 with value: 0.6496273811877121.
regularization_factors, val_score: 0.649627:  70%|#######   | 14/20 [00:20<00:08,  1.43s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000478 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644511	valid's binary_logloss: 0.650251
[900]	train's binary_logloss: 0.643837	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649627:  75%|#######5  | 15/20 [00:22<00:07,  1.54s/it][I 2020-09-27 04:45:05,051] Trial 54 finished with value: 0.6497208216187388 and parameters: {'lambda_l1': 0.0006042412420761168, 'lambda_l2': 1.2746381508914564e-06}. Best is trial 48 with value: 0.6496273811877121.
regularization_factors, val_score: 0.649627:  75%|#######5  | 15/20 [00:22<00:07,  1.54s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000393 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644511	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643837	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649627:  80%|########  | 16/20 [00:23<00:05,  1.50s/it][I 2020-09-27 04:45:06,437] Trial 55 finished with value: 0.6497208030993801 and parameters: {'lambda_l1': 0.00012765660632242594, 'lambda_l2': 3.053198787844645e-06}. Best is trial 48 with value: 0.6496273811877121.
regularization_factors, val_score: 0.649627:  80%|########  | 16/20 [00:23<00:05,  1.50s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.005265 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647221	valid's binary_logloss: 0.651249
[600]	train's binary_logloss: 0.646154	valid's binary_logloss: 0.650938
[700]	train's binary_logloss: 0.645276	valid's binary_logloss: 0.650398
[800]	train's binary_logloss: 0.644511	valid's binary_logloss: 0.65025
[900]	train's binary_logloss: 0.643837	valid's binary_logloss: 0.650107
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643265	valid's binary_logloss: 0.649721
regularization_factors, val_score: 0.649627:  85%|########5 | 17/20 [00:25<00:04,  1.47s/it][I 2020-09-27 04:45:07,858] Trial 56 finished with value: 0.6497208043349164 and parameters: {'lambda_l1': 4.556413554561587e-05, 'lambda_l2': 1.5942089823943423e-05}. Best is trial 48 with value: 0.6496273811877121.
regularization_factors, val_score: 0.649627:  85%|########5 | 17/20 [00:25<00:04,  1.47s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000436 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667565	valid's binary_logloss: 0.669845
[200]	train's binary_logloss: 0.657208	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651961	valid's binary_logloss: 0.655819
[400]	train's binary_logloss: 0.648963	valid's binary_logloss: 0.65306
[500]	train's binary_logloss: 0.647204	valid's binary_logloss: 0.651318
[600]	train's binary_logloss: 0.646126	valid's binary_logloss: 0.650793
[700]	train's binary_logloss: 0.645263	valid's binary_logloss: 0.650638
[800]	train's binary_logloss: 0.644496	valid's binary_logloss: 0.650417
[900]	train's binary_logloss: 0.643824	valid's binary_logloss: 0.650127
[1000]	train's binary_logloss: 0.643243	valid's binary_logloss: 0.649775
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643243	valid's binary_logloss: 0.649775
regularization_factors, val_score: 0.649627:  90%|######### | 18/20 [00:26<00:03,  1.52s/it][I 2020-09-27 04:45:09,475] Trial 57 finished with value: 0.6497749853707365 and parameters: {'lambda_l1': 9.790223265016275e-08, 'lambda_l2': 0.04135439527004478}. Best is trial 48 with value: 0.6496273811877121.
regularization_factors, val_score: 0.649627:  90%|######### | 18/20 [00:26<00:03,  1.52s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000415 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669845
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.65196	valid's binary_logloss: 0.655819
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647227	valid's binary_logloss: 0.65118
[600]	train's binary_logloss: 0.64616	valid's binary_logloss: 0.650744
[700]	train's binary_logloss: 0.64528	valid's binary_logloss: 0.650352
[800]	train's binary_logloss: 0.644543	valid's binary_logloss: 0.65007
[900]	train's binary_logloss: 0.64386	valid's binary_logloss: 0.649948
[1000]	train's binary_logloss: 0.643285	valid's binary_logloss: 0.649628
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643285	valid's binary_logloss: 0.649628
regularization_factors, val_score: 0.649627:  95%|#########5| 19/20 [00:28<00:01,  1.46s/it][I 2020-09-27 04:45:10,790] Trial 58 finished with value: 0.6496276632112322 and parameters: {'lambda_l1': 0.0038892159453483817, 'lambda_l2': 3.162000398193428e-08}. Best is trial 48 with value: 0.6496273811877121.
regularization_factors, val_score: 0.649627:  95%|#########5| 19/20 [00:28<00:01,  1.46s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000477 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667565	valid's binary_logloss: 0.669846
[200]	train's binary_logloss: 0.657208	valid's binary_logloss: 0.660603
[300]	train's binary_logloss: 0.651961	valid's binary_logloss: 0.65582
[400]	train's binary_logloss: 0.648963	valid's binary_logloss: 0.65306
[500]	train's binary_logloss: 0.647204	valid's binary_logloss: 0.651318
[600]	train's binary_logloss: 0.646126	valid's binary_logloss: 0.650793
[700]	train's binary_logloss: 0.645262	valid's binary_logloss: 0.650639
[800]	train's binary_logloss: 0.644495	valid's binary_logloss: 0.650418
[900]	train's binary_logloss: 0.643823	valid's binary_logloss: 0.650127
[1000]	train's binary_logloss: 0.643243	valid's binary_logloss: 0.649775
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643243	valid's binary_logloss: 0.649775
regularization_factors, val_score: 0.649627: 100%|##########| 20/20 [00:30<00:00,  1.59s/it][I 2020-09-27 04:45:12,683] Trial 59 finished with value: 0.6497745906504417 and parameters: {'lambda_l1': 0.012927722016684172, 'lambda_l2': 1.3099842616085507e-08}. Best is trial 48 with value: 0.6496273811877121.
regularization_factors, val_score: 0.649627: 100%|##########| 20/20 [00:30<00:00,  1.51s/it]
min_data_in_leaf, val_score: 0.649627:   0%|          | 0/5 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000494 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.652009	valid's binary_logloss: 0.655779
[400]	train's binary_logloss: 0.649064	valid's binary_logloss: 0.653022
[500]	train's binary_logloss: 0.647474	valid's binary_logloss: 0.651157
[600]	train's binary_logloss: 0.646465	valid's binary_logloss: 0.650811
[700]	train's binary_logloss: 0.645712	valid's binary_logloss: 0.650417
[800]	train's binary_logloss: 0.645041	valid's binary_logloss: 0.650168
[900]	train's binary_logloss: 0.644428	valid's binary_logloss: 0.650337
Early stopping, best iteration is:
[811]	train's binary_logloss: 0.644942	valid's binary_logloss: 0.65
min_data_in_leaf, val_score: 0.649627:  20%|##        | 1/5 [00:01<00:05,  1.26s/it][I 2020-09-27 04:45:13,958] Trial 60 finished with value: 0.6500004379994676 and parameters: {'min_child_samples': 100}. Best is trial 60 with value: 0.6500004379994676.
min_data_in_leaf, val_score: 0.649627:  20%|##        | 1/5 [00:01<00:05,  1.26s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000400 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648932	valid's binary_logloss: 0.652828
[500]	train's binary_logloss: 0.647269	valid's binary_logloss: 0.651176
[600]	train's binary_logloss: 0.646216	valid's binary_logloss: 0.650912
[700]	train's binary_logloss: 0.64537	valid's binary_logloss: 0.650588
[800]	train's binary_logloss: 0.644654	valid's binary_logloss: 0.65043
[900]	train's binary_logloss: 0.643973	valid's binary_logloss: 0.650308
Early stopping, best iteration is:
[811]	train's binary_logloss: 0.644548	valid's binary_logloss: 0.650197
min_data_in_leaf, val_score: 0.649627:  40%|####      | 2/5 [00:02<00:03,  1.26s/it][I 2020-09-27 04:45:15,217] Trial 61 finished with value: 0.6501970617141319 and parameters: {'min_child_samples': 50}. Best is trial 60 with value: 0.6500004379994676.
min_data_in_leaf, val_score: 0.649627:  40%|####      | 2/5 [00:02<00:03,  1.26s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000404 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.65195	valid's binary_logloss: 0.65587
[400]	train's binary_logloss: 0.648925	valid's binary_logloss: 0.65304
[500]	train's binary_logloss: 0.647138	valid's binary_logloss: 0.651221
[600]	train's binary_logloss: 0.646029	valid's binary_logloss: 0.650864
[700]	train's binary_logloss: 0.645079	valid's binary_logloss: 0.650534
[800]	train's binary_logloss: 0.644313	valid's binary_logloss: 0.650116
[900]	train's binary_logloss: 0.643631	valid's binary_logloss: 0.650098
Early stopping, best iteration is:
[811]	train's binary_logloss: 0.644204	valid's binary_logloss: 0.649909
min_data_in_leaf, val_score: 0.649627:  60%|######    | 3/5 [00:04<00:02,  1.36s/it][I 2020-09-27 04:45:16,819] Trial 62 finished with value: 0.6499094175863028 and parameters: {'min_child_samples': 5}. Best is trial 62 with value: 0.6499094175863028.
min_data_in_leaf, val_score: 0.649627:  60%|######    | 3/5 [00:04<00:02,  1.36s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000374 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.65195	valid's binary_logloss: 0.65587
[400]	train's binary_logloss: 0.648927	valid's binary_logloss: 0.653037
[500]	train's binary_logloss: 0.647134	valid's binary_logloss: 0.651144
[600]	train's binary_logloss: 0.646045	valid's binary_logloss: 0.650746
[700]	train's binary_logloss: 0.64509	valid's binary_logloss: 0.650337
[800]	train's binary_logloss: 0.644337	valid's binary_logloss: 0.650057
[900]	train's binary_logloss: 0.643666	valid's binary_logloss: 0.649979
Early stopping, best iteration is:
[811]	train's binary_logloss: 0.64422	valid's binary_logloss: 0.649764
min_data_in_leaf, val_score: 0.649627:  80%|########  | 4/5 [00:05<00:01,  1.32s/it][I 2020-09-27 04:45:18,047] Trial 63 finished with value: 0.649763563681734 and parameters: {'min_child_samples': 10}. Best is trial 63 with value: 0.649763563681734.
min_data_in_leaf, val_score: 0.649627:  80%|########  | 4/5 [00:05<00:01,  1.32s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000400 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.667564	valid's binary_logloss: 0.669844
[200]	train's binary_logloss: 0.657207	valid's binary_logloss: 0.660602
[300]	train's binary_logloss: 0.651959	valid's binary_logloss: 0.655818
[400]	train's binary_logloss: 0.648965	valid's binary_logloss: 0.653001
[500]	train's binary_logloss: 0.647226	valid's binary_logloss: 0.65118
[600]	train's binary_logloss: 0.646171	valid's binary_logloss: 0.650777
[700]	train's binary_logloss: 0.645308	valid's binary_logloss: 0.650525
[800]	train's binary_logloss: 0.644563	valid's binary_logloss: 0.650372
[900]	train's binary_logloss: 0.643935	valid's binary_logloss: 0.650147
[1000]	train's binary_logloss: 0.643333	valid's binary_logloss: 0.649925
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.643333	valid's binary_logloss: 0.649925
min_data_in_leaf, val_score: 0.649627: 100%|##########| 5/5 [00:06<00:00,  1.32s/it][I 2020-09-27 04:45:19,357] Trial 64 finished with value: 0.6499246052491834 and parameters: {'min_child_samples': 25}. Best is trial 63 with value: 0.649763563681734.
min_data_in_leaf, val_score: 0.649627: 100%|##########| 5/5 [00:06<00:00,  1.33s/it]
Fold : 7
[I 2020-09-27 04:45:19,501] A new study created in memory with name: no-name-a7907594-6487-49b5-9a75-d6ef7c2f3b45
feature_fraction, val_score: inf:   0%|          | 0/7 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008857 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.576754	valid's binary_logloss: 0.65897
Early stopping, best iteration is:
[69]	train's binary_logloss: 0.59731	valid's binary_logloss: 0.657355
feature_fraction, val_score: 0.657355:  14%|#4        | 1/7 [00:00<00:05,  1.05it/s][I 2020-09-27 04:45:20,471] Trial 0 finished with value: 0.6573548252031896 and parameters: {'feature_fraction': 0.6}. Best is trial 0 with value: 0.6573548252031896.
feature_fraction, val_score: 0.657355:  14%|#4        | 1/7 [00:00<00:05,  1.05it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000492 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.578741	valid's binary_logloss: 0.657569
[200]	train's binary_logloss: 0.527428	valid's binary_logloss: 0.66217
Early stopping, best iteration is:
[101]	train's binary_logloss: 0.578076	valid's binary_logloss: 0.657517
feature_fraction, val_score: 0.657355:  29%|##8       | 2/7 [00:01<00:04,  1.19it/s][I 2020-09-27 04:45:21,053] Trial 1 finished with value: 0.6575168666853799 and parameters: {'feature_fraction': 0.5}. Best is trial 0 with value: 0.6573548252031896.
feature_fraction, val_score: 0.657355:  29%|##8       | 2/7 [00:01<00:04,  1.19it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000966 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.572623	valid's binary_logloss: 0.65727
Early stopping, best iteration is:
[81]	train's binary_logloss: 0.585015	valid's binary_logloss: 0.656834
feature_fraction, val_score: 0.656834:  43%|####2     | 3/7 [00:02<00:03,  1.31it/s][I 2020-09-27 04:45:21,630] Trial 2 finished with value: 0.6568343510931819 and parameters: {'feature_fraction': 0.8999999999999999}. Best is trial 2 with value: 0.6568343510931819.
feature_fraction, val_score: 0.656834:  43%|####2     | 3/7 [00:02<00:03,  1.31it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008151 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.573507	valid's binary_logloss: 0.657017
[200]	train's binary_logloss: 0.517816	valid's binary_logloss: 0.65931
Early stopping, best iteration is:
[131]	train's binary_logloss: 0.55462	valid's binary_logloss: 0.655904
feature_fraction, val_score: 0.655904:  57%|#####7    | 4/7 [00:02<00:02,  1.36it/s][I 2020-09-27 04:45:22,306] Trial 3 finished with value: 0.6559041635016774 and parameters: {'feature_fraction': 0.8}. Best is trial 3 with value: 0.6559041635016774.
feature_fraction, val_score: 0.655904:  57%|#####7    | 4/7 [00:02<00:02,  1.36it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005133 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.576598	valid's binary_logloss: 0.657946
Early stopping, best iteration is:
[91]	train's binary_logloss: 0.582318	valid's binary_logloss: 0.656196
feature_fraction, val_score: 0.655904:  71%|#######1  | 5/7 [00:03<00:01,  1.49it/s][I 2020-09-27 04:45:22,831] Trial 4 finished with value: 0.6561957860921139 and parameters: {'feature_fraction': 0.7}. Best is trial 3 with value: 0.6559041635016774.
feature_fraction, val_score: 0.655904:  71%|#######1  | 5/7 [00:03<00:01,  1.49it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000955 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.570187	valid's binary_logloss: 0.658113
Early stopping, best iteration is:
[86]	train's binary_logloss: 0.579352	valid's binary_logloss: 0.657006
feature_fraction, val_score: 0.655904:  86%|########5 | 6/7 [00:04<00:00,  1.27it/s][I 2020-09-27 04:45:23,895] Trial 5 finished with value: 0.6570061744677883 and parameters: {'feature_fraction': 1.0}. Best is trial 3 with value: 0.6559041635016774.
feature_fraction, val_score: 0.655904:  86%|########5 | 6/7 [00:04<00:00,  1.27it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000543 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.582693	valid's binary_logloss: 0.657081
[200]	train's binary_logloss: 0.532209	valid's binary_logloss: 0.662404
Early stopping, best iteration is:
[100]	train's binary_logloss: 0.582693	valid's binary_logloss: 0.657081
feature_fraction, val_score: 0.655904: 100%|##########| 7/7 [00:04<00:00,  1.39it/s][I 2020-09-27 04:45:24,444] Trial 6 finished with value: 0.6570805680910938 and parameters: {'feature_fraction': 0.4}. Best is trial 3 with value: 0.6559041635016774.
feature_fraction, val_score: 0.655904: 100%|##########| 7/7 [00:04<00:00,  1.42it/s]
num_leaves, val_score: 0.655904:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004860 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.466323	valid's binary_logloss: 0.658116
Early stopping, best iteration is:
[54]	train's binary_logloss: 0.534041	valid's binary_logloss: 0.655259
num_leaves, val_score: 0.655259:   5%|5         | 1/20 [00:00<00:14,  1.35it/s][I 2020-09-27 04:45:25,204] Trial 7 finished with value: 0.6552590163821485 and parameters: {'num_leaves': 85}. Best is trial 7 with value: 0.6552590163821485.
num_leaves, val_score: 0.655259:   5%|5         | 1/20 [00:00<00:14,  1.35it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000895 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.540475	valid's binary_logloss: 0.662426
Early stopping, best iteration is:
[83]	train's binary_logloss: 0.556094	valid's binary_logloss: 0.65974
num_leaves, val_score: 0.655259:  10%|#         | 2/20 [00:01<00:12,  1.40it/s][I 2020-09-27 04:45:25,845] Trial 8 finished with value: 0.6597395208090874 and parameters: {'num_leaves': 46}. Best is trial 7 with value: 0.6552590163821485.
num_leaves, val_score: 0.655259:  10%|#         | 2/20 [00:01<00:12,  1.40it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004966 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.60271	valid's binary_logloss: 0.656639
[200]	train's binary_logloss: 0.564039	valid's binary_logloss: 0.658129
Early stopping, best iteration is:
[108]	train's binary_logloss: 0.59936	valid's binary_logloss: 0.65622
num_leaves, val_score: 0.655259:  15%|#5        | 3/20 [00:01<00:11,  1.53it/s][I 2020-09-27 04:45:26,358] Trial 9 finished with value: 0.6562201154063112 and parameters: {'num_leaves': 19}. Best is trial 7 with value: 0.6552590163821485.
num_leaves, val_score: 0.655259:  15%|#5        | 3/20 [00:01<00:11,  1.53it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000795 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.314803	valid's binary_logloss: 0.675418
Early stopping, best iteration is:
[17]	train's binary_logloss: 0.568221	valid's binary_logloss: 0.663895
num_leaves, val_score: 0.655259:  20%|##        | 4/20 [00:03<00:15,  1.05it/s][I 2020-09-27 04:45:28,019] Trial 10 finished with value: 0.6638954298043613 and parameters: {'num_leaves': 201}. Best is trial 7 with value: 0.6552590163821485.
num_leaves, val_score: 0.655259:  20%|##        | 4/20 [00:03<00:15,  1.05it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001158 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.415347	valid's binary_logloss: 0.669083
Early stopping, best iteration is:
[25]	train's binary_logloss: 0.574128	valid's binary_logloss: 0.660544
num_leaves, val_score: 0.655259:  25%|##5       | 5/20 [00:04<00:13,  1.10it/s][I 2020-09-27 04:45:28,824] Trial 11 finished with value: 0.6605435697983096 and parameters: {'num_leaves': 118}. Best is trial 7 with value: 0.6552590163821485.
num_leaves, val_score: 0.655259:  25%|##5       | 5/20 [00:04<00:13,  1.10it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000788 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.424268	valid's binary_logloss: 0.667432
Early stopping, best iteration is:
[24]	train's binary_logloss: 0.581498	valid's binary_logloss: 0.659094
num_leaves, val_score: 0.655259:  30%|###       | 6/20 [00:05<00:12,  1.16it/s][I 2020-09-27 04:45:29,584] Trial 12 finished with value: 0.6590937806617987 and parameters: {'num_leaves': 112}. Best is trial 7 with value: 0.6552590163821485.
num_leaves, val_score: 0.655259:  30%|###       | 6/20 [00:05<00:12,  1.16it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000922 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264846	valid's binary_logloss: 0.665644
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.441545	valid's binary_logloss: 0.654899
num_leaves, val_score: 0.654899:  35%|###5      | 7/20 [00:07<00:15,  1.17s/it][I 2020-09-27 04:45:31,476] Trial 13 finished with value: 0.6548985372701535 and parameters: {'num_leaves': 256}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899:  35%|###5      | 7/20 [00:07<00:15,  1.17s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000853 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.276152	valid's binary_logloss: 0.674007
Early stopping, best iteration is:
[19]	train's binary_logloss: 0.540876	valid's binary_logloss: 0.662293
num_leaves, val_score: 0.654899:  40%|####      | 8/20 [00:08<00:14,  1.17s/it][I 2020-09-27 04:45:32,652] Trial 14 finished with value: 0.6622934249190215 and parameters: {'num_leaves': 244}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899:  40%|####      | 8/20 [00:08<00:14,  1.17s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000843 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.475337	valid's binary_logloss: 0.661813
Early stopping, best iteration is:
[30]	train's binary_logloss: 0.586366	valid's binary_logloss: 0.658471
num_leaves, val_score: 0.654899:  45%|####5     | 9/20 [00:08<00:11,  1.02s/it][I 2020-09-27 04:45:33,298] Trial 15 finished with value: 0.6584705405485446 and parameters: {'num_leaves': 81}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899:  45%|####5     | 9/20 [00:08<00:11,  1.02s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.009872 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.32841	valid's binary_logloss: 0.676457
Early stopping, best iteration is:
[23]	train's binary_logloss: 0.544414	valid's binary_logloss: 0.664223
num_leaves, val_score: 0.654899:  50%|#####     | 10/20 [00:09<00:09,  1.01it/s][I 2020-09-27 04:45:34,239] Trial 16 finished with value: 0.664222940902035 and parameters: {'num_leaves': 189}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899:  50%|#####     | 10/20 [00:09<00:09,  1.01it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010397 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.353294	valid's binary_logloss: 0.668112
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.499593	valid's binary_logloss: 0.659529
num_leaves, val_score: 0.654899:  55%|#####5    | 11/20 [00:11<00:10,  1.17s/it][I 2020-09-27 04:45:35,832] Trial 17 finished with value: 0.6595289898494625 and parameters: {'num_leaves': 166}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899:  55%|#####5    | 11/20 [00:11<00:10,  1.17s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000949 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.263628	valid's binary_logloss: 0.683347
Early stopping, best iteration is:
[22]	train's binary_logloss: 0.520072	valid's binary_logloss: 0.662143
num_leaves, val_score: 0.654899:  60%|######    | 12/20 [00:12<00:09,  1.22s/it][I 2020-09-27 04:45:37,170] Trial 18 finished with value: 0.6621427253155968 and parameters: {'num_leaves': 254}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899:  60%|######    | 12/20 [00:12<00:09,  1.22s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004879 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.505464	valid's binary_logloss: 0.660159
Early stopping, best iteration is:
[34]	train's binary_logloss: 0.592348	valid's binary_logloss: 0.657805
num_leaves, val_score: 0.654899:  65%|######5   | 13/20 [00:13<00:07,  1.02s/it][I 2020-09-27 04:45:37,726] Trial 19 finished with value: 0.6578051021077662 and parameters: {'num_leaves': 64}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899:  65%|######5   | 13/20 [00:13<00:07,  1.02s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000866 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.373419	valid's binary_logloss: 0.673149
Early stopping, best iteration is:
[25]	train's binary_logloss: 0.55608	valid's binary_logloss: 0.664823
num_leaves, val_score: 0.654899:  70%|#######   | 14/20 [00:14<00:06,  1.13s/it][I 2020-09-27 04:45:39,108] Trial 20 finished with value: 0.6648226497863418 and parameters: {'num_leaves': 150}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899:  70%|#######   | 14/20 [00:14<00:06,  1.13s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000806 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.447378	valid's binary_logloss: 0.661981
Early stopping, best iteration is:
[54]	train's binary_logloss: 0.521425	valid's binary_logloss: 0.658385
num_leaves, val_score: 0.654899:  75%|#######5  | 15/20 [00:15<00:05,  1.04s/it][I 2020-09-27 04:45:39,938] Trial 21 finished with value: 0.6583849556869334 and parameters: {'num_leaves': 97}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899:  75%|#######5  | 15/20 [00:15<00:05,  1.04s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004821 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650824	valid's binary_logloss: 0.661115
[200]	train's binary_logloss: 0.639749	valid's binary_logloss: 0.657461
[300]	train's binary_logloss: 0.632853	valid's binary_logloss: 0.656219
Early stopping, best iteration is:
[286]	train's binary_logloss: 0.633724	valid's binary_logloss: 0.656027
num_leaves, val_score: 0.654899:  80%|########  | 16/20 [00:16<00:03,  1.08it/s][I 2020-09-27 04:45:40,584] Trial 22 finished with value: 0.6560272050011768 and parameters: {'num_leaves': 4}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899:  80%|########  | 16/20 [00:16<00:03,  1.08it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000847 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.295072	valid's binary_logloss: 0.672517
Early stopping, best iteration is:
[35]	train's binary_logloss: 0.474148	valid's binary_logloss: 0.65871
num_leaves, val_score: 0.654899:  85%|########5 | 17/20 [00:17<00:03,  1.02s/it][I 2020-09-27 04:45:41,847] Trial 23 finished with value: 0.6587097853479928 and parameters: {'num_leaves': 222}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899:  85%|########5 | 17/20 [00:17<00:03,  1.02s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.006283 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.559059	valid's binary_logloss: 0.657941
Early stopping, best iteration is:
[78]	train's binary_logloss: 0.57617	valid's binary_logloss: 0.656731
num_leaves, val_score: 0.654899:  90%|######### | 18/20 [00:18<00:02,  1.04s/it][I 2020-09-27 04:45:42,913] Trial 24 finished with value: 0.6567308128514865 and parameters: {'num_leaves': 37}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899:  90%|######### | 18/20 [00:18<00:02,  1.04s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010220 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.388417	valid's binary_logloss: 0.668977
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.521183	valid's binary_logloss: 0.659883
num_leaves, val_score: 0.654899:  95%|#########5| 19/20 [00:19<00:01,  1.02s/it][I 2020-09-27 04:45:43,893] Trial 25 finished with value: 0.659882543351282 and parameters: {'num_leaves': 138}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899:  95%|#########5| 19/20 [00:19<00:01,  1.02s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001162 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.465903	valid's binary_logloss: 0.661386
Early stopping, best iteration is:
[58]	train's binary_logloss: 0.526905	valid's binary_logloss: 0.657041
num_leaves, val_score: 0.654899: 100%|##########| 20/20 [00:20<00:00,  1.03it/s][I 2020-09-27 04:45:44,739] Trial 26 finished with value: 0.6570410039446851 and parameters: {'num_leaves': 86}. Best is trial 13 with value: 0.6548985372701535.
num_leaves, val_score: 0.654899: 100%|##########| 20/20 [00:20<00:00,  1.01s/it]
bagging, val_score: 0.654899:   0%|          | 0/10 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004721 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.278384	valid's binary_logloss: 0.695429
Early stopping, best iteration is:
[15]	train's binary_logloss: 0.574718	valid's binary_logloss: 0.670457
bagging, val_score: 0.654899:  10%|#         | 1/10 [00:01<00:10,  1.20s/it][I 2020-09-27 04:45:45,957] Trial 27 finished with value: 0.6704565961142148 and parameters: {'bagging_fraction': 0.5211974818467775, 'bagging_freq': 5}. Best is trial 27 with value: 0.6704565961142148.
bagging, val_score: 0.654899:  10%|#         | 1/10 [00:01<00:10,  1.20s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.018210 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.262977	valid's binary_logloss: 0.676276
Early stopping, best iteration is:
[30]	train's binary_logloss: 0.4766	valid's binary_logloss: 0.659897
bagging, val_score: 0.654899:  20%|##        | 2/10 [00:02<00:10,  1.37s/it][I 2020-09-27 04:45:47,718] Trial 28 finished with value: 0.6598967542506259 and parameters: {'bagging_fraction': 0.9934815132316437, 'bagging_freq': 1}. Best is trial 28 with value: 0.6598967542506259.
bagging, val_score: 0.654899:  20%|##        | 2/10 [00:02<00:10,  1.37s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000944 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264681	valid's binary_logloss: 0.684914
Early stopping, best iteration is:
[17]	train's binary_logloss: 0.549363	valid's binary_logloss: 0.665808
bagging, val_score: 0.654899:  30%|###       | 3/10 [00:04<00:09,  1.35s/it][I 2020-09-27 04:45:49,025] Trial 29 finished with value: 0.6658075550967804 and parameters: {'bagging_fraction': 0.9461366644242709, 'bagging_freq': 7}. Best is trial 28 with value: 0.6598967542506259.
bagging, val_score: 0.654899:  30%|###       | 3/10 [00:04<00:09,  1.35s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000798 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.281086	valid's binary_logloss: 0.707672
Early stopping, best iteration is:
[17]	train's binary_logloss: 0.564424	valid's binary_logloss: 0.669487
bagging, val_score: 0.654899:  40%|####      | 4/10 [00:06<00:08,  1.49s/it][I 2020-09-27 04:45:50,825] Trial 30 finished with value: 0.6694865333360387 and parameters: {'bagging_fraction': 0.4198139622048719, 'bagging_freq': 1}. Best is trial 28 with value: 0.6598967542506259.
bagging, val_score: 0.654899:  40%|####      | 4/10 [00:06<00:08,  1.49s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000847 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.266862	valid's binary_logloss: 0.689038
Early stopping, best iteration is:
[24]	train's binary_logloss: 0.514004	valid's binary_logloss: 0.66343
bagging, val_score: 0.654899:  50%|#####     | 5/10 [00:07<00:07,  1.51s/it][I 2020-09-27 04:45:52,393] Trial 31 finished with value: 0.663429926936811 and parameters: {'bagging_fraction': 0.7512986978703702, 'bagging_freq': 4}. Best is trial 28 with value: 0.6598967542506259.
bagging, val_score: 0.654899:  50%|#####     | 5/10 [00:07<00:07,  1.51s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004195 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.265958	valid's binary_logloss: 0.68931
Early stopping, best iteration is:
[18]	train's binary_logloss: 0.549252	valid's binary_logloss: 0.6651
bagging, val_score: 0.654899:  60%|######    | 6/10 [00:09<00:05,  1.48s/it][I 2020-09-27 04:45:53,804] Trial 32 finished with value: 0.6650999977344688 and parameters: {'bagging_fraction': 0.7159980215103693, 'bagging_freq': 7}. Best is trial 28 with value: 0.6598967542506259.
bagging, val_score: 0.654899:  60%|######    | 6/10 [00:09<00:05,  1.48s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.014957 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.272788	valid's binary_logloss: 0.699964
Early stopping, best iteration is:
[22]	train's binary_logloss: 0.532238	valid's binary_logloss: 0.667015
bagging, val_score: 0.654899:  70%|#######   | 7/10 [00:10<00:04,  1.52s/it][I 2020-09-27 04:45:55,408] Trial 33 finished with value: 0.6670147726739027 and parameters: {'bagging_fraction': 0.5563678837629615, 'bagging_freq': 2}. Best is trial 28 with value: 0.6598967542506259.
bagging, val_score: 0.654899:  70%|#######   | 7/10 [00:10<00:04,  1.52s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000817 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.267946	valid's binary_logloss: 0.681235
Early stopping, best iteration is:
[21]	train's binary_logloss: 0.52815	valid's binary_logloss: 0.661059
bagging, val_score: 0.654899:  80%|########  | 8/10 [00:12<00:02,  1.47s/it][I 2020-09-27 04:45:56,765] Trial 34 finished with value: 0.6610593718244844 and parameters: {'bagging_fraction': 0.8313758530340752, 'bagging_freq': 3}. Best is trial 28 with value: 0.6598967542506259.
bagging, val_score: 0.654899:  80%|########  | 8/10 [00:12<00:02,  1.47s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001483 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.292079	valid's binary_logloss: 0.717802
Early stopping, best iteration is:
[16]	train's binary_logloss: 0.575885	valid's binary_logloss: 0.667398
bagging, val_score: 0.654899:  90%|######### | 9/10 [00:13<00:01,  1.56s/it][I 2020-09-27 04:45:58,520] Trial 35 finished with value: 0.6673978773322778 and parameters: {'bagging_fraction': 0.4012246265638672, 'bagging_freq': 5}. Best is trial 28 with value: 0.6598967542506259.
bagging, val_score: 0.654899:  90%|######### | 9/10 [00:13<00:01,  1.56s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.006489 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.274203	valid's binary_logloss: 0.693572
Early stopping, best iteration is:
[23]	train's binary_logloss: 0.525889	valid's binary_logloss: 0.667929
bagging, val_score: 0.654899: 100%|##########| 10/10 [00:14<00:00,  1.46s/it][I 2020-09-27 04:45:59,749] Trial 36 finished with value: 0.667928676492384 and parameters: {'bagging_fraction': 0.6154723571783186, 'bagging_freq': 6}. Best is trial 28 with value: 0.6598967542506259.
bagging, val_score: 0.654899: 100%|##########| 10/10 [00:15<00:00,  1.50s/it]
feature_fraction_stage2, val_score: 0.654899:   0%|          | 0/6 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004865 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.269387	valid's binary_logloss: 0.674759
Early stopping, best iteration is:
[25]	train's binary_logloss: 0.505622	valid's binary_logloss: 0.660533
feature_fraction_stage2, val_score: 0.654899:  17%|#6        | 1/6 [00:01<00:05,  1.11s/it][I 2020-09-27 04:46:00,871] Trial 37 finished with value: 0.6605329410813362 and parameters: {'feature_fraction': 0.7200000000000001}. Best is trial 37 with value: 0.6605329410813362.
feature_fraction_stage2, val_score: 0.654899:  17%|#6        | 1/6 [00:01<00:05,  1.11s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001084 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264846	valid's binary_logloss: 0.665644
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.441545	valid's binary_logloss: 0.654899
feature_fraction_stage2, val_score: 0.654899:  33%|###3      | 2/6 [00:03<00:05,  1.35s/it][I 2020-09-27 04:46:02,793] Trial 38 finished with value: 0.6548985372701535 and parameters: {'feature_fraction': 0.8160000000000001}. Best is trial 38 with value: 0.6548985372701535.
feature_fraction_stage2, val_score: 0.654899:  33%|###3      | 2/6 [00:03<00:05,  1.35s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004430 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.261299	valid's binary_logloss: 0.680887
Early stopping, best iteration is:
[20]	train's binary_logloss: 0.526475	valid's binary_logloss: 0.664407
feature_fraction_stage2, val_score: 0.654899:  50%|#####     | 3/6 [00:04<00:03,  1.29s/it][I 2020-09-27 04:46:03,948] Trial 39 finished with value: 0.6644073193077987 and parameters: {'feature_fraction': 0.88}. Best is trial 38 with value: 0.6548985372701535.
feature_fraction_stage2, val_score: 0.654899:  50%|#####     | 3/6 [00:04<00:03,  1.29s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000627 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.265461	valid's binary_logloss: 0.682897
Early stopping, best iteration is:
[24]	train's binary_logloss: 0.509383	valid's binary_logloss: 0.662438
feature_fraction_stage2, val_score: 0.654899:  67%|######6   | 4/6 [00:06<00:03,  1.74s/it][I 2020-09-27 04:46:06,737] Trial 40 finished with value: 0.6624379814652182 and parameters: {'feature_fraction': 0.7520000000000001}. Best is trial 38 with value: 0.6548985372701535.
feature_fraction_stage2, val_score: 0.654899:  67%|######6   | 4/6 [00:06<00:03,  1.74s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005011 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.265461	valid's binary_logloss: 0.682897
Early stopping, best iteration is:
[24]	train's binary_logloss: 0.509383	valid's binary_logloss: 0.662438
feature_fraction_stage2, val_score: 0.654899:  83%|########3 | 5/6 [00:08<00:01,  1.57s/it][I 2020-09-27 04:46:07,913] Trial 41 finished with value: 0.6624379814652182 and parameters: {'feature_fraction': 0.784}. Best is trial 38 with value: 0.6548985372701535.
feature_fraction_stage2, val_score: 0.654899:  83%|########3 | 5/6 [00:08<00:01,  1.57s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000490 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.261804	valid's binary_logloss: 0.683679
Early stopping, best iteration is:
[25]	train's binary_logloss: 0.500705	valid's binary_logloss: 0.666858
feature_fraction_stage2, val_score: 0.654899: 100%|##########| 6/6 [00:09<00:00,  1.64s/it][I 2020-09-27 04:46:09,700] Trial 42 finished with value: 0.6668584743708171 and parameters: {'feature_fraction': 0.8480000000000001}. Best is trial 38 with value: 0.6548985372701535.
feature_fraction_stage2, val_score: 0.654899: 100%|##########| 6/6 [00:09<00:00,  1.66s/it]
regularization_factors, val_score: 0.654899:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001030 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.265642	valid's binary_logloss: 0.678724
Early stopping, best iteration is:
[32]	train's binary_logloss: 0.469659	valid's binary_logloss: 0.660888
regularization_factors, val_score: 0.654899:   5%|5         | 1/20 [00:01<00:29,  1.58s/it][I 2020-09-27 04:46:11,296] Trial 43 finished with value: 0.6608876654466368 and parameters: {'lambda_l1': 2.395209311350816e-06, 'lambda_l2': 0.13259120292527007}. Best is trial 43 with value: 0.6608876654466368.
regularization_factors, val_score: 0.654899:   5%|5         | 1/20 [00:01<00:29,  1.58s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000545 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.390048	valid's binary_logloss: 0.672992
Early stopping, best iteration is:
[35]	train's binary_logloss: 0.535479	valid's binary_logloss: 0.6619
regularization_factors, val_score: 0.654899:  10%|#         | 2/20 [00:03<00:29,  1.62s/it][I 2020-09-27 04:46:13,025] Trial 44 finished with value: 0.6619000150383266 and parameters: {'lambda_l1': 6.3824374533786505, 'lambda_l2': 2.9167015559251105e-08}. Best is trial 43 with value: 0.6608876654466368.
regularization_factors, val_score: 0.654899:  10%|#         | 2/20 [00:03<00:29,  1.62s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000772 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.295559	valid's binary_logloss: 0.675205
Early stopping, best iteration is:
[31]	train's binary_logloss: 0.4927	valid's binary_logloss: 0.660235
regularization_factors, val_score: 0.654899:  15%|#5        | 3/20 [00:05<00:30,  1.78s/it][I 2020-09-27 04:46:15,164] Trial 45 finished with value: 0.6602349076536601 and parameters: {'lambda_l1': 1.6485218908617958, 'lambda_l2': 1.1064051625349032e-08}. Best is trial 45 with value: 0.6602349076536601.
regularization_factors, val_score: 0.654899:  15%|#5        | 3/20 [00:05<00:30,  1.78s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000792 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.263724	valid's binary_logloss: 0.666125
Early stopping, best iteration is:
[37]	train's binary_logloss: 0.44514	valid's binary_logloss: 0.65487
regularization_factors, val_score: 0.654870:  20%|##        | 4/20 [00:06<00:27,  1.71s/it][I 2020-09-27 04:46:16,718] Trial 46 finished with value: 0.654869889386949 and parameters: {'lambda_l1': 3.0122223559697e-08, 'lambda_l2': 0.00021523441226079608}. Best is trial 46 with value: 0.654869889386949.
regularization_factors, val_score: 0.654870:  20%|##        | 4/20 [00:07<00:27,  1.71s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000873 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.263726	valid's binary_logloss: 0.666214
Early stopping, best iteration is:
[37]	train's binary_logloss: 0.445141	valid's binary_logloss: 0.654998
regularization_factors, val_score: 0.654870:  25%|##5       | 5/20 [00:09<00:27,  1.80s/it][I 2020-09-27 04:46:18,736] Trial 47 finished with value: 0.6549975103511791 and parameters: {'lambda_l1': 5.742885425790403e-08, 'lambda_l2': 0.0002767008648395505}. Best is trial 46 with value: 0.654869889386949.
regularization_factors, val_score: 0.654870:  25%|##5       | 5/20 [00:09<00:27,  1.80s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001046 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264666	valid's binary_logloss: 0.666984
Early stopping, best iteration is:
[37]	train's binary_logloss: 0.445139	valid's binary_logloss: 0.654961
regularization_factors, val_score: 0.654870:  30%|###       | 6/20 [00:10<00:24,  1.73s/it][I 2020-09-27 04:46:20,296] Trial 48 finished with value: 0.6549605608644221 and parameters: {'lambda_l1': 8.607940177177832e-08, 'lambda_l2': 0.0001457048379141767}. Best is trial 46 with value: 0.654869889386949.
regularization_factors, val_score: 0.654870:  30%|###       | 6/20 [00:10<00:24,  1.73s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000921 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.263723	valid's binary_logloss: 0.666273
Early stopping, best iteration is:
[37]	train's binary_logloss: 0.445139	valid's binary_logloss: 0.655013
regularization_factors, val_score: 0.654870:  35%|###5      | 7/20 [00:12<00:24,  1.87s/it][I 2020-09-27 04:46:22,480] Trial 49 finished with value: 0.6550133681715706 and parameters: {'lambda_l1': 1.014605269031525e-08, 'lambda_l2': 0.00017168614663450724}. Best is trial 46 with value: 0.654869889386949.
regularization_factors, val_score: 0.654870:  35%|###5      | 7/20 [00:12<00:24,  1.87s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004784 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.263723	valid's binary_logloss: 0.666224
Early stopping, best iteration is:
[37]	train's binary_logloss: 0.445139	valid's binary_logloss: 0.654928
regularization_factors, val_score: 0.654870:  40%|####      | 8/20 [00:14<00:20,  1.72s/it][I 2020-09-27 04:46:23,853] Trial 50 finished with value: 0.6549280045293944 and parameters: {'lambda_l1': 1.4571400606028584e-08, 'lambda_l2': 0.00018545051962322963}. Best is trial 46 with value: 0.654869889386949.
regularization_factors, val_score: 0.654870:  40%|####      | 8/20 [00:14<00:20,  1.72s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001114 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.263723	valid's binary_logloss: 0.666238
Early stopping, best iteration is:
[37]	train's binary_logloss: 0.445139	valid's binary_logloss: 0.654958
regularization_factors, val_score: 0.654870:  45%|####5     | 9/20 [00:16<00:20,  1.83s/it][I 2020-09-27 04:46:25,937] Trial 51 finished with value: 0.6549580786078956 and parameters: {'lambda_l1': 1.1439375782625129e-08, 'lambda_l2': 0.00020040952933499966}. Best is trial 46 with value: 0.654869889386949.
regularization_factors, val_score: 0.654870:  45%|####5     | 9/20 [00:16<00:20,  1.83s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004784 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264844	valid's binary_logloss: 0.665542
Early stopping, best iteration is:
[43]	train's binary_logloss: 0.421453	valid's binary_logloss: 0.654899
regularization_factors, val_score: 0.654870:  50%|#####     | 10/20 [00:17<00:17,  1.71s/it][I 2020-09-27 04:46:27,383] Trial 52 finished with value: 0.6548985889877852 and parameters: {'lambda_l1': 1.3842789551680254e-08, 'lambda_l2': 5.6495755460565416e-05}. Best is trial 46 with value: 0.654869889386949.
regularization_factors, val_score: 0.654870:  50%|#####     | 10/20 [00:17<00:17,  1.71s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004465 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264846	valid's binary_logloss: 0.665605
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.441545	valid's binary_logloss: 0.654867
regularization_factors, val_score: 0.654867:  55%|#####5    | 11/20 [00:19<00:16,  1.80s/it][I 2020-09-27 04:46:29,389] Trial 53 finished with value: 0.6548668432849638 and parameters: {'lambda_l1': 1.189642528157414e-08, 'lambda_l2': 5.203275530782603e-06}. Best is trial 53 with value: 0.6548668432849638.
regularization_factors, val_score: 0.654867:  55%|#####5    | 11/20 [00:19<00:16,  1.80s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000851 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264846	valid's binary_logloss: 0.665516
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.441545	valid's binary_logloss: 0.654838
regularization_factors, val_score: 0.654838:  60%|######    | 12/20 [00:21<00:13,  1.73s/it][I 2020-09-27 04:46:30,946] Trial 54 finished with value: 0.6548376800578837 and parameters: {'lambda_l1': 2.1670289548196517e-06, 'lambda_l2': 1.661838347245463e-06}. Best is trial 54 with value: 0.6548376800578837.
regularization_factors, val_score: 0.654838:  60%|######    | 12/20 [00:21<00:13,  1.73s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000858 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264846	valid's binary_logloss: 0.66563
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.441545	valid's binary_logloss: 0.654905
regularization_factors, val_score: 0.654838:  65%|######5   | 13/20 [00:23<00:12,  1.84s/it][I 2020-09-27 04:46:33,033] Trial 55 finished with value: 0.6549047704174789 and parameters: {'lambda_l1': 1.3047485080865155e-05, 'lambda_l2': 1.1388372443834088e-06}. Best is trial 54 with value: 0.6548376800578837.
regularization_factors, val_score: 0.654838:  65%|######5   | 13/20 [00:23<00:12,  1.84s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000859 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264846	valid's binary_logloss: 0.665482
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.441545	valid's binary_logloss: 0.654874
regularization_factors, val_score: 0.654838:  70%|#######   | 14/20 [00:24<00:10,  1.75s/it][I 2020-09-27 04:46:34,599] Trial 56 finished with value: 0.6548744228748046 and parameters: {'lambda_l1': 1.677207413460934e-06, 'lambda_l2': 2.3547411212794384e-06}. Best is trial 54 with value: 0.6548376800578837.
regularization_factors, val_score: 0.654838:  70%|#######   | 14/20 [00:24<00:10,  1.75s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000869 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264846	valid's binary_logloss: 0.665557
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.441545	valid's binary_logloss: 0.654891
regularization_factors, val_score: 0.654838:  75%|#######5  | 15/20 [00:27<00:09,  1.87s/it][I 2020-09-27 04:46:36,725] Trial 57 finished with value: 0.65489106531928 and parameters: {'lambda_l1': 2.808990582697963e-06, 'lambda_l2': 3.680142856828128e-06}. Best is trial 54 with value: 0.6548376800578837.
regularization_factors, val_score: 0.654838:  75%|#######5  | 15/20 [00:27<00:09,  1.87s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005001 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264846	valid's binary_logloss: 0.665537
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.441545	valid's binary_logloss: 0.654883
regularization_factors, val_score: 0.654838:  80%|########  | 16/20 [00:28<00:07,  1.75s/it][I 2020-09-27 04:46:38,214] Trial 58 finished with value: 0.6548833147276146 and parameters: {'lambda_l1': 3.548471771976571e-06, 'lambda_l2': 1.3069611448432158e-06}. Best is trial 54 with value: 0.6548376800578837.
regularization_factors, val_score: 0.654838:  80%|########  | 16/20 [00:28<00:07,  1.75s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000875 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264846	valid's binary_logloss: 0.665518
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.441545	valid's binary_logloss: 0.654891
regularization_factors, val_score: 0.654838:  85%|########5 | 17/20 [00:30<00:05,  1.86s/it][I 2020-09-27 04:46:40,318] Trial 59 finished with value: 0.654890562836266 and parameters: {'lambda_l1': 3.6573101362801625e-06, 'lambda_l2': 1.3655499334616953e-06}. Best is trial 54 with value: 0.6548376800578837.
regularization_factors, val_score: 0.654838:  85%|########5 | 17/20 [00:30<00:05,  1.86s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000836 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.263781	valid's binary_logloss: 0.676626
Early stopping, best iteration is:
[41]	train's binary_logloss: 0.42877	valid's binary_logloss: 0.654741
regularization_factors, val_score: 0.654741:  90%|######### | 18/20 [00:32<00:03,  1.78s/it][I 2020-09-27 04:46:41,916] Trial 60 finished with value: 0.6547406780914222 and parameters: {'lambda_l1': 8.237457561826076e-05, 'lambda_l2': 7.033731015507948e-07}. Best is trial 60 with value: 0.6547406780914222.
regularization_factors, val_score: 0.654741:  90%|######### | 18/20 [00:32<00:03,  1.78s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000852 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264845	valid's binary_logloss: 0.66557
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.441546	valid's binary_logloss: 0.654905
regularization_factors, val_score: 0.654741:  95%|#########5| 19/20 [00:34<00:01,  1.87s/it][I 2020-09-27 04:46:43,993] Trial 61 finished with value: 0.6549054519516266 and parameters: {'lambda_l1': 8.016334404835216e-05, 'lambda_l2': 8.515948497846046e-07}. Best is trial 60 with value: 0.6547406780914222.
regularization_factors, val_score: 0.654741:  95%|#########5| 19/20 [00:34<00:01,  1.87s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000852 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.264846	valid's binary_logloss: 0.665533
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.441545	valid's binary_logloss: 0.65488
regularization_factors, val_score: 0.654741: 100%|##########| 20/20 [00:35<00:00,  1.78s/it][I 2020-09-27 04:46:45,569] Trial 62 finished with value: 0.6548801766746073 and parameters: {'lambda_l1': 6.807709893234408e-07, 'lambda_l2': 1.3368946882298463e-06}. Best is trial 60 with value: 0.6547406780914222.
regularization_factors, val_score: 0.654741: 100%|##########| 20/20 [00:35<00:00,  1.79s/it]
min_data_in_leaf, val_score: 0.654741:   0%|          | 0/5 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000788 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.26674	valid's binary_logloss: 0.673371
Early stopping, best iteration is:
[37]	train's binary_logloss: 0.44854	valid's binary_logloss: 0.659997
min_data_in_leaf, val_score: 0.654741:  20%|##        | 1/5 [00:02<00:08,  2.00s/it][I 2020-09-27 04:46:47,591] Trial 63 finished with value: 0.6599970673875724 and parameters: {'min_child_samples': 25}. Best is trial 63 with value: 0.6599970673875724.
min_data_in_leaf, val_score: 0.654741:  20%|##        | 1/5 [00:02<00:08,  2.00s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000851 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.259932	valid's binary_logloss: 0.674681
Early stopping, best iteration is:
[27]	train's binary_logloss: 0.485136	valid's binary_logloss: 0.663118
min_data_in_leaf, val_score: 0.654741:  40%|####      | 2/5 [00:03<00:05,  1.90s/it][I 2020-09-27 04:46:49,234] Trial 64 finished with value: 0.6631181004712279 and parameters: {'min_child_samples': 5}. Best is trial 63 with value: 0.6599970673875724.
min_data_in_leaf, val_score: 0.654741:  40%|####      | 2/5 [00:03<00:05,  1.90s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000813 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.277711	valid's binary_logloss: 0.677648
Early stopping, best iteration is:
[35]	train's binary_logloss: 0.469589	valid's binary_logloss: 0.66334
min_data_in_leaf, val_score: 0.654741:  60%|######    | 3/5 [00:05<00:03,  1.93s/it][I 2020-09-27 04:46:51,228] Trial 65 finished with value: 0.6633400797460791 and parameters: {'min_child_samples': 50}. Best is trial 63 with value: 0.6599970673875724.
min_data_in_leaf, val_score: 0.654741:  60%|######    | 3/5 [00:05<00:03,  1.93s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000892 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.262769	valid's binary_logloss: 0.674802
Early stopping, best iteration is:
[35]	train's binary_logloss: 0.44869	valid's binary_logloss: 0.662608
min_data_in_leaf, val_score: 0.654741:  80%|########  | 4/5 [00:07<00:01,  1.86s/it][I 2020-09-27 04:46:52,928] Trial 66 finished with value: 0.6626078522269558 and parameters: {'min_child_samples': 10}. Best is trial 63 with value: 0.6599970673875724.
min_data_in_leaf, val_score: 0.654741:  80%|########  | 4/5 [00:07<00:01,  1.86s/it][LightGBM] [Info] Number of positive: 12849, number of negative: 13150
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000877 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4238
[LightGBM] [Info] Number of data points in the train set: 25999, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494211 -> initscore=-0.023156
[LightGBM] [Info] Start training from score -0.023156
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
Training until validation scores don't improve for 100 rounds
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[100]	train's binary_logloss: 0.355756	valid's binary_logloss: 0.678018
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf
Early stopping, best iteration is:
[32]	train's binary_logloss: 0.530938	valid's binary_logloss: 0.659107
min_data_in_leaf, val_score: 0.654741: 100%|##########| 5/5 [00:10<00:00,  2.23s/it][I 2020-09-27 04:46:56,042] Trial 67 finished with value: 0.6591070735201429 and parameters: {'min_child_samples': 100}. Best is trial 67 with value: 0.6591070735201429.
min_data_in_leaf, val_score: 0.654741: 100%|##########| 5/5 [00:10<00:00,  2.09s/it]
Fold : 8
[I 2020-09-27 04:46:56,086] A new study created in memory with name: no-name-0ac6def8-9fc0-44ca-a007-4dfc936cce3b
feature_fraction, val_score: inf:   0%|          | 0/7 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000362 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.582362	valid's binary_logloss: 0.657172
Early stopping, best iteration is:
[75]	train's binary_logloss: 0.598686	valid's binary_logloss: 0.656832
feature_fraction, val_score: 0.656832:  14%|#4        | 1/7 [00:00<00:03,  1.84it/s][I 2020-09-27 04:46:56,647] Trial 0 finished with value: 0.6568322981471987 and parameters: {'feature_fraction': 0.4}. Best is trial 0 with value: 0.6568322981471987.
feature_fraction, val_score: 0.656832:  14%|#4        | 1/7 [00:00<00:03,  1.84it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004661 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.575682	valid's binary_logloss: 0.657591
Early stopping, best iteration is:
[65]	train's binary_logloss: 0.598895	valid's binary_logloss: 0.65488
feature_fraction, val_score: 0.654880:  29%|##8       | 2/7 [00:01<00:02,  1.90it/s][I 2020-09-27 04:46:57,132] Trial 1 finished with value: 0.654879918185127 and parameters: {'feature_fraction': 0.7}. Best is trial 1 with value: 0.654879918185127.
feature_fraction, val_score: 0.654880:  29%|##8       | 2/7 [00:01<00:02,  1.90it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000886 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.573899	valid's binary_logloss: 0.657206
Early stopping, best iteration is:
[70]	train's binary_logloss: 0.593743	valid's binary_logloss: 0.655846
feature_fraction, val_score: 0.654880:  43%|####2     | 3/7 [00:01<00:02,  1.85it/s][I 2020-09-27 04:46:57,704] Trial 2 finished with value: 0.6558455623356558 and parameters: {'feature_fraction': 0.8}. Best is trial 1 with value: 0.654879918185127.
feature_fraction, val_score: 0.654880:  43%|####2     | 3/7 [00:01<00:02,  1.85it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000972 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.57084	valid's binary_logloss: 0.654546
[200]	train's binary_logloss: 0.513329	valid's binary_logloss: 0.658828
Early stopping, best iteration is:
[113]	train's binary_logloss: 0.562928	valid's binary_logloss: 0.654073
feature_fraction, val_score: 0.654073:  57%|#####7    | 4/7 [00:02<00:01,  1.73it/s][I 2020-09-27 04:46:58,369] Trial 3 finished with value: 0.6540730143979465 and parameters: {'feature_fraction': 1.0}. Best is trial 3 with value: 0.6540730143979465.
feature_fraction, val_score: 0.654073:  57%|#####7    | 4/7 [00:02<00:01,  1.73it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004601 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.579014	valid's binary_logloss: 0.655596
Early stopping, best iteration is:
[59]	train's binary_logloss: 0.606682	valid's binary_logloss: 0.6549
feature_fraction, val_score: 0.654073:  71%|#######1  | 5/7 [00:02<00:01,  1.63it/s][I 2020-09-27 04:46:59,068] Trial 4 finished with value: 0.6548998752828071 and parameters: {'feature_fraction': 0.5}. Best is trial 3 with value: 0.6540730143979465.
feature_fraction, val_score: 0.654073:  71%|#######1  | 5/7 [00:02<00:01,  1.63it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.011929 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.572184	valid's binary_logloss: 0.658179
Early stopping, best iteration is:
[67]	train's binary_logloss: 0.594979	valid's binary_logloss: 0.656087
feature_fraction, val_score: 0.654073:  86%|########5 | 6/7 [00:03<00:00,  1.52it/s][I 2020-09-27 04:46:59,834] Trial 5 finished with value: 0.6560873120955287 and parameters: {'feature_fraction': 0.8999999999999999}. Best is trial 3 with value: 0.6540730143979465.
feature_fraction, val_score: 0.654073:  86%|########5 | 6/7 [00:03<00:00,  1.52it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004851 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.576475	valid's binary_logloss: 0.652717
Early stopping, best iteration is:
[82]	train's binary_logloss: 0.587886	valid's binary_logloss: 0.65223
feature_fraction, val_score: 0.652230: 100%|##########| 7/7 [00:04<00:00,  1.64it/s][I 2020-09-27 04:47:00,328] Trial 6 finished with value: 0.6522296496116577 and parameters: {'feature_fraction': 0.6}. Best is trial 6 with value: 0.6522296496116577.
feature_fraction, val_score: 0.652230: 100%|##########| 7/7 [00:04<00:00,  1.65it/s]
num_leaves, val_score: 0.652230:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004842 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.326219	valid's binary_logloss: 0.670645
Early stopping, best iteration is:
[35]	train's binary_logloss: 0.497287	valid's binary_logloss: 0.658257
num_leaves, val_score: 0.652230:   5%|5         | 1/20 [00:00<00:17,  1.08it/s][I 2020-09-27 04:47:01,277] Trial 7 finished with value: 0.6582573078988242 and parameters: {'num_leaves': 200}. Best is trial 7 with value: 0.6582573078988242.
num_leaves, val_score: 0.652230:   5%|5         | 1/20 [00:00<00:17,  1.08it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004568 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.281259	valid's binary_logloss: 0.679658
Early stopping, best iteration is:
[36]	train's binary_logloss: 0.462423	valid's binary_logloss: 0.663543
num_leaves, val_score: 0.652230:  10%|#         | 2/20 [00:02<00:17,  1.03it/s][I 2020-09-27 04:47:02,352] Trial 8 finished with value: 0.6635430249480666 and parameters: {'num_leaves': 251}. Best is trial 7 with value: 0.6582573078988242.
num_leaves, val_score: 0.652230:  10%|#         | 2/20 [00:02<00:17,  1.03it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005025 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657954	valid's binary_logloss: 0.664118
[200]	train's binary_logloss: 0.647761	valid's binary_logloss: 0.655763
[300]	train's binary_logloss: 0.642622	valid's binary_logloss: 0.652688
[400]	train's binary_logloss: 0.638865	valid's binary_logloss: 0.651406
[500]	train's binary_logloss: 0.635649	valid's binary_logloss: 0.650648
[600]	train's binary_logloss: 0.632824	valid's binary_logloss: 0.650562
[700]	train's binary_logloss: 0.630241	valid's binary_logloss: 0.650393
Early stopping, best iteration is:
[694]	train's binary_logloss: 0.630392	valid's binary_logloss: 0.650237
num_leaves, val_score: 0.650237:  15%|#5        | 3/20 [00:03<00:19,  1.17s/it][I 2020-09-27 04:47:03,969] Trial 9 finished with value: 0.65023693049223 and parameters: {'num_leaves': 3}. Best is trial 9 with value: 0.65023693049223.
num_leaves, val_score: 0.650237:  15%|#5        | 3/20 [00:03<00:19,  1.17s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004669 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657954	valid's binary_logloss: 0.664118
[200]	train's binary_logloss: 0.647761	valid's binary_logloss: 0.655763
[300]	train's binary_logloss: 0.642622	valid's binary_logloss: 0.652688
[400]	train's binary_logloss: 0.638865	valid's binary_logloss: 0.651406
[500]	train's binary_logloss: 0.635649	valid's binary_logloss: 0.650648
[600]	train's binary_logloss: 0.632824	valid's binary_logloss: 0.650562
[700]	train's binary_logloss: 0.630241	valid's binary_logloss: 0.650393
Early stopping, best iteration is:
[694]	train's binary_logloss: 0.630392	valid's binary_logloss: 0.650237
num_leaves, val_score: 0.650237:  20%|##        | 4/20 [00:04<00:18,  1.18s/it][I 2020-09-27 04:47:05,175] Trial 10 finished with value: 0.6502369304922299 and parameters: {'num_leaves': 3}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  20%|##        | 4/20 [00:04<00:18,  1.18s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005031 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.628976	valid's binary_logloss: 0.654294
[200]	train's binary_logloss: 0.608343	valid's binary_logloss: 0.65336
Early stopping, best iteration is:
[176]	train's binary_logloss: 0.613043	valid's binary_logloss: 0.652178
num_leaves, val_score: 0.650237:  25%|##5       | 5/20 [00:05<00:14,  1.01it/s][I 2020-09-27 04:47:05,732] Trial 11 finished with value: 0.6521775910881051 and parameters: {'num_leaves': 10}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  25%|##5       | 5/20 [00:05<00:14,  1.01it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004975 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.635408	valid's binary_logloss: 0.65585
[200]	train's binary_logloss: 0.618644	valid's binary_logloss: 0.654948
Early stopping, best iteration is:
[161]	train's binary_logloss: 0.624573	valid's binary_logloss: 0.6542
num_leaves, val_score: 0.650237:  30%|###       | 6/20 [00:05<00:11,  1.17it/s][I 2020-09-27 04:47:06,273] Trial 12 finished with value: 0.6541997386028066 and parameters: {'num_leaves': 8}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  30%|###       | 6/20 [00:05<00:11,  1.17it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000898 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.488946	valid's binary_logloss: 0.659659
Early stopping, best iteration is:
[39]	train's binary_logloss: 0.5773	valid's binary_logloss: 0.65642
num_leaves, val_score: 0.650237:  35%|###5      | 7/20 [00:07<00:12,  1.04it/s][I 2020-09-27 04:47:07,468] Trial 13 finished with value: 0.6564196118293095 and parameters: {'num_leaves': 76}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  35%|###5      | 7/20 [00:07<00:12,  1.04it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000884 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.498154	valid's binary_logloss: 0.662208
Early stopping, best iteration is:
[39]	train's binary_logloss: 0.581162	valid's binary_logloss: 0.657338
num_leaves, val_score: 0.650237:  40%|####      | 8/20 [00:07<00:10,  1.16it/s][I 2020-09-27 04:47:08,100] Trial 14 finished with value: 0.6573383283023841 and parameters: {'num_leaves': 71}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  40%|####      | 8/20 [00:07<00:10,  1.16it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004661 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.646593	valid's binary_logloss: 0.657131
[200]	train's binary_logloss: 0.634403	valid's binary_logloss: 0.652331
[300]	train's binary_logloss: 0.625932	valid's binary_logloss: 0.651212
Early stopping, best iteration is:
[289]	train's binary_logloss: 0.626825	valid's binary_logloss: 0.650977
num_leaves, val_score: 0.650237:  45%|####5     | 9/20 [00:08<00:08,  1.27it/s][I 2020-09-27 04:47:08,723] Trial 15 finished with value: 0.6509770662129648 and parameters: {'num_leaves': 5}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  45%|####5     | 9/20 [00:08<00:08,  1.27it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004588 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.498154	valid's binary_logloss: 0.662208
Early stopping, best iteration is:
[39]	train's binary_logloss: 0.581162	valid's binary_logloss: 0.657338
num_leaves, val_score: 0.650237:  50%|#####     | 10/20 [00:08<00:07,  1.40it/s][I 2020-09-27 04:47:09,256] Trial 16 finished with value: 0.6573383283023841 and parameters: {'num_leaves': 71}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  50%|#####     | 10/20 [00:08<00:07,  1.40it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012770 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.392369	valid's binary_logloss: 0.669429
Early stopping, best iteration is:
[45]	train's binary_logloss: 0.507218	valid's binary_logloss: 0.65932
num_leaves, val_score: 0.650237:  55%|#####5    | 11/20 [00:09<00:06,  1.35it/s][I 2020-09-27 04:47:10,058] Trial 17 finished with value: 0.659319879723429 and parameters: {'num_leaves': 141}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  55%|#####5    | 11/20 [00:09<00:06,  1.35it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010521 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.553596	valid's binary_logloss: 0.658319
Early stopping, best iteration is:
[54]	train's binary_logloss: 0.593584	valid's binary_logloss: 0.656096
num_leaves, val_score: 0.650237:  60%|######    | 12/20 [00:10<00:06,  1.25it/s][I 2020-09-27 04:47:10,992] Trial 18 finished with value: 0.6560958307756177 and parameters: {'num_leaves': 42}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  60%|######    | 12/20 [00:10<00:06,  1.25it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001062 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.409616	valid's binary_logloss: 0.665328
Early stopping, best iteration is:
[37]	train's binary_logloss: 0.53918	valid's binary_logloss: 0.658274
num_leaves, val_score: 0.650237:  65%|######5   | 13/20 [00:11<00:05,  1.19it/s][I 2020-09-27 04:47:11,928] Trial 19 finished with value: 0.6582743650249848 and parameters: {'num_leaves': 128}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  65%|######5   | 13/20 [00:11<00:05,  1.19it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004694 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.563913	valid's binary_logloss: 0.657197
Early stopping, best iteration is:
[77]	train's binary_logloss: 0.580794	valid's binary_logloss: 0.654988
num_leaves, val_score: 0.650237:  70%|#######   | 14/20 [00:12<00:04,  1.37it/s][I 2020-09-27 04:47:12,412] Trial 20 finished with value: 0.6549884052454669 and parameters: {'num_leaves': 37}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  70%|#######   | 14/20 [00:12<00:04,  1.37it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.009279 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657954	valid's binary_logloss: 0.664118
[200]	train's binary_logloss: 0.647761	valid's binary_logloss: 0.655763
[300]	train's binary_logloss: 0.642622	valid's binary_logloss: 0.652688
[400]	train's binary_logloss: 0.638865	valid's binary_logloss: 0.651406
[500]	train's binary_logloss: 0.635649	valid's binary_logloss: 0.650648
[600]	train's binary_logloss: 0.632824	valid's binary_logloss: 0.650562
[700]	train's binary_logloss: 0.630241	valid's binary_logloss: 0.650393
Early stopping, best iteration is:
[694]	train's binary_logloss: 0.630392	valid's binary_logloss: 0.650237
num_leaves, val_score: 0.650237:  75%|#######5  | 15/20 [00:13<00:04,  1.18it/s][I 2020-09-27 04:47:13,534] Trial 21 finished with value: 0.6502369304922299 and parameters: {'num_leaves': 3}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  75%|#######5  | 15/20 [00:13<00:04,  1.18it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000785 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.646593	valid's binary_logloss: 0.657131
[200]	train's binary_logloss: 0.634403	valid's binary_logloss: 0.652331
[300]	train's binary_logloss: 0.625932	valid's binary_logloss: 0.651212
Early stopping, best iteration is:
[289]	train's binary_logloss: 0.626825	valid's binary_logloss: 0.650977
num_leaves, val_score: 0.650237:  80%|########  | 16/20 [00:14<00:03,  1.08it/s][I 2020-09-27 04:47:14,641] Trial 22 finished with value: 0.6509770662129648 and parameters: {'num_leaves': 5}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  80%|########  | 16/20 [00:14<00:03,  1.08it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.019722 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.565177	valid's binary_logloss: 0.655115
Early stopping, best iteration is:
[72]	train's binary_logloss: 0.585373	valid's binary_logloss: 0.65339
num_leaves, val_score: 0.650237:  85%|########5 | 17/20 [00:14<00:02,  1.20it/s][I 2020-09-27 04:47:15,254] Trial 23 finished with value: 0.6533900763638031 and parameters: {'num_leaves': 36}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  85%|########5 | 17/20 [00:14<00:02,  1.20it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004882 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.565177	valid's binary_logloss: 0.655115
Early stopping, best iteration is:
[72]	train's binary_logloss: 0.585373	valid's binary_logloss: 0.65339
num_leaves, val_score: 0.650237:  90%|######### | 18/20 [00:15<00:01,  1.36it/s][I 2020-09-27 04:47:15,759] Trial 24 finished with value: 0.6533900763638031 and parameters: {'num_leaves': 36}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  90%|######### | 18/20 [00:15<00:01,  1.36it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.009941 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.668764	valid's binary_logloss: 0.672745
[200]	train's binary_logloss: 0.65923	valid's binary_logloss: 0.663108
[300]	train's binary_logloss: 0.654067	valid's binary_logloss: 0.658349
[400]	train's binary_logloss: 0.651004	valid's binary_logloss: 0.655269
[500]	train's binary_logloss: 0.649032	valid's binary_logloss: 0.653548
[600]	train's binary_logloss: 0.647704	valid's binary_logloss: 0.652575
[700]	train's binary_logloss: 0.646761	valid's binary_logloss: 0.651831
[800]	train's binary_logloss: 0.646051	valid's binary_logloss: 0.651481
[900]	train's binary_logloss: 0.645487	valid's binary_logloss: 0.651115
[1000]	train's binary_logloss: 0.645019	valid's binary_logloss: 0.651086
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.645019	valid's binary_logloss: 0.651086
num_leaves, val_score: 0.650237:  95%|#########5| 19/20 [00:16<00:00,  1.12it/s][I 2020-09-27 04:47:17,024] Trial 25 finished with value: 0.6510861135332809 and parameters: {'num_leaves': 2}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237:  95%|#########5| 19/20 [00:16<00:00,  1.12it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005016 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.424526	valid's binary_logloss: 0.659895
Early stopping, best iteration is:
[33]	train's binary_logloss: 0.556295	valid's binary_logloss: 0.654704
num_leaves, val_score: 0.650237: 100%|##########| 20/20 [00:17<00:00,  1.20it/s][I 2020-09-27 04:47:17,712] Trial 26 finished with value: 0.6547044966874118 and parameters: {'num_leaves': 119}. Best is trial 10 with value: 0.6502369304922299.
num_leaves, val_score: 0.650237: 100%|##########| 20/20 [00:17<00:00,  1.15it/s]
bagging, val_score: 0.650237:   0%|          | 0/10 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004738 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.658045	valid's binary_logloss: 0.664127
[200]	train's binary_logloss: 0.647774	valid's binary_logloss: 0.655872
[300]	train's binary_logloss: 0.642497	valid's binary_logloss: 0.65234
[400]	train's binary_logloss: 0.638669	valid's binary_logloss: 0.651347
[500]	train's binary_logloss: 0.635478	valid's binary_logloss: 0.651168
Early stopping, best iteration is:
[453]	train's binary_logloss: 0.63696	valid's binary_logloss: 0.650827
bagging, val_score: 0.650237:  10%|#         | 1/10 [00:01<00:12,  1.44s/it][I 2020-09-27 04:47:19,168] Trial 27 finished with value: 0.6508272913841253 and parameters: {'bagging_fraction': 0.9963833291336945, 'bagging_freq': 6}. Best is trial 27 with value: 0.6508272913841253.
bagging, val_score: 0.650237:  10%|#         | 1/10 [00:01<00:12,  1.44s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004596 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.655889	valid's binary_logloss: 0.659928
[200]	train's binary_logloss: 0.646169	valid's binary_logloss: 0.654108
[300]	train's binary_logloss: 0.641381	valid's binary_logloss: 0.651946
[400]	train's binary_logloss: 0.637836	valid's binary_logloss: 0.650942
Early stopping, best iteration is:
[389]	train's binary_logloss: 0.638252	valid's binary_logloss: 0.650231
bagging, val_score: 0.650231:  20%|##        | 2/10 [00:02<00:10,  1.27s/it][I 2020-09-27 04:47:20,036] Trial 28 finished with value: 0.6502306281879949 and parameters: {'bagging_fraction': 0.4514619063105707, 'bagging_freq': 1}. Best is trial 28 with value: 0.6502306281879949.
bagging, val_score: 0.650231:  20%|##        | 2/10 [00:02<00:10,  1.27s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004540 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.655825	valid's binary_logloss: 0.660152
[200]	train's binary_logloss: 0.646253	valid's binary_logloss: 0.653794
[300]	train's binary_logloss: 0.64168	valid's binary_logloss: 0.652426
[400]	train's binary_logloss: 0.638103	valid's binary_logloss: 0.651436
Early stopping, best iteration is:
[381]	train's binary_logloss: 0.63871	valid's binary_logloss: 0.650844
bagging, val_score: 0.650231:  30%|###       | 3/10 [00:03<00:07,  1.13s/it][I 2020-09-27 04:47:20,849] Trial 29 finished with value: 0.6508442537589473 and parameters: {'bagging_fraction': 0.4102073189555854, 'bagging_freq': 1}. Best is trial 28 with value: 0.6502306281879949.
bagging, val_score: 0.650231:  30%|###       | 3/10 [00:03<00:07,  1.13s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004944 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.655879	valid's binary_logloss: 0.660227
[200]	train's binary_logloss: 0.646389	valid's binary_logloss: 0.653638
[300]	train's binary_logloss: 0.641982	valid's binary_logloss: 0.652522
[400]	train's binary_logloss: 0.638289	valid's binary_logloss: 0.652467
Early stopping, best iteration is:
[387]	train's binary_logloss: 0.63872	valid's binary_logloss: 0.652011
bagging, val_score: 0.650231:  40%|####      | 4/10 [00:03<00:06,  1.04s/it][I 2020-09-27 04:47:21,683] Trial 30 finished with value: 0.6520113408937251 and parameters: {'bagging_fraction': 0.4058702957626291, 'bagging_freq': 1}. Best is trial 28 with value: 0.6502306281879949.
bagging, val_score: 0.650231:  40%|####      | 4/10 [00:03<00:06,  1.04s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008559 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656756	valid's binary_logloss: 0.662314
[200]	train's binary_logloss: 0.646612	valid's binary_logloss: 0.653269
[300]	train's binary_logloss: 0.641725	valid's binary_logloss: 0.650745
[400]	train's binary_logloss: 0.638127	valid's binary_logloss: 0.650437
Early stopping, best iteration is:
[386]	train's binary_logloss: 0.638596	valid's binary_logloss: 0.649682
bagging, val_score: 0.649682:  50%|#####     | 5/10 [00:05<00:05,  1.10s/it][I 2020-09-27 04:47:22,921] Trial 31 finished with value: 0.6496818435732342 and parameters: {'bagging_fraction': 0.6494590447691004, 'bagging_freq': 3}. Best is trial 31 with value: 0.6496818435732342.
bagging, val_score: 0.649682:  50%|#####     | 5/10 [00:05<00:05,  1.10s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004698 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.65681	valid's binary_logloss: 0.661741
[200]	train's binary_logloss: 0.646579	valid's binary_logloss: 0.653526
[300]	train's binary_logloss: 0.641414	valid's binary_logloss: 0.651164
[400]	train's binary_logloss: 0.637939	valid's binary_logloss: 0.650273
Early stopping, best iteration is:
[386]	train's binary_logloss: 0.638417	valid's binary_logloss: 0.649953
bagging, val_score: 0.649682:  60%|######    | 6/10 [00:05<00:03,  1.00it/s][I 2020-09-27 04:47:23,670] Trial 32 finished with value: 0.6499526525656477 and parameters: {'bagging_fraction': 0.6912299232252367, 'bagging_freq': 3}. Best is trial 31 with value: 0.6496818435732342.
bagging, val_score: 0.649682:  60%|######    | 6/10 [00:05<00:03,  1.00it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004577 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656917	valid's binary_logloss: 0.662005
[200]	train's binary_logloss: 0.646694	valid's binary_logloss: 0.653638
[300]	train's binary_logloss: 0.641556	valid's binary_logloss: 0.651449
[400]	train's binary_logloss: 0.637865	valid's binary_logloss: 0.650759
Early stopping, best iteration is:
[387]	train's binary_logloss: 0.638305	valid's binary_logloss: 0.650278
bagging, val_score: 0.649682:  70%|#######   | 7/10 [00:06<00:02,  1.09it/s][I 2020-09-27 04:47:24,413] Trial 33 finished with value: 0.6502781308682245 and parameters: {'bagging_fraction': 0.7006917452683191, 'bagging_freq': 3}. Best is trial 31 with value: 0.6496818435732342.
bagging, val_score: 0.649682:  70%|#######   | 7/10 [00:06<00:02,  1.09it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005647 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656936	valid's binary_logloss: 0.66276
[200]	train's binary_logloss: 0.646675	valid's binary_logloss: 0.653779
[300]	train's binary_logloss: 0.641779	valid's binary_logloss: 0.651157
[400]	train's binary_logloss: 0.638073	valid's binary_logloss: 0.650615
Early stopping, best iteration is:
[385]	train's binary_logloss: 0.638614	valid's binary_logloss: 0.649988
bagging, val_score: 0.649682:  80%|########  | 8/10 [00:07<00:01,  1.14it/s][I 2020-09-27 04:47:25,180] Trial 34 finished with value: 0.6499876467524935 and parameters: {'bagging_fraction': 0.6697359412743568, 'bagging_freq': 3}. Best is trial 31 with value: 0.6496818435732342.
bagging, val_score: 0.649682:  80%|########  | 8/10 [00:07<00:01,  1.14it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005015 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656928	valid's binary_logloss: 0.661654
[200]	train's binary_logloss: 0.646722	valid's binary_logloss: 0.652579
[300]	train's binary_logloss: 0.641759	valid's binary_logloss: 0.65045
[400]	train's binary_logloss: 0.638172	valid's binary_logloss: 0.649763
[500]	train's binary_logloss: 0.634789	valid's binary_logloss: 0.649596
Early stopping, best iteration is:
[472]	train's binary_logloss: 0.635712	valid's binary_logloss: 0.649063
bagging, val_score: 0.649063:  90%|######### | 9/10 [00:08<00:01,  1.00s/it][I 2020-09-27 04:47:26,487] Trial 35 finished with value: 0.649063213862809 and parameters: {'bagging_fraction': 0.6949061090676492, 'bagging_freq': 3}. Best is trial 35 with value: 0.649063213862809.
bagging, val_score: 0.649063:  90%|######### | 9/10 [00:08<00:01,  1.00s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.006251 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656842	valid's binary_logloss: 0.662647
[200]	train's binary_logloss: 0.646752	valid's binary_logloss: 0.654227
[300]	train's binary_logloss: 0.641722	valid's binary_logloss: 0.65187
[400]	train's binary_logloss: 0.638055	valid's binary_logloss: 0.651604
[500]	train's binary_logloss: 0.634682	valid's binary_logloss: 0.651328
Early stopping, best iteration is:
[464]	train's binary_logloss: 0.63593	valid's binary_logloss: 0.650751
bagging, val_score: 0.649063: 100%|##########| 10/10 [00:09<00:00,  1.01s/it][I 2020-09-27 04:47:27,517] Trial 36 finished with value: 0.6507510423009876 and parameters: {'bagging_fraction': 0.6850010940586274, 'bagging_freq': 3}. Best is trial 35 with value: 0.649063213862809.
bagging, val_score: 0.649063: 100%|##########| 10/10 [00:09<00:00,  1.02it/s]
feature_fraction_stage2, val_score: 0.649063:   0%|          | 0/6 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005016 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656928	valid's binary_logloss: 0.661654
[200]	train's binary_logloss: 0.646722	valid's binary_logloss: 0.652579
[300]	train's binary_logloss: 0.641759	valid's binary_logloss: 0.65045
[400]	train's binary_logloss: 0.638172	valid's binary_logloss: 0.649763
[500]	train's binary_logloss: 0.634789	valid's binary_logloss: 0.649596
Early stopping, best iteration is:
[472]	train's binary_logloss: 0.635712	valid's binary_logloss: 0.649063
feature_fraction_stage2, val_score: 0.649063:  17%|#6        | 1/6 [00:00<00:04,  1.11it/s][I 2020-09-27 04:47:28,434] Trial 37 finished with value: 0.649063213862809 and parameters: {'feature_fraction': 0.616}. Best is trial 37 with value: 0.649063213862809.
feature_fraction_stage2, val_score: 0.649063:  17%|#6        | 1/6 [00:00<00:04,  1.11it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004551 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656692	valid's binary_logloss: 0.661077
[200]	train's binary_logloss: 0.646429	valid's binary_logloss: 0.652636
[300]	train's binary_logloss: 0.641474	valid's binary_logloss: 0.650303
[400]	train's binary_logloss: 0.637696	valid's binary_logloss: 0.649448
[500]	train's binary_logloss: 0.634319	valid's binary_logloss: 0.64952
Early stopping, best iteration is:
[468]	train's binary_logloss: 0.635429	valid's binary_logloss: 0.648813
feature_fraction_stage2, val_score: 0.648813:  33%|###3      | 2/6 [00:01<00:03,  1.11it/s][I 2020-09-27 04:47:29,319] Trial 38 finished with value: 0.6488132191521305 and parameters: {'feature_fraction': 0.6479999999999999}. Best is trial 38 with value: 0.6488132191521305.
feature_fraction_stage2, val_score: 0.648813:  33%|###3      | 2/6 [00:01<00:03,  1.11it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000576 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657062	valid's binary_logloss: 0.661671
[200]	train's binary_logloss: 0.647032	valid's binary_logloss: 0.653351
[300]	train's binary_logloss: 0.641928	valid's binary_logloss: 0.650494
[400]	train's binary_logloss: 0.638384	valid's binary_logloss: 0.649678
[500]	train's binary_logloss: 0.634981	valid's binary_logloss: 0.649656
[600]	train's binary_logloss: 0.631952	valid's binary_logloss: 0.649533
Early stopping, best iteration is:
[551]	train's binary_logloss: 0.633314	valid's binary_logloss: 0.649149
feature_fraction_stage2, val_score: 0.648813:  50%|#####     | 3/6 [00:03<00:03,  1.06s/it][I 2020-09-27 04:47:30,759] Trial 39 finished with value: 0.6491494764152731 and parameters: {'feature_fraction': 0.552}. Best is trial 38 with value: 0.6488132191521305.
feature_fraction_stage2, val_score: 0.648813:  50%|#####     | 3/6 [00:03<00:03,  1.06s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000576 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656892	valid's binary_logloss: 0.661978
[200]	train's binary_logloss: 0.646753	valid's binary_logloss: 0.653932
[300]	train's binary_logloss: 0.641708	valid's binary_logloss: 0.652504
[400]	train's binary_logloss: 0.638016	valid's binary_logloss: 0.651
[500]	train's binary_logloss: 0.634749	valid's binary_logloss: 0.651944
Early stopping, best iteration is:
[412]	train's binary_logloss: 0.637606	valid's binary_logloss: 0.65077
feature_fraction_stage2, val_score: 0.648813:  67%|######6   | 4/6 [00:04<00:02,  1.00s/it][I 2020-09-27 04:47:31,621] Trial 40 finished with value: 0.6507703781531419 and parameters: {'feature_fraction': 0.584}. Best is trial 38 with value: 0.6488132191521305.
feature_fraction_stage2, val_score: 0.648813:  67%|######6   | 4/6 [00:04<00:02,  1.00s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004974 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661816
[200]	train's binary_logloss: 0.646462	valid's binary_logloss: 0.653161
[300]	train's binary_logloss: 0.641483	valid's binary_logloss: 0.650694
[400]	train's binary_logloss: 0.6377	valid's binary_logloss: 0.649329
[500]	train's binary_logloss: 0.634322	valid's binary_logloss: 0.649442
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635675	valid's binary_logloss: 0.648669
feature_fraction_stage2, val_score: 0.648669:  83%|########3 | 5/6 [00:05<00:00,  1.01it/s][I 2020-09-27 04:47:32,586] Trial 41 finished with value: 0.6486693704765705 and parameters: {'feature_fraction': 0.6799999999999999}. Best is trial 41 with value: 0.6486693704765705.
feature_fraction_stage2, val_score: 0.648669:  83%|########3 | 5/6 [00:05<00:00,  1.01it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004631 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657062	valid's binary_logloss: 0.661671
[200]	train's binary_logloss: 0.647032	valid's binary_logloss: 0.653351
[300]	train's binary_logloss: 0.641928	valid's binary_logloss: 0.650494
[400]	train's binary_logloss: 0.638384	valid's binary_logloss: 0.649678
[500]	train's binary_logloss: 0.634981	valid's binary_logloss: 0.649656
[600]	train's binary_logloss: 0.631952	valid's binary_logloss: 0.649533
Early stopping, best iteration is:
[551]	train's binary_logloss: 0.633314	valid's binary_logloss: 0.649149
feature_fraction_stage2, val_score: 0.648669: 100%|##########| 6/6 [00:06<00:00,  1.11s/it][I 2020-09-27 04:47:33,975] Trial 42 finished with value: 0.6491494764152731 and parameters: {'feature_fraction': 0.52}. Best is trial 41 with value: 0.6486693704765705.
feature_fraction_stage2, val_score: 0.648669: 100%|##########| 6/6 [00:06<00:00,  1.08s/it]
regularization_factors, val_score: 0.648669:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012670 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656679	valid's binary_logloss: 0.661907
[200]	train's binary_logloss: 0.646312	valid's binary_logloss: 0.653757
[300]	train's binary_logloss: 0.641354	valid's binary_logloss: 0.651207
[400]	train's binary_logloss: 0.637772	valid's binary_logloss: 0.650071
[500]	train's binary_logloss: 0.634407	valid's binary_logloss: 0.649928
Early stopping, best iteration is:
[473]	train's binary_logloss: 0.635291	valid's binary_logloss: 0.649607
regularization_factors, val_score: 0.648669:   5%|5         | 1/20 [00:01<00:20,  1.08s/it][I 2020-09-27 04:47:35,081] Trial 43 finished with value: 0.6496074604263665 and parameters: {'lambda_l1': 4.193685786846261e-08, 'lambda_l2': 0.3229663189729637}. Best is trial 43 with value: 0.6496074604263665.
regularization_factors, val_score: 0.648669:   5%|5         | 1/20 [00:01<00:20,  1.08s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.011232 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657809	valid's binary_logloss: 0.663236
[200]	train's binary_logloss: 0.647792	valid's binary_logloss: 0.654479
[300]	train's binary_logloss: 0.643617	valid's binary_logloss: 0.650947
[400]	train's binary_logloss: 0.640751	valid's binary_logloss: 0.650509
[500]	train's binary_logloss: 0.638188	valid's binary_logloss: 0.650185
Early stopping, best iteration is:
[464]	train's binary_logloss: 0.639104	valid's binary_logloss: 0.649722
regularization_factors, val_score: 0.648669:  10%|#         | 2/20 [00:01<00:18,  1.03s/it][I 2020-09-27 04:47:35,996] Trial 44 finished with value: 0.6497218154257693 and parameters: {'lambda_l1': 7.661272890098566, 'lambda_l2': 1.6647681314940023e-07}. Best is trial 43 with value: 0.6496074604263665.
regularization_factors, val_score: 0.648669:  10%|#         | 2/20 [00:02<00:18,  1.03s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005094 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656808	valid's binary_logloss: 0.661435
[200]	train's binary_logloss: 0.646494	valid's binary_logloss: 0.653055
[300]	train's binary_logloss: 0.641675	valid's binary_logloss: 0.650859
[400]	train's binary_logloss: 0.637945	valid's binary_logloss: 0.649955
[500]	train's binary_logloss: 0.634406	valid's binary_logloss: 0.649662
Early stopping, best iteration is:
[469]	train's binary_logloss: 0.635491	valid's binary_logloss: 0.649128
regularization_factors, val_score: 0.648669:  15%|#5        | 3/20 [00:02<00:17,  1.01s/it][I 2020-09-27 04:47:36,939] Trial 45 finished with value: 0.6491279609766097 and parameters: {'lambda_l1': 0.038151596318754866, 'lambda_l2': 1.3493847157523479e-08}. Best is trial 45 with value: 0.6491279609766097.
regularization_factors, val_score: 0.648669:  15%|#5        | 3/20 [00:02<00:17,  1.01s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000809 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656763	valid's binary_logloss: 0.661526
[200]	train's binary_logloss: 0.646359	valid's binary_logloss: 0.653842
[300]	train's binary_logloss: 0.641324	valid's binary_logloss: 0.650914
[400]	train's binary_logloss: 0.637727	valid's binary_logloss: 0.650544
[500]	train's binary_logloss: 0.634443	valid's binary_logloss: 0.65063
Early stopping, best iteration is:
[457]	train's binary_logloss: 0.635863	valid's binary_logloss: 0.649897
regularization_factors, val_score: 0.648669:  20%|##        | 4/20 [00:04<00:17,  1.12s/it][I 2020-09-27 04:47:38,329] Trial 46 finished with value: 0.6498965604608956 and parameters: {'lambda_l1': 0.061724342064673816, 'lambda_l2': 1.1191098300700474e-08}. Best is trial 45 with value: 0.6491279609766097.
regularization_factors, val_score: 0.648669:  20%|##        | 4/20 [00:04<00:17,  1.12s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005050 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661816
[200]	train's binary_logloss: 0.646462	valid's binary_logloss: 0.653161
[300]	train's binary_logloss: 0.641483	valid's binary_logloss: 0.650694
[400]	train's binary_logloss: 0.637704	valid's binary_logloss: 0.649345
[500]	train's binary_logloss: 0.634326	valid's binary_logloss: 0.649458
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635679	valid's binary_logloss: 0.648685
regularization_factors, val_score: 0.648669:  25%|##5       | 5/20 [00:05<00:16,  1.09s/it][I 2020-09-27 04:47:39,333] Trial 47 finished with value: 0.6486854676681457 and parameters: {'lambda_l1': 7.797235764236721e-05, 'lambda_l2': 4.7617608412875624e-05}. Best is trial 47 with value: 0.6486854676681457.
regularization_factors, val_score: 0.648669:  25%|##5       | 5/20 [00:05<00:16,  1.09s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004751 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661816
[200]	train's binary_logloss: 0.646462	valid's binary_logloss: 0.653161
[300]	train's binary_logloss: 0.641483	valid's binary_logloss: 0.650694
[400]	train's binary_logloss: 0.6377	valid's binary_logloss: 0.649329
[500]	train's binary_logloss: 0.634322	valid's binary_logloss: 0.649442
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635675	valid's binary_logloss: 0.648669
regularization_factors, val_score: 0.648669:  30%|###       | 6/20 [00:06<00:14,  1.03s/it][I 2020-09-27 04:47:40,228] Trial 48 finished with value: 0.6486693564289175 and parameters: {'lambda_l1': 4.6907860757180705e-07, 'lambda_l2': 0.0001968190366629125}. Best is trial 48 with value: 0.6486693564289175.
regularization_factors, val_score: 0.648669:  30%|###       | 6/20 [00:06<00:14,  1.03s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000809 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661816
[200]	train's binary_logloss: 0.646461	valid's binary_logloss: 0.653161
[300]	train's binary_logloss: 0.641483	valid's binary_logloss: 0.650694
[400]	train's binary_logloss: 0.637703	valid's binary_logloss: 0.649345
[500]	train's binary_logloss: 0.634325	valid's binary_logloss: 0.649458
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635679	valid's binary_logloss: 0.648685
regularization_factors, val_score: 0.648669:  35%|###5      | 7/20 [00:07<00:12,  1.01it/s][I 2020-09-27 04:47:41,143] Trial 49 finished with value: 0.6486853650384133 and parameters: {'lambda_l1': 7.106958303444078e-07, 'lambda_l2': 0.00031503172943699124}. Best is trial 48 with value: 0.6486693564289175.
regularization_factors, val_score: 0.648669:  35%|###5      | 7/20 [00:07<00:12,  1.01it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004879 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661816
[200]	train's binary_logloss: 0.646462	valid's binary_logloss: 0.653161
[300]	train's binary_logloss: 0.641483	valid's binary_logloss: 0.650694
[400]	train's binary_logloss: 0.637704	valid's binary_logloss: 0.649345
[500]	train's binary_logloss: 0.634326	valid's binary_logloss: 0.649458
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635679	valid's binary_logloss: 0.648685
regularization_factors, val_score: 0.648669:  40%|####      | 8/20 [00:08<00:13,  1.09s/it][I 2020-09-27 04:47:42,447] Trial 50 finished with value: 0.6486854411094597 and parameters: {'lambda_l1': 4.4792617037779606e-07, 'lambda_l2': 0.00040119958572656174}. Best is trial 48 with value: 0.6486693564289175.
regularization_factors, val_score: 0.648669:  40%|####      | 8/20 [00:08<00:13,  1.09s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000826 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661816
[200]	train's binary_logloss: 0.646462	valid's binary_logloss: 0.653161
[300]	train's binary_logloss: 0.641483	valid's binary_logloss: 0.650694
[400]	train's binary_logloss: 0.6377	valid's binary_logloss: 0.649329
[500]	train's binary_logloss: 0.634322	valid's binary_logloss: 0.649442
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635675	valid's binary_logloss: 0.648669
regularization_factors, val_score: 0.648669:  45%|####5     | 9/20 [00:09<00:11,  1.05s/it][I 2020-09-27 04:47:43,405] Trial 51 finished with value: 0.6486693484397528 and parameters: {'lambda_l1': 3.4434140107307795e-07, 'lambda_l2': 0.00030839964549718207}. Best is trial 51 with value: 0.6486693484397528.
regularization_factors, val_score: 0.648669:  45%|####5     | 9/20 [00:09<00:11,  1.05s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000859 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661816
[200]	train's binary_logloss: 0.646462	valid's binary_logloss: 0.653161
[300]	train's binary_logloss: 0.641483	valid's binary_logloss: 0.650694
[400]	train's binary_logloss: 0.637704	valid's binary_logloss: 0.649345
[500]	train's binary_logloss: 0.634326	valid's binary_logloss: 0.649458
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635679	valid's binary_logloss: 0.648685
regularization_factors, val_score: 0.648669:  50%|#####     | 10/20 [00:10<00:10,  1.01s/it][I 2020-09-27 04:47:44,335] Trial 52 finished with value: 0.6486854548458479 and parameters: {'lambda_l1': 2.9104522149216847e-07, 'lambda_l2': 0.00021910741939920448}. Best is trial 51 with value: 0.6486693484397528.
regularization_factors, val_score: 0.648669:  50%|#####     | 10/20 [00:10<00:10,  1.01s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.009619 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661816
[200]	train's binary_logloss: 0.646462	valid's binary_logloss: 0.653161
[300]	train's binary_logloss: 0.641483	valid's binary_logloss: 0.650694
[400]	train's binary_logloss: 0.637704	valid's binary_logloss: 0.649345
[500]	train's binary_logloss: 0.634326	valid's binary_logloss: 0.649458
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635679	valid's binary_logloss: 0.648685
regularization_factors, val_score: 0.648669:  55%|#####5    | 11/20 [00:11<00:09,  1.03s/it][I 2020-09-27 04:47:45,395] Trial 53 finished with value: 0.6486854423114767 and parameters: {'lambda_l1': 1.8051040240033813e-07, 'lambda_l2': 0.0003854090838672508}. Best is trial 51 with value: 0.6486693484397528.
regularization_factors, val_score: 0.648669:  55%|#####5    | 11/20 [00:11<00:09,  1.03s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.002587 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661816
[200]	train's binary_logloss: 0.646462	valid's binary_logloss: 0.653161
[300]	train's binary_logloss: 0.641483	valid's binary_logloss: 0.650694
[400]	train's binary_logloss: 0.637704	valid's binary_logloss: 0.649345
[500]	train's binary_logloss: 0.634326	valid's binary_logloss: 0.649458
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635679	valid's binary_logloss: 0.648685
regularization_factors, val_score: 0.648669:  60%|######    | 12/20 [00:12<00:08,  1.11s/it][I 2020-09-27 04:47:46,692] Trial 54 finished with value: 0.6486853768130912 and parameters: {'lambda_l1': 1.238928167433891e-06, 'lambda_l2': 0.001256677247104136}. Best is trial 51 with value: 0.6486693484397528.
regularization_factors, val_score: 0.648669:  60%|######    | 12/20 [00:12<00:08,  1.11s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.009674 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661816
[200]	train's binary_logloss: 0.646462	valid's binary_logloss: 0.653161
[300]	train's binary_logloss: 0.641484	valid's binary_logloss: 0.650694
[400]	train's binary_logloss: 0.637701	valid's binary_logloss: 0.649329
[500]	train's binary_logloss: 0.634323	valid's binary_logloss: 0.649442
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635676	valid's binary_logloss: 0.648669
regularization_factors, val_score: 0.648669:  65%|######5   | 13/20 [00:13<00:07,  1.05s/it][I 2020-09-27 04:47:47,595] Trial 55 finished with value: 0.6486691459174234 and parameters: {'lambda_l1': 4.463187899572695e-06, 'lambda_l2': 0.0031440113191708815}. Best is trial 55 with value: 0.6486691459174234.
regularization_factors, val_score: 0.648669:  65%|######5   | 13/20 [00:13<00:07,  1.05s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004983 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661816
[200]	train's binary_logloss: 0.646463	valid's binary_logloss: 0.653161
[300]	train's binary_logloss: 0.641496	valid's binary_logloss: 0.650455
[400]	train's binary_logloss: 0.637726	valid's binary_logloss: 0.649177
[500]	train's binary_logloss: 0.634323	valid's binary_logloss: 0.649235
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635717	valid's binary_logloss: 0.648475
regularization_factors, val_score: 0.648475:  70%|#######   | 14/20 [00:14<00:06,  1.00s/it][I 2020-09-27 04:47:48,497] Trial 56 finished with value: 0.6484745400508054 and parameters: {'lambda_l1': 9.72103845454671e-06, 'lambda_l2': 0.011377674391625028}. Best is trial 56 with value: 0.6484745400508054.
regularization_factors, val_score: 0.648475:  70%|#######   | 14/20 [00:14<00:06,  1.00s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001108 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656804	valid's binary_logloss: 0.661432
[200]	train's binary_logloss: 0.646512	valid's binary_logloss: 0.653055
[300]	train's binary_logloss: 0.641411	valid's binary_logloss: 0.650947
[400]	train's binary_logloss: 0.637723	valid's binary_logloss: 0.649987
[500]	train's binary_logloss: 0.634326	valid's binary_logloss: 0.650052
Early stopping, best iteration is:
[464]	train's binary_logloss: 0.635548	valid's binary_logloss: 0.649488
regularization_factors, val_score: 0.648475:  75%|#######5  | 15/20 [00:15<00:05,  1.08s/it][I 2020-09-27 04:47:49,768] Trial 57 finished with value: 0.6494877361955586 and parameters: {'lambda_l1': 1.3431278103969494e-05, 'lambda_l2': 0.02447854418046466}. Best is trial 56 with value: 0.6484745400508054.
regularization_factors, val_score: 0.648475:  75%|#######5  | 15/20 [00:15<00:05,  1.08s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.015062 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661816
[200]	train's binary_logloss: 0.646462	valid's binary_logloss: 0.653161
[300]	train's binary_logloss: 0.641495	valid's binary_logloss: 0.650455
[400]	train's binary_logloss: 0.637725	valid's binary_logloss: 0.649177
[500]	train's binary_logloss: 0.634322	valid's binary_logloss: 0.649235
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635715	valid's binary_logloss: 0.648475
regularization_factors, val_score: 0.648475:  80%|########  | 16/20 [00:16<00:04,  1.08s/it][I 2020-09-27 04:47:50,844] Trial 58 finished with value: 0.6484747613620405 and parameters: {'lambda_l1': 5.558047158433351e-06, 'lambda_l2': 0.007068676788026479}. Best is trial 56 with value: 0.6484745400508054.
regularization_factors, val_score: 0.648475:  80%|########  | 16/20 [00:16<00:04,  1.08s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000877 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661817
[200]	train's binary_logloss: 0.646463	valid's binary_logloss: 0.653162
[300]	train's binary_logloss: 0.641497	valid's binary_logloss: 0.650455
[400]	train's binary_logloss: 0.637728	valid's binary_logloss: 0.649176
[500]	train's binary_logloss: 0.634326	valid's binary_logloss: 0.649234
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635719	valid's binary_logloss: 0.648474
regularization_factors, val_score: 0.648474:  85%|########5 | 17/20 [00:17<00:03,  1.09s/it][I 2020-09-27 04:47:51,940] Trial 59 finished with value: 0.6484741879925516 and parameters: {'lambda_l1': 1.1402866509712536e-05, 'lambda_l2': 0.018245531549826995}. Best is trial 59 with value: 0.6484741879925516.
regularization_factors, val_score: 0.648474:  85%|########5 | 17/20 [00:17<00:03,  1.09s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004790 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661817
[200]	train's binary_logloss: 0.646463	valid's binary_logloss: 0.653162
[300]	train's binary_logloss: 0.641497	valid's binary_logloss: 0.650455
[400]	train's binary_logloss: 0.637728	valid's binary_logloss: 0.649176
[500]	train's binary_logloss: 0.634326	valid's binary_logloss: 0.649234
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635719	valid's binary_logloss: 0.648474
regularization_factors, val_score: 0.648474:  90%|######### | 18/20 [00:18<00:02,  1.03s/it][I 2020-09-27 04:47:52,825] Trial 60 finished with value: 0.6484741937796707 and parameters: {'lambda_l1': 1.9414476880011605e-05, 'lambda_l2': 0.018137780474702055}. Best is trial 59 with value: 0.6484741879925516.
regularization_factors, val_score: 0.648474:  90%|######### | 18/20 [00:18<00:02,  1.03s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004872 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656759	valid's binary_logloss: 0.661816
[200]	train's binary_logloss: 0.646463	valid's binary_logloss: 0.653161
[300]	train's binary_logloss: 0.641496	valid's binary_logloss: 0.650455
[400]	train's binary_logloss: 0.637727	valid's binary_logloss: 0.649177
[500]	train's binary_logloss: 0.634324	valid's binary_logloss: 0.649235
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.635717	valid's binary_logloss: 0.648474
regularization_factors, val_score: 0.648474:  95%|#########5| 19/20 [00:20<00:01,  1.14s/it][I 2020-09-27 04:47:54,216] Trial 61 finished with value: 0.6484744650230337 and parameters: {'lambda_l1': 1.6059055968360456e-05, 'lambda_l2': 0.012844094468385317}. Best is trial 59 with value: 0.6484741879925516.
regularization_factors, val_score: 0.648474:  95%|#########5| 19/20 [00:20<00:01,  1.14s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004738 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.65676	valid's binary_logloss: 0.661817
[200]	train's binary_logloss: 0.646463	valid's binary_logloss: 0.653162
[300]	train's binary_logloss: 0.641498	valid's binary_logloss: 0.650454
[400]	train's binary_logloss: 0.637729	valid's binary_logloss: 0.649176
[500]	train's binary_logloss: 0.634329	valid's binary_logloss: 0.649256
Early stopping, best iteration is:
[458]	train's binary_logloss: 0.63572	valid's binary_logloss: 0.648474
regularization_factors, val_score: 0.648474: 100%|##########| 20/20 [00:21<00:00,  1.09s/it][I 2020-09-27 04:47:55,185] Trial 62 finished with value: 0.6484739197241262 and parameters: {'lambda_l1': 1.4740300654925883e-05, 'lambda_l2': 0.02349270671135121}. Best is trial 62 with value: 0.6484739197241262.
regularization_factors, val_score: 0.648474: 100%|##########| 20/20 [00:21<00:00,  1.06s/it]
min_data_in_leaf, val_score: 0.648474:   0%|          | 0/5 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000891 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.65676	valid's binary_logloss: 0.661817
[200]	train's binary_logloss: 0.646525	valid's binary_logloss: 0.653595
[300]	train's binary_logloss: 0.641846	valid's binary_logloss: 0.650251
[400]	train's binary_logloss: 0.638553	valid's binary_logloss: 0.649577
[500]	train's binary_logloss: 0.63552	valid's binary_logloss: 0.64935
Early stopping, best iteration is:
[464]	train's binary_logloss: 0.636644	valid's binary_logloss: 0.648879
min_data_in_leaf, val_score: 0.648474:  20%|##        | 1/5 [00:01<00:04,  1.01s/it][I 2020-09-27 04:47:56,208] Trial 63 finished with value: 0.648879477038123 and parameters: {'min_child_samples': 100}. Best is trial 63 with value: 0.648879477038123.
min_data_in_leaf, val_score: 0.648474:  20%|##        | 1/5 [00:01<00:04,  1.01s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.010338 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.65676	valid's binary_logloss: 0.661817
[200]	train's binary_logloss: 0.646463	valid's binary_logloss: 0.653162
[300]	train's binary_logloss: 0.6415	valid's binary_logloss: 0.650452
[400]	train's binary_logloss: 0.637791	valid's binary_logloss: 0.649375
[500]	train's binary_logloss: 0.634508	valid's binary_logloss: 0.649431
Early stopping, best iteration is:
[451]	train's binary_logloss: 0.636075	valid's binary_logloss: 0.648808
min_data_in_leaf, val_score: 0.648474:  40%|####      | 2/5 [00:02<00:03,  1.03s/it][I 2020-09-27 04:47:57,286] Trial 64 finished with value: 0.6488082911033146 and parameters: {'min_child_samples': 25}. Best is trial 64 with value: 0.6488082911033146.
min_data_in_leaf, val_score: 0.648474:  40%|####      | 2/5 [00:02<00:03,  1.03s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.015161 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.65676	valid's binary_logloss: 0.661817
[200]	train's binary_logloss: 0.646493	valid's binary_logloss: 0.65325
[300]	train's binary_logloss: 0.641497	valid's binary_logloss: 0.650822
[400]	train's binary_logloss: 0.637788	valid's binary_logloss: 0.649913
[500]	train's binary_logloss: 0.634246	valid's binary_logloss: 0.649666
Early stopping, best iteration is:
[468]	train's binary_logloss: 0.635328	valid's binary_logloss: 0.649123
min_data_in_leaf, val_score: 0.648474:  60%|######    | 3/5 [00:03<00:02,  1.10s/it][I 2020-09-27 04:47:58,535] Trial 65 finished with value: 0.649122937898972 and parameters: {'min_child_samples': 10}. Best is trial 64 with value: 0.6488082911033146.
min_data_in_leaf, val_score: 0.648474:  60%|######    | 3/5 [00:03<00:02,  1.10s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001909 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.65676	valid's binary_logloss: 0.661817
[200]	train's binary_logloss: 0.64648	valid's binary_logloss: 0.653437
[300]	train's binary_logloss: 0.641645	valid's binary_logloss: 0.650437
[400]	train's binary_logloss: 0.638049	valid's binary_logloss: 0.649312
[500]	train's binary_logloss: 0.634894	valid's binary_logloss: 0.649312
Early stopping, best iteration is:
[447]	train's binary_logloss: 0.63654	valid's binary_logloss: 0.648959
min_data_in_leaf, val_score: 0.648474:  80%|########  | 4/5 [00:04<00:01,  1.04s/it][I 2020-09-27 04:47:59,458] Trial 66 finished with value: 0.6489591737909831 and parameters: {'min_child_samples': 50}. Best is trial 64 with value: 0.6488082911033146.
min_data_in_leaf, val_score: 0.648474:  80%|########  | 4/5 [00:04<00:01,  1.04s/it][LightGBM] [Info] Number of positive: 12855, number of negative: 13145
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001493 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4242
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494423 -> initscore=-0.022309
[LightGBM] [Info] Start training from score -0.022309
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.65676	valid's binary_logloss: 0.661817
[200]	train's binary_logloss: 0.646493	valid's binary_logloss: 0.65325
[300]	train's binary_logloss: 0.64151	valid's binary_logloss: 0.65102
[400]	train's binary_logloss: 0.637718	valid's binary_logloss: 0.650205
[500]	train's binary_logloss: 0.634224	valid's binary_logloss: 0.649812
Early stopping, best iteration is:
[459]	train's binary_logloss: 0.635631	valid's binary_logloss: 0.649459
min_data_in_leaf, val_score: 0.648474: 100%|##########| 5/5 [00:05<00:00,  1.02s/it][I 2020-09-27 04:48:00,406] Trial 67 finished with value: 0.6494585505180116 and parameters: {'min_child_samples': 5}. Best is trial 64 with value: 0.6488082911033146.
min_data_in_leaf, val_score: 0.648474: 100%|##########| 5/5 [00:05<00:00,  1.04s/it]
Fold : 9
[I 2020-09-27 04:48:00,469] A new study created in memory with name: no-name-7b517b46-d774-4255-8991-4b95254b4cf4
feature_fraction, val_score: inf:   0%|          | 0/7 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000996 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.571415	valid's binary_logloss: 0.664205
Early stopping, best iteration is:
[56]	train's binary_logloss: 0.602615	valid's binary_logloss: 0.661775
feature_fraction, val_score: 0.661775:  14%|#4        | 1/7 [00:00<00:05,  1.12it/s][I 2020-09-27 04:48:01,375] Trial 0 finished with value: 0.6617754129584598 and parameters: {'feature_fraction': 0.8999999999999999}. Best is trial 0 with value: 0.6617754129584598.
feature_fraction, val_score: 0.661775:  14%|#4        | 1/7 [00:00<00:05,  1.12it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001704 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.577907	valid's binary_logloss: 0.66332
Early stopping, best iteration is:
[71]	train's binary_logloss: 0.596884	valid's binary_logloss: 0.662116
feature_fraction, val_score: 0.661775:  29%|##8       | 2/7 [00:02<00:04,  1.04it/s][I 2020-09-27 04:48:02,518] Trial 1 finished with value: 0.6621160574043524 and parameters: {'feature_fraction': 0.5}. Best is trial 0 with value: 0.6617754129584598.
feature_fraction, val_score: 0.661775:  29%|##8       | 2/7 [00:02<00:04,  1.04it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000446 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.581403	valid's binary_logloss: 0.664323
Early stopping, best iteration is:
[69]	train's binary_logloss: 0.601689	valid's binary_logloss: 0.663133
feature_fraction, val_score: 0.661775:  43%|####2     | 3/7 [00:02<00:03,  1.21it/s][I 2020-09-27 04:48:03,026] Trial 2 finished with value: 0.6631333345540273 and parameters: {'feature_fraction': 0.4}. Best is trial 0 with value: 0.6617754129584598.
feature_fraction, val_score: 0.661775:  43%|####2     | 3/7 [00:02<00:03,  1.21it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000926 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.569848	valid's binary_logloss: 0.664707
Early stopping, best iteration is:
[67]	train's binary_logloss: 0.59321	valid's binary_logloss: 0.662756
feature_fraction, val_score: 0.661775:  57%|#####7    | 4/7 [00:03<00:02,  1.35it/s][I 2020-09-27 04:48:03,555] Trial 3 finished with value: 0.6627562719479632 and parameters: {'feature_fraction': 1.0}. Best is trial 0 with value: 0.6617754129584598.
feature_fraction, val_score: 0.661775:  57%|#####7    | 4/7 [00:03<00:02,  1.35it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000814 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.573306	valid's binary_logloss: 0.661877
[200]	train's binary_logloss: 0.517495	valid's binary_logloss: 0.664844
Early stopping, best iteration is:
[113]	train's binary_logloss: 0.565412	valid's binary_logloss: 0.660855
feature_fraction, val_score: 0.660855:  71%|#######1  | 5/7 [00:03<00:01,  1.42it/s][I 2020-09-27 04:48:04,180] Trial 4 finished with value: 0.6608549078376256 and parameters: {'feature_fraction': 0.8}. Best is trial 4 with value: 0.6608549078376256.
feature_fraction, val_score: 0.660855:  71%|#######1  | 5/7 [00:03<00:01,  1.42it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004908 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.574579	valid's binary_logloss: 0.662991
Early stopping, best iteration is:
[54]	train's binary_logloss: 0.606825	valid's binary_logloss: 0.661109
feature_fraction, val_score: 0.660855:  86%|########5 | 6/7 [00:04<00:00,  1.58it/s][I 2020-09-27 04:48:04,649] Trial 5 finished with value: 0.66110942300072 and parameters: {'feature_fraction': 0.7}. Best is trial 4 with value: 0.6608549078376256.
feature_fraction, val_score: 0.660855:  86%|########5 | 6/7 [00:04<00:00,  1.58it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004781 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.575121	valid's binary_logloss: 0.665199
Early stopping, best iteration is:
[58]	train's binary_logloss: 0.604585	valid's binary_logloss: 0.663434
feature_fraction, val_score: 0.660855: 100%|##########| 7/7 [00:04<00:00,  1.58it/s][I 2020-09-27 04:48:05,278] Trial 6 finished with value: 0.6634341152776219 and parameters: {'feature_fraction': 0.6}. Best is trial 4 with value: 0.6608549078376256.
feature_fraction, val_score: 0.660855: 100%|##########| 7/7 [00:04<00:00,  1.46it/s]
num_leaves, val_score: 0.660855:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004289 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.573306	valid's binary_logloss: 0.661877
[200]	train's binary_logloss: 0.517495	valid's binary_logloss: 0.664844
Early stopping, best iteration is:
[113]	train's binary_logloss: 0.565412	valid's binary_logloss: 0.660855
num_leaves, val_score: 0.660855:   5%|5         | 1/20 [00:01<00:37,  1.99s/it][I 2020-09-27 04:48:07,289] Trial 7 finished with value: 0.6608549078376256 and parameters: {'num_leaves': 31}. Best is trial 7 with value: 0.6608549078376256.
num_leaves, val_score: 0.660855:   5%|5         | 1/20 [00:02<00:37,  1.99s/it][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.009560 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.291282	valid's binary_logloss: 0.683399
Early stopping, best iteration is:
[25]	train's binary_logloss: 0.515751	valid's binary_logloss: 0.662419
num_leaves, val_score: 0.660855:  10%|#         | 2/20 [00:03<00:32,  1.79s/it][I 2020-09-27 04:48:08,615] Trial 8 finished with value: 0.6624189569421138 and parameters: {'num_leaves': 227}. Best is trial 7 with value: 0.6608549078376256.
num_leaves, val_score: 0.660855:  10%|#         | 2/20 [00:03<00:32,  1.79s/it][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000898 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.633793	valid's binary_logloss: 0.66169
[200]	train's binary_logloss: 0.616176	valid's binary_logloss: 0.662116
Early stopping, best iteration is:
[127]	train's binary_logloss: 0.628472	valid's binary_logloss: 0.661027
num_leaves, val_score: 0.660855:  15%|#5        | 3/20 [00:03<00:24,  1.41s/it][I 2020-09-27 04:48:09,146] Trial 9 finished with value: 0.6610265822259738 and parameters: {'num_leaves': 8}. Best is trial 7 with value: 0.6608549078376256.
num_leaves, val_score: 0.660855:  15%|#5        | 3/20 [00:03<00:24,  1.41s/it][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000937 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.277955	valid's binary_logloss: 0.68283
Early stopping, best iteration is:
[29]	train's binary_logloss: 0.489512	valid's binary_logloss: 0.667966
num_leaves, val_score: 0.660855:  20%|##        | 4/20 [00:05<00:24,  1.52s/it][I 2020-09-27 04:48:10,921] Trial 10 finished with value: 0.6679657755747807 and parameters: {'num_leaves': 241}. Best is trial 7 with value: 0.6608549078376256.
num_leaves, val_score: 0.660855:  20%|##        | 4/20 [00:05<00:24,  1.52s/it][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000991 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.586674	valid's binary_logloss: 0.66365
Early stopping, best iteration is:
[74]	train's binary_logloss: 0.601053	valid's binary_logloss: 0.662634
num_leaves, val_score: 0.660855:  25%|##5       | 5/20 [00:06<00:18,  1.21s/it][I 2020-09-27 04:48:11,417] Trial 11 finished with value: 0.6626335832631651 and parameters: {'num_leaves': 25}. Best is trial 7 with value: 0.6608549078376256.
num_leaves, val_score: 0.660855:  25%|##5       | 5/20 [00:06<00:18,  1.21s/it][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000938 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.437791	valid's binary_logloss: 0.672762
Early stopping, best iteration is:
[35]	train's binary_logloss: 0.555775	valid's binary_logloss: 0.664193
num_leaves, val_score: 0.660855:  30%|###       | 6/20 [00:06<00:15,  1.07s/it][I 2020-09-27 04:48:12,160] Trial 12 finished with value: 0.664192632015902 and parameters: {'num_leaves': 104}. Best is trial 7 with value: 0.6608549078376256.
num_leaves, val_score: 0.660855:  30%|###       | 6/20 [00:06<00:15,  1.07s/it][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000963 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.458686	valid's binary_logloss: 0.668395
Early stopping, best iteration is:
[40]	train's binary_logloss: 0.556167	valid's binary_logloss: 0.664721
num_leaves, val_score: 0.660855:  35%|###5      | 7/20 [00:07<00:12,  1.03it/s][I 2020-09-27 04:48:12,890] Trial 13 finished with value: 0.6647214922013416 and parameters: {'num_leaves': 90}. Best is trial 7 with value: 0.6608549078376256.
num_leaves, val_score: 0.660855:  35%|###5      | 7/20 [00:07<00:12,  1.03it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.009252 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.350518	valid's binary_logloss: 0.684373
Early stopping, best iteration is:
[22]	train's binary_logloss: 0.558627	valid's binary_logloss: 0.66659
num_leaves, val_score: 0.660855:  40%|####      | 8/20 [00:08<00:11,  1.07it/s][I 2020-09-27 04:48:13,742] Trial 14 finished with value: 0.6665899695534304 and parameters: {'num_leaves': 167}. Best is trial 7 with value: 0.6608549078376256.
num_leaves, val_score: 0.660855:  40%|####      | 8/20 [00:08<00:11,  1.07it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004731 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.532376	valid's binary_logloss: 0.669923
Early stopping, best iteration is:
[40]	train's binary_logloss: 0.596466	valid's binary_logloss: 0.665493
num_leaves, val_score: 0.660855:  45%|####5     | 9/20 [00:09<00:10,  1.04it/s][I 2020-09-27 04:48:14,759] Trial 15 finished with value: 0.665493198325232 and parameters: {'num_leaves': 50}. Best is trial 7 with value: 0.6608549078376256.
num_leaves, val_score: 0.660855:  45%|####5     | 9/20 [00:09<00:10,  1.04it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000797 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.350518	valid's binary_logloss: 0.684373
Early stopping, best iteration is:
[22]	train's binary_logloss: 0.558627	valid's binary_logloss: 0.66659
num_leaves, val_score: 0.660855:  50%|#####     | 10/20 [00:10<00:09,  1.04it/s][I 2020-09-27 04:48:15,722] Trial 16 finished with value: 0.6665899695534304 and parameters: {'num_leaves': 167}. Best is trial 7 with value: 0.6608549078376256.
num_leaves, val_score: 0.660855:  50%|#####     | 10/20 [00:10<00:09,  1.04it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004981 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.512177	valid's binary_logloss: 0.666645
Early stopping, best iteration is:
[29]	train's binary_logloss: 0.60465	valid's binary_logloss: 0.662506
num_leaves, val_score: 0.660855:  55%|#####5    | 11/20 [00:10<00:07,  1.20it/s][I 2020-09-27 04:48:16,266] Trial 17 finished with value: 0.6625059680700698 and parameters: {'num_leaves': 60}. Best is trial 7 with value: 0.6608549078376256.
num_leaves, val_score: 0.660855:  55%|#####5    | 11/20 [00:10<00:07,  1.20it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001403 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650402	valid's binary_logloss: 0.665088
[200]	train's binary_logloss: 0.639194	valid's binary_logloss: 0.660755
[300]	train's binary_logloss: 0.632319	valid's binary_logloss: 0.66028
[400]	train's binary_logloss: 0.626788	valid's binary_logloss: 0.660425
Early stopping, best iteration is:
[360]	train's binary_logloss: 0.628825	valid's binary_logloss: 0.659947
num_leaves, val_score: 0.659947:  60%|######    | 12/20 [00:11<00:06,  1.22it/s][I 2020-09-27 04:48:17,046] Trial 18 finished with value: 0.659946568502427 and parameters: {'num_leaves': 4}. Best is trial 18 with value: 0.659946568502427.
num_leaves, val_score: 0.659947:  60%|######    | 12/20 [00:11<00:06,  1.22it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001058 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.637047	valid's binary_logloss: 0.663485
[200]	train's binary_logloss: 0.621339	valid's binary_logloss: 0.662464
[300]	train's binary_logloss: 0.608631	valid's binary_logloss: 0.662754
Early stopping, best iteration is:
[218]	train's binary_logloss: 0.618989	valid's binary_logloss: 0.661824
num_leaves, val_score: 0.659947:  65%|######5   | 13/20 [00:12<00:05,  1.31it/s][I 2020-09-27 04:48:17,672] Trial 19 finished with value: 0.6618235788394593 and parameters: {'num_leaves': 7}. Best is trial 18 with value: 0.659946568502427.
num_leaves, val_score: 0.659947:  65%|######5   | 13/20 [00:12<00:05,  1.31it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.006735 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657096	valid's binary_logloss: 0.667367
[200]	train's binary_logloss: 0.646565	valid's binary_logloss: 0.661837
[300]	train's binary_logloss: 0.641199	valid's binary_logloss: 0.660556
[400]	train's binary_logloss: 0.637372	valid's binary_logloss: 0.660306
Early stopping, best iteration is:
[357]	train's binary_logloss: 0.638883	valid's binary_logloss: 0.659984
num_leaves, val_score: 0.659947:  70%|#######   | 14/20 [00:13<00:05,  1.13it/s][I 2020-09-27 04:48:18,847] Trial 20 finished with value: 0.6599838058476127 and parameters: {'num_leaves': 3}. Best is trial 18 with value: 0.659946568502427.
num_leaves, val_score: 0.659947:  70%|#######   | 14/20 [00:13<00:05,  1.13it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001022 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.645096	valid's binary_logloss: 0.662705
[200]	train's binary_logloss: 0.632688	valid's binary_logloss: 0.660038
Early stopping, best iteration is:
[193]	train's binary_logloss: 0.633406	valid's binary_logloss: 0.660001
num_leaves, val_score: 0.659947:  75%|#######5  | 15/20 [00:14<00:03,  1.26it/s][I 2020-09-27 04:48:19,418] Trial 21 finished with value: 0.6600011118675012 and parameters: {'num_leaves': 5}. Best is trial 18 with value: 0.659946568502427.
num_leaves, val_score: 0.659947:  75%|#######5  | 15/20 [00:14<00:03,  1.26it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001022 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.629941	valid's binary_logloss: 0.66087
[200]	train's binary_logloss: 0.610687	valid's binary_logloss: 0.662744
Early stopping, best iteration is:
[125]	train's binary_logloss: 0.624522	valid's binary_logloss: 0.66051
num_leaves, val_score: 0.659947:  80%|########  | 16/20 [00:14<00:02,  1.38it/s][I 2020-09-27 04:48:19,995] Trial 22 finished with value: 0.6605096965777126 and parameters: {'num_leaves': 9}. Best is trial 18 with value: 0.659946568502427.
num_leaves, val_score: 0.659947:  80%|########  | 16/20 [00:14<00:02,  1.38it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.006817 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.512177	valid's binary_logloss: 0.666645
Early stopping, best iteration is:
[29]	train's binary_logloss: 0.60465	valid's binary_logloss: 0.662506
num_leaves, val_score: 0.659947:  85%|########5 | 17/20 [00:15<00:02,  1.48it/s][I 2020-09-27 04:48:20,555] Trial 23 finished with value: 0.6625059680700698 and parameters: {'num_leaves': 60}. Best is trial 18 with value: 0.659946568502427.
num_leaves, val_score: 0.659947:  85%|########5 | 17/20 [00:15<00:02,  1.48it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001075 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650402	valid's binary_logloss: 0.665088
[200]	train's binary_logloss: 0.639194	valid's binary_logloss: 0.660755
[300]	train's binary_logloss: 0.632319	valid's binary_logloss: 0.66028
[400]	train's binary_logloss: 0.626788	valid's binary_logloss: 0.660425
Early stopping, best iteration is:
[360]	train's binary_logloss: 0.628825	valid's binary_logloss: 0.659947
num_leaves, val_score: 0.659947:  90%|######### | 18/20 [00:16<00:01,  1.40it/s][I 2020-09-27 04:48:21,359] Trial 24 finished with value: 0.6599465685024268 and parameters: {'num_leaves': 4}. Best is trial 24 with value: 0.6599465685024268.
num_leaves, val_score: 0.659947:  90%|######### | 18/20 [00:16<00:01,  1.40it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000833 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.569877	valid's binary_logloss: 0.667063
Early stopping, best iteration is:
[43]	train's binary_logloss: 0.613515	valid's binary_logloss: 0.663355
num_leaves, val_score: 0.659947:  95%|#########5| 19/20 [00:17<00:00,  1.27it/s][I 2020-09-27 04:48:22,313] Trial 25 finished with value: 0.6633553868179578 and parameters: {'num_leaves': 32}. Best is trial 24 with value: 0.6599465685024268.
num_leaves, val_score: 0.659947:  95%|#########5| 19/20 [00:17<00:00,  1.27it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000762 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.466555	valid's binary_logloss: 0.674499
Early stopping, best iteration is:
[32]	train's binary_logloss: 0.577932	valid's binary_logloss: 0.665241
num_leaves, val_score: 0.659947: 100%|##########| 20/20 [00:17<00:00,  1.30it/s][I 2020-09-27 04:48:23,045] Trial 26 finished with value: 0.6652408810612898 and parameters: {'num_leaves': 85}. Best is trial 24 with value: 0.6599465685024268.
num_leaves, val_score: 0.659947: 100%|##########| 20/20 [00:17<00:00,  1.13it/s]
bagging, val_score: 0.659947:   0%|          | 0/10 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000831 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.649887	valid's binary_logloss: 0.663942
[200]	train's binary_logloss: 0.639122	valid's binary_logloss: 0.660082
[300]	train's binary_logloss: 0.632108	valid's binary_logloss: 0.660084
Early stopping, best iteration is:
[239]	train's binary_logloss: 0.636191	valid's binary_logloss: 0.659231
bagging, val_score: 0.659231:  10%|#         | 1/10 [00:00<00:06,  1.48it/s][I 2020-09-27 04:48:23,737] Trial 27 finished with value: 0.6592308295470516 and parameters: {'bagging_fraction': 0.8273856061488931, 'bagging_freq': 3}. Best is trial 27 with value: 0.6592308295470516.
bagging, val_score: 0.659231:  10%|#         | 1/10 [00:00<00:06,  1.48it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004809 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.649764	valid's binary_logloss: 0.663649
[200]	train's binary_logloss: 0.638988	valid's binary_logloss: 0.65972
[300]	train's binary_logloss: 0.632186	valid's binary_logloss: 0.659366
[400]	train's binary_logloss: 0.626354	valid's binary_logloss: 0.659639
Early stopping, best iteration is:
[322]	train's binary_logloss: 0.630893	valid's binary_logloss: 0.659046
bagging, val_score: 0.659046:  20%|##        | 2/10 [00:01<00:05,  1.42it/s][I 2020-09-27 04:48:24,511] Trial 28 finished with value: 0.659045827636086 and parameters: {'bagging_fraction': 0.8505997166142588, 'bagging_freq': 3}. Best is trial 28 with value: 0.659045827636086.
bagging, val_score: 0.659046:  20%|##        | 2/10 [00:01<00:05,  1.42it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.006005 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631879	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633137	valid's binary_logloss: 0.657825
bagging, val_score: 0.657825:  30%|###       | 3/10 [00:02<00:05,  1.40it/s][I 2020-09-27 04:48:25,250] Trial 29 finished with value: 0.6578254229578112 and parameters: {'bagging_fraction': 0.8638496845155332, 'bagging_freq': 3}. Best is trial 29 with value: 0.6578254229578112.
bagging, val_score: 0.657825:  30%|###       | 3/10 [00:02<00:05,  1.40it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001015 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.649823	valid's binary_logloss: 0.663663
[200]	train's binary_logloss: 0.638792	valid's binary_logloss: 0.659288
[300]	train's binary_logloss: 0.631889	valid's binary_logloss: 0.658487
Early stopping, best iteration is:
[241]	train's binary_logloss: 0.635846	valid's binary_logloss: 0.658154
bagging, val_score: 0.657825:  40%|####      | 4/10 [00:03<00:04,  1.28it/s][I 2020-09-27 04:48:26,194] Trial 30 finished with value: 0.6581539755446953 and parameters: {'bagging_fraction': 0.8557672308630124, 'bagging_freq': 3}. Best is trial 29 with value: 0.6578254229578112.
bagging, val_score: 0.657825:  40%|####      | 4/10 [00:03<00:04,  1.28it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.002755 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650081	valid's binary_logloss: 0.663971
[200]	train's binary_logloss: 0.639132	valid's binary_logloss: 0.660066
[300]	train's binary_logloss: 0.632409	valid's binary_logloss: 0.659247
[400]	train's binary_logloss: 0.626618	valid's binary_logloss: 0.65931
Early stopping, best iteration is:
[341]	train's binary_logloss: 0.630063	valid's binary_logloss: 0.659011
bagging, val_score: 0.657825:  50%|#####     | 5/10 [00:04<00:04,  1.12it/s][I 2020-09-27 04:48:27,332] Trial 31 finished with value: 0.6590114175208794 and parameters: {'bagging_fraction': 0.8641735815144664, 'bagging_freq': 3}. Best is trial 29 with value: 0.6578254229578112.
bagging, val_score: 0.657825:  50%|#####     | 5/10 [00:04<00:04,  1.12it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000894 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.64988	valid's binary_logloss: 0.664431
[200]	train's binary_logloss: 0.639026	valid's binary_logloss: 0.660245
[300]	train's binary_logloss: 0.632354	valid's binary_logloss: 0.659402
[400]	train's binary_logloss: 0.626336	valid's binary_logloss: 0.659571
Early stopping, best iteration is:
[322]	train's binary_logloss: 0.630965	valid's binary_logloss: 0.659137
bagging, val_score: 0.657825:  60%|######    | 6/10 [00:05<00:03,  1.17it/s][I 2020-09-27 04:48:28,113] Trial 32 finished with value: 0.6591365814377496 and parameters: {'bagging_fraction': 0.8683729064753997, 'bagging_freq': 3}. Best is trial 29 with value: 0.6578254229578112.
bagging, val_score: 0.657825:  60%|######    | 6/10 [00:05<00:03,  1.17it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.006198 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.649852	valid's binary_logloss: 0.663933
[200]	train's binary_logloss: 0.639035	valid's binary_logloss: 0.65961
[300]	train's binary_logloss: 0.632241	valid's binary_logloss: 0.658575
[400]	train's binary_logloss: 0.62642	valid's binary_logloss: 0.658789
Early stopping, best iteration is:
[362]	train's binary_logloss: 0.628634	valid's binary_logloss: 0.658088
bagging, val_score: 0.657825:  70%|#######   | 7/10 [00:05<00:02,  1.16it/s][I 2020-09-27 04:48:28,978] Trial 33 finished with value: 0.6580884313848213 and parameters: {'bagging_fraction': 0.8714478568851503, 'bagging_freq': 3}. Best is trial 29 with value: 0.6578254229578112.
bagging, val_score: 0.657825:  70%|#######   | 7/10 [00:05<00:02,  1.16it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000833 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.649854	valid's binary_logloss: 0.66428
[200]	train's binary_logloss: 0.638784	valid's binary_logloss: 0.660031
[300]	train's binary_logloss: 0.631851	valid's binary_logloss: 0.659733
Early stopping, best iteration is:
[259]	train's binary_logloss: 0.634487	valid's binary_logloss: 0.659495
bagging, val_score: 0.657825:  80%|########  | 8/10 [00:06<00:01,  1.20it/s][I 2020-09-27 04:48:29,747] Trial 34 finished with value: 0.6594947433812736 and parameters: {'bagging_fraction': 0.9161669099599079, 'bagging_freq': 3}. Best is trial 29 with value: 0.6578254229578112.
bagging, val_score: 0.657825:  80%|########  | 8/10 [00:06<00:01,  1.20it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.003234 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.649122	valid's binary_logloss: 0.66312
[200]	train's binary_logloss: 0.638562	valid's binary_logloss: 0.660896
Early stopping, best iteration is:
[170]	train's binary_logloss: 0.640905	valid's binary_logloss: 0.660106
bagging, val_score: 0.657825:  90%|######### | 9/10 [00:07<00:00,  1.18it/s][I 2020-09-27 04:48:30,634] Trial 35 finished with value: 0.6601056363582504 and parameters: {'bagging_fraction': 0.5253666376896976, 'bagging_freq': 2}. Best is trial 29 with value: 0.6578254229578112.
bagging, val_score: 0.657825:  90%|######### | 9/10 [00:07<00:00,  1.18it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000924 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.64967	valid's binary_logloss: 0.663468
[200]	train's binary_logloss: 0.638905	valid's binary_logloss: 0.660271
[300]	train's binary_logloss: 0.632375	valid's binary_logloss: 0.660007
[400]	train's binary_logloss: 0.626598	valid's binary_logloss: 0.659422
[500]	train's binary_logloss: 0.621008	valid's binary_logloss: 0.658741
Early stopping, best iteration is:
[478]	train's binary_logloss: 0.622298	valid's binary_logloss: 0.658486
bagging, val_score: 0.657825: 100%|##########| 10/10 [00:08<00:00,  1.12it/s][I 2020-09-27 04:48:31,633] Trial 36 finished with value: 0.658485788443909 and parameters: {'bagging_fraction': 0.7531472629337493, 'bagging_freq': 6}. Best is trial 29 with value: 0.6578254229578112.
bagging, val_score: 0.657825: 100%|##########| 10/10 [00:08<00:00,  1.17it/s]
feature_fraction_stage2, val_score: 0.657825:   0%|          | 0/6 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000919 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.649945	valid's binary_logloss: 0.663854
[200]	train's binary_logloss: 0.638845	valid's binary_logloss: 0.659895
[300]	train's binary_logloss: 0.631902	valid's binary_logloss: 0.659391
[400]	train's binary_logloss: 0.626123	valid's binary_logloss: 0.659153
Early stopping, best iteration is:
[373]	train's binary_logloss: 0.627702	valid's binary_logloss: 0.658836
feature_fraction_stage2, val_score: 0.657825:  17%|#6        | 1/6 [00:00<00:04,  1.09it/s][I 2020-09-27 04:48:32,563] Trial 37 finished with value: 0.6588361247453068 and parameters: {'feature_fraction': 0.8480000000000001}. Best is trial 37 with value: 0.6588361247453068.
feature_fraction_stage2, val_score: 0.657825:  17%|#6        | 1/6 [00:00<00:04,  1.09it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000843 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631879	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633137	valid's binary_logloss: 0.657825
feature_fraction_stage2, val_score: 0.657825:  33%|###3      | 2/6 [00:01<00:03,  1.16it/s][I 2020-09-27 04:48:33,305] Trial 38 finished with value: 0.6578254229578112 and parameters: {'feature_fraction': 0.8160000000000001}. Best is trial 38 with value: 0.6578254229578112.
feature_fraction_stage2, val_score: 0.657825:  33%|###3      | 2/6 [00:01<00:03,  1.16it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000906 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.649964	valid's binary_logloss: 0.664299
[200]	train's binary_logloss: 0.638952	valid's binary_logloss: 0.66021
[300]	train's binary_logloss: 0.632195	valid's binary_logloss: 0.6599
Early stopping, best iteration is:
[265]	train's binary_logloss: 0.634377	valid's binary_logloss: 0.659183
feature_fraction_stage2, val_score: 0.657825:  50%|#####     | 3/6 [00:02<00:02,  1.11it/s][I 2020-09-27 04:48:34,293] Trial 39 finished with value: 0.6591825632130718 and parameters: {'feature_fraction': 0.88}. Best is trial 38 with value: 0.6578254229578112.
feature_fraction_stage2, val_score: 0.657825:  50%|#####     | 3/6 [00:02<00:02,  1.11it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016186 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650087	valid's binary_logloss: 0.664287
[200]	train's binary_logloss: 0.639299	valid's binary_logloss: 0.66052
[300]	train's binary_logloss: 0.632483	valid's binary_logloss: 0.659821
Early stopping, best iteration is:
[265]	train's binary_logloss: 0.634723	valid's binary_logloss: 0.659476
feature_fraction_stage2, val_score: 0.657825:  67%|######6   | 4/6 [00:03<00:01,  1.12it/s][I 2020-09-27 04:48:35,160] Trial 40 finished with value: 0.6594764202643758 and parameters: {'feature_fraction': 0.784}. Best is trial 38 with value: 0.6578254229578112.
feature_fraction_stage2, val_score: 0.657825:  67%|######6   | 4/6 [00:03<00:01,  1.12it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005048 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650002	valid's binary_logloss: 0.663921
[200]	train's binary_logloss: 0.639037	valid's binary_logloss: 0.659453
[300]	train's binary_logloss: 0.632258	valid's binary_logloss: 0.658764
Early stopping, best iteration is:
[263]	train's binary_logloss: 0.634669	valid's binary_logloss: 0.658316
feature_fraction_stage2, val_score: 0.657825:  83%|########3 | 5/6 [00:04<00:00,  1.21it/s][I 2020-09-27 04:48:35,831] Trial 41 finished with value: 0.6583158730564244 and parameters: {'feature_fraction': 0.7200000000000001}. Best is trial 38 with value: 0.6578254229578112.
feature_fraction_stage2, val_score: 0.657825:  83%|########3 | 5/6 [00:04<00:00,  1.21it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004831 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650087	valid's binary_logloss: 0.664287
[200]	train's binary_logloss: 0.639299	valid's binary_logloss: 0.66052
[300]	train's binary_logloss: 0.632483	valid's binary_logloss: 0.659821
Early stopping, best iteration is:
[265]	train's binary_logloss: 0.634723	valid's binary_logloss: 0.659476
feature_fraction_stage2, val_score: 0.657825: 100%|##########| 6/6 [00:04<00:00,  1.29it/s][I 2020-09-27 04:48:36,497] Trial 42 finished with value: 0.6594764202643758 and parameters: {'feature_fraction': 0.7520000000000001}. Best is trial 38 with value: 0.6578254229578112.
feature_fraction_stage2, val_score: 0.657825: 100%|##########| 6/6 [00:04<00:00,  1.24it/s]
regularization_factors, val_score: 0.657825:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000810 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650995	valid's binary_logloss: 0.663984
[200]	train's binary_logloss: 0.640816	valid's binary_logloss: 0.660341
[300]	train's binary_logloss: 0.635324	valid's binary_logloss: 0.659307
Early stopping, best iteration is:
[279]	train's binary_logloss: 0.636364	valid's binary_logloss: 0.659099
regularization_factors, val_score: 0.657825:   5%|5         | 1/20 [00:00<00:15,  1.24it/s][I 2020-09-27 04:48:37,323] Trial 43 finished with value: 0.659098571564232 and parameters: {'lambda_l1': 6.368584839611204, 'lambda_l2': 0.011340463712795283}. Best is trial 43 with value: 0.659098571564232.
regularization_factors, val_score: 0.657825:   5%|5         | 1/20 [00:00<00:15,  1.24it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004832 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631873	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633131	valid's binary_logloss: 0.657825
regularization_factors, val_score: 0.657825:  10%|#         | 2/20 [00:01<00:15,  1.13it/s][I 2020-09-27 04:48:38,394] Trial 44 finished with value: 0.6578254666437432 and parameters: {'lambda_l1': 2.7238144783184155e-08, 'lambda_l2': 1.008805152305263e-07}. Best is trial 44 with value: 0.6578254666437432.
regularization_factors, val_score: 0.657825:  10%|#         | 2/20 [00:01<00:15,  1.13it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.021204 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631873	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633131	valid's binary_logloss: 0.657825
regularization_factors, val_score: 0.657825:  15%|#5        | 3/20 [00:02<00:14,  1.13it/s][I 2020-09-27 04:48:39,260] Trial 45 finished with value: 0.6578254666410345 and parameters: {'lambda_l1': 1.0435559181670616e-08, 'lambda_l2': 1.585450495098862e-08}. Best is trial 45 with value: 0.6578254666410345.
regularization_factors, val_score: 0.657825:  15%|#5        | 3/20 [00:02<00:14,  1.13it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000902 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631876	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633134	valid's binary_logloss: 0.657826
regularization_factors, val_score: 0.657825:  20%|##        | 4/20 [00:03<00:13,  1.19it/s][I 2020-09-27 04:48:39,997] Trial 46 finished with value: 0.6578256353646856 and parameters: {'lambda_l1': 1.0219085602778263e-08, 'lambda_l2': 1.1221944608727081e-08}. Best is trial 45 with value: 0.6578254666410345.
regularization_factors, val_score: 0.657825:  20%|##        | 4/20 [00:03<00:13,  1.19it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001003 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631876	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633134	valid's binary_logloss: 0.657826
regularization_factors, val_score: 0.657825:  25%|##5       | 5/20 [00:04<00:12,  1.22it/s][I 2020-09-27 04:48:40,774] Trial 47 finished with value: 0.6578256353644696 and parameters: {'lambda_l1': 1.0248936810635019e-08, 'lambda_l2': 1.3858509908150066e-08}. Best is trial 45 with value: 0.6578254666410345.
regularization_factors, val_score: 0.657825:  25%|##5       | 5/20 [00:04<00:12,  1.22it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000912 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631876	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633134	valid's binary_logloss: 0.657826
regularization_factors, val_score: 0.657825:  30%|###       | 6/20 [00:04<00:11,  1.26it/s][I 2020-09-27 04:48:41,504] Trial 48 finished with value: 0.6578256353646091 and parameters: {'lambda_l1': 1.4330977464180962e-08, 'lambda_l2': 1.3975871503118737e-08}. Best is trial 45 with value: 0.6578254666410345.
regularization_factors, val_score: 0.657825:  30%|###       | 6/20 [00:04<00:11,  1.26it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.008196 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631873	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633131	valid's binary_logloss: 0.657825
regularization_factors, val_score: 0.657825:  35%|###5      | 7/20 [00:06<00:11,  1.10it/s][I 2020-09-27 04:48:42,673] Trial 49 finished with value: 0.6578254666417223 and parameters: {'lambda_l1': 1.3796608447685168e-08, 'lambda_l2': 4.7462022137544425e-08}. Best is trial 45 with value: 0.6578254666410345.
regularization_factors, val_score: 0.657825:  35%|###5      | 7/20 [00:06<00:11,  1.10it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000848 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631873	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633131	valid's binary_logloss: 0.657825
regularization_factors, val_score: 0.657825:  40%|####      | 8/20 [00:07<00:10,  1.13it/s][I 2020-09-27 04:48:43,519] Trial 50 finished with value: 0.6578254844386293 and parameters: {'lambda_l1': 5.179213600763938e-06, 'lambda_l2': 5.7935679948530245e-06}. Best is trial 45 with value: 0.6578254666410345.
regularization_factors, val_score: 0.657825:  40%|####      | 8/20 [00:07<00:10,  1.13it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000955 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631876	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633134	valid's binary_logloss: 0.657825
regularization_factors, val_score: 0.657825:  45%|####5     | 9/20 [00:07<00:09,  1.19it/s][I 2020-09-27 04:48:44,255] Trial 51 finished with value: 0.6578252553356203 and parameters: {'lambda_l1': 1.558360532559241e-05, 'lambda_l2': 4.619266691410038e-06}. Best is trial 51 with value: 0.6578252553356203.
regularization_factors, val_score: 0.657825:  45%|####5     | 9/20 [00:07<00:09,  1.19it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000894 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631876	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633134	valid's binary_logloss: 0.657826
regularization_factors, val_score: 0.657825:  50%|#####     | 10/20 [00:08<00:08,  1.23it/s][I 2020-09-27 04:48:44,994] Trial 52 finished with value: 0.6578256356080326 and parameters: {'lambda_l1': 2.7250062441770353e-06, 'lambda_l2': 2.367717105214214e-06}. Best is trial 51 with value: 0.6578252553356203.
regularization_factors, val_score: 0.657825:  50%|#####     | 10/20 [00:08<00:08,  1.23it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001566 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631876	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633134	valid's binary_logloss: 0.657826
regularization_factors, val_score: 0.657825:  55%|#####5    | 11/20 [00:09<00:07,  1.26it/s][I 2020-09-27 04:48:45,738] Trial 53 finished with value: 0.6578256354185161 and parameters: {'lambda_l1': 3.3319990687032807e-07, 'lambda_l2': 9.864870800228018e-07}. Best is trial 51 with value: 0.6578252553356203.
regularization_factors, val_score: 0.657825:  55%|#####5    | 11/20 [00:09<00:07,  1.26it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000843 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650018	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638999	valid's binary_logloss: 0.659122
[300]	train's binary_logloss: 0.632026	valid's binary_logloss: 0.658235
Early stopping, best iteration is:
[290]	train's binary_logloss: 0.632696	valid's binary_logloss: 0.658147
regularization_factors, val_score: 0.657825:  60%|######    | 12/20 [00:10<00:07,  1.08it/s][I 2020-09-27 04:48:46,984] Trial 54 finished with value: 0.6581469442698293 and parameters: {'lambda_l1': 0.006687572450441869, 'lambda_l2': 2.42706687165855e-07}. Best is trial 51 with value: 0.6578252553356203.
regularization_factors, val_score: 0.657825:  60%|######    | 12/20 [00:10<00:07,  1.08it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000800 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631879	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633137	valid's binary_logloss: 0.657825
regularization_factors, val_score: 0.657825:  65%|######5   | 13/20 [00:11<00:06,  1.14it/s][I 2020-09-27 04:48:47,745] Trial 55 finished with value: 0.6578254639325656 and parameters: {'lambda_l1': 0.0006287455216203462, 'lambda_l2': 8.873368461516288e-05}. Best is trial 51 with value: 0.6578252553356203.
regularization_factors, val_score: 0.657825:  65%|######5   | 13/20 [00:11<00:06,  1.14it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000955 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638998	valid's binary_logloss: 0.659122
[300]	train's binary_logloss: 0.632035	valid's binary_logloss: 0.658192
[400]	train's binary_logloss: 0.626308	valid's binary_logloss: 0.65838
Early stopping, best iteration is:
[307]	train's binary_logloss: 0.631536	valid's binary_logloss: 0.658173
regularization_factors, val_score: 0.657825:  70%|#######   | 14/20 [00:12<00:05,  1.18it/s][I 2020-09-27 04:48:48,519] Trial 56 finished with value: 0.6581734662791846 and parameters: {'lambda_l1': 0.0023930535028393245, 'lambda_l2': 0.00017393958564666502}. Best is trial 51 with value: 0.6578252553356203.
regularization_factors, val_score: 0.657825:  70%|#######   | 14/20 [00:12<00:05,  1.18it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000968 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631879	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633137	valid's binary_logloss: 0.657825
regularization_factors, val_score: 0.657825:  75%|#######5  | 15/20 [00:12<00:04,  1.23it/s][I 2020-09-27 04:48:49,262] Trial 57 finished with value: 0.6578254362526145 and parameters: {'lambda_l1': 0.0001139415321377858, 'lambda_l2': 0.00018316916066158292}. Best is trial 51 with value: 0.6578252553356203.
regularization_factors, val_score: 0.657825:  75%|#######5  | 15/20 [00:12<00:04,  1.23it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001132 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.63892	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.63188	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633137	valid's binary_logloss: 0.657825
regularization_factors, val_score: 0.657825:  80%|########  | 16/20 [00:13<00:03,  1.16it/s][I 2020-09-27 04:48:50,236] Trial 58 finished with value: 0.6578254556273145 and parameters: {'lambda_l1': 0.00013528915615576415, 'lambda_l2': 0.0002063652161436333}. Best is trial 51 with value: 0.6578252553356203.
regularization_factors, val_score: 0.657825:  80%|########  | 16/20 [00:13<00:03,  1.16it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.015940 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631876	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633134	valid's binary_logloss: 0.657826
regularization_factors, val_score: 0.657825:  85%|########5 | 17/20 [00:14<00:02,  1.13it/s][I 2020-09-27 04:48:51,159] Trial 59 finished with value: 0.6578256499831218 and parameters: {'lambda_l1': 0.00013602675042342853, 'lambda_l2': 0.00018327214045940748}. Best is trial 51 with value: 0.6578252553356203.
regularization_factors, val_score: 0.657825:  85%|########5 | 17/20 [00:14<00:02,  1.13it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001556 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.63892	valid's binary_logloss: 0.659405
[300]	train's binary_logloss: 0.631879	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633136	valid's binary_logloss: 0.657826
regularization_factors, val_score: 0.657825:  90%|######### | 18/20 [00:15<00:01,  1.20it/s][I 2020-09-27 04:48:51,873] Trial 60 finished with value: 0.6578258958216437 and parameters: {'lambda_l1': 0.00015253198331570547, 'lambda_l2': 0.0071388436435203435}. Best is trial 51 with value: 0.6578252553356203.
regularization_factors, val_score: 0.657825:  90%|######### | 18/20 [00:15<00:01,  1.20it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000941 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.63188	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633137	valid's binary_logloss: 0.657826
regularization_factors, val_score: 0.657825:  95%|#########5| 19/20 [00:16<00:00,  1.23it/s][I 2020-09-27 04:48:52,641] Trial 61 finished with value: 0.657825517008465 and parameters: {'lambda_l1': 0.0015382833109009928, 'lambda_l2': 4.130282202202608e-05}. Best is trial 51 with value: 0.6578252553356203.
regularization_factors, val_score: 0.657825:  95%|#########5| 19/20 [00:16<00:00,  1.23it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000998 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638919	valid's binary_logloss: 0.659404
[300]	train's binary_logloss: 0.631874	valid's binary_logloss: 0.657967
Early stopping, best iteration is:
[281]	train's binary_logloss: 0.633131	valid's binary_logloss: 0.657826
regularization_factors, val_score: 0.657825: 100%|##########| 20/20 [00:16<00:00,  1.24it/s][I 2020-09-27 04:48:53,434] Trial 62 finished with value: 0.6578255316540529 and parameters: {'lambda_l1': 3.339088375221171e-05, 'lambda_l2': 0.0017874187322570777}. Best is trial 51 with value: 0.6578252553356203.
regularization_factors, val_score: 0.657825: 100%|##########| 20/20 [00:16<00:00,  1.18it/s]
min_data_in_leaf, val_score: 0.657825:   0%|          | 0/5 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000868 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638922	valid's binary_logloss: 0.659362
[300]	train's binary_logloss: 0.632151	valid's binary_logloss: 0.658294
[400]	train's binary_logloss: 0.626333	valid's binary_logloss: 0.659009
Early stopping, best iteration is:
[300]	train's binary_logloss: 0.632151	valid's binary_logloss: 0.658294
min_data_in_leaf, val_score: 0.657825:  20%|##        | 1/5 [00:01<00:05,  1.28s/it][I 2020-09-27 04:48:54,726] Trial 63 finished with value: 0.6582938352043944 and parameters: {'min_child_samples': 25}. Best is trial 63 with value: 0.6582938352043944.
min_data_in_leaf, val_score: 0.657825:  20%|##        | 1/5 [00:01<00:05,  1.28s/it][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.005115 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638826	valid's binary_logloss: 0.659409
[300]	train's binary_logloss: 0.63179	valid's binary_logloss: 0.658711
Early stopping, best iteration is:
[232]	train's binary_logloss: 0.636374	valid's binary_logloss: 0.658604
min_data_in_leaf, val_score: 0.657825:  40%|####      | 2/5 [00:01<00:03,  1.10s/it][I 2020-09-27 04:48:55,420] Trial 64 finished with value: 0.6586044923878769 and parameters: {'min_child_samples': 5}. Best is trial 63 with value: 0.6582938352043944.
min_data_in_leaf, val_score: 0.657825:  40%|####      | 2/5 [00:01<00:03,  1.10s/it][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000876 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.65002	valid's binary_logloss: 0.663404
[200]	train's binary_logloss: 0.639166	valid's binary_logloss: 0.659615
[300]	train's binary_logloss: 0.632623	valid's binary_logloss: 0.659448
Early stopping, best iteration is:
[240]	train's binary_logloss: 0.636379	valid's binary_logloss: 0.65892
min_data_in_leaf, val_score: 0.657825:  60%|######    | 3/5 [00:02<00:01,  1.02it/s][I 2020-09-27 04:48:56,123] Trial 65 finished with value: 0.658919688761251 and parameters: {'min_child_samples': 50}. Best is trial 63 with value: 0.6582938352043944.
min_data_in_leaf, val_score: 0.657825:  60%|######    | 3/5 [00:02<00:01,  1.02it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.004999 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.65002	valid's binary_logloss: 0.663404
[200]	train's binary_logloss: 0.639125	valid's binary_logloss: 0.659535
[300]	train's binary_logloss: 0.632949	valid's binary_logloss: 0.658458
Early stopping, best iteration is:
[266]	train's binary_logloss: 0.634844	valid's binary_logloss: 0.658133
min_data_in_leaf, val_score: 0.657825:  80%|########  | 4/5 [00:03<00:00,  1.10it/s][I 2020-09-27 04:48:56,850] Trial 66 finished with value: 0.6581334712449208 and parameters: {'min_child_samples': 100}. Best is trial 66 with value: 0.6581334712449208.
min_data_in_leaf, val_score: 0.657825:  80%|########  | 4/5 [00:03<00:00,  1.10it/s][LightGBM] [Info] Number of positive: 12844, number of negative: 13156
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000869 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4241
[LightGBM] [Info] Number of data points in the train set: 26000, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.494000 -> initscore=-0.024001
[LightGBM] [Info] Start training from score -0.024001
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.650017	valid's binary_logloss: 0.663448
[200]	train's binary_logloss: 0.638826	valid's binary_logloss: 0.659409
[300]	train's binary_logloss: 0.631757	valid's binary_logloss: 0.658712
Early stopping, best iteration is:
[234]	train's binary_logloss: 0.636229	valid's binary_logloss: 0.65847
min_data_in_leaf, val_score: 0.657825: 100%|##########| 5/5 [00:04<00:00,  1.19it/s][I 2020-09-27 04:48:57,537] Trial 67 finished with value: 0.6584702414733883 and parameters: {'min_child_samples': 10}. Best is trial 66 with value: 0.6581334712449208.
min_data_in_leaf, val_score: 0.657825: 100%|##########| 5/5 [00:04<00:00,  1.22it/s]

################################
CV_score:0.6150646280811062

Fold : 0
[I 2020-09-27 04:48:57,616] A new study created in memory with name: no-name-8ff89ddf-7641-453c-86d0-fc507cfacb11
feature_fraction, val_score: inf:   0%|          | 0/7 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.026169 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.663649	valid's binary_logloss: 0.690213
Early stopping, best iteration is:
[45]	train's binary_logloss: 0.676671	valid's binary_logloss: 0.689632
feature_fraction, val_score: 0.689632:  14%|#4        | 1/7 [00:01<00:08,  1.45s/it][I 2020-09-27 04:48:59,075] Trial 0 finished with value: 0.6896322280348909 and parameters: {'feature_fraction': 0.7}. Best is trial 0 with value: 0.6896322280348909.
feature_fraction, val_score: 0.689632:  14%|#4        | 1/7 [00:01<00:08,  1.45s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001410 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.66617	valid's binary_logloss: 0.690438
Early stopping, best iteration is:
[41]	train's binary_logloss: 0.679097	valid's binary_logloss: 0.689783
feature_fraction, val_score: 0.689632:  29%|##8       | 2/7 [00:02<00:06,  1.24s/it][I 2020-09-27 04:48:59,837] Trial 1 finished with value: 0.6897826459094011 and parameters: {'feature_fraction': 0.4}. Best is trial 0 with value: 0.6896322280348909.
feature_fraction, val_score: 0.689632:  29%|##8       | 2/7 [00:02<00:06,  1.24s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.003154 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.662081	valid's binary_logloss: 0.690003
Early stopping, best iteration is:
[41]	train's binary_logloss: 0.677046	valid's binary_logloss: 0.68968
feature_fraction, val_score: 0.689632:  43%|####2     | 3/7 [00:03<00:04,  1.13s/it][I 2020-09-27 04:49:00,694] Trial 2 finished with value: 0.6896804566917485 and parameters: {'feature_fraction': 1.0}. Best is trial 0 with value: 0.6896322280348909.
feature_fraction, val_score: 0.689632:  43%|####2     | 3/7 [00:03<00:04,  1.13s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.015461 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.66415	valid's binary_logloss: 0.689593
Early stopping, best iteration is:
[39]	train's binary_logloss: 0.678462	valid's binary_logloss: 0.689021
feature_fraction, val_score: 0.689021:  57%|#####7    | 4/7 [00:03<00:03,  1.01s/it][I 2020-09-27 04:49:01,437] Trial 3 finished with value: 0.6890207233896298 and parameters: {'feature_fraction': 0.6}. Best is trial 3 with value: 0.6890207233896298.
feature_fraction, val_score: 0.689021:  57%|#####7    | 4/7 [00:03<00:03,  1.01s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.025191 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.66251	valid's binary_logloss: 0.689598
Early stopping, best iteration is:
[42]	train's binary_logloss: 0.676825	valid's binary_logloss: 0.68942
feature_fraction, val_score: 0.689021:  71%|#######1  | 5/7 [00:05<00:02,  1.13s/it][I 2020-09-27 04:49:02,837] Trial 4 finished with value: 0.6894197474355058 and parameters: {'feature_fraction': 0.8999999999999999}. Best is trial 3 with value: 0.6890207233896298.
feature_fraction, val_score: 0.689021:  71%|#######1  | 5/7 [00:05<00:02,  1.13s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.017648 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.663408	valid's binary_logloss: 0.690473
Early stopping, best iteration is:
[45]	train's binary_logloss: 0.676529	valid's binary_logloss: 0.689828
feature_fraction, val_score: 0.689021:  86%|########5 | 6/7 [00:06<00:01,  1.04s/it][I 2020-09-27 04:49:03,685] Trial 5 finished with value: 0.68982789311769 and parameters: {'feature_fraction': 0.8}. Best is trial 3 with value: 0.6890207233896298.
feature_fraction, val_score: 0.689021:  86%|########5 | 6/7 [00:06<00:01,  1.04s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001730 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.665183	valid's binary_logloss: 0.689631
[200]	train's binary_logloss: 0.645808	valid's binary_logloss: 0.690591
Early stopping, best iteration is:
[102]	train's binary_logloss: 0.664779	valid's binary_logloss: 0.689546
feature_fraction, val_score: 0.689021: 100%|##########| 7/7 [00:07<00:00,  1.07s/it][I 2020-09-27 04:49:04,813] Trial 6 finished with value: 0.6895458054912846 and parameters: {'feature_fraction': 0.5}. Best is trial 3 with value: 0.6890207233896298.
feature_fraction, val_score: 0.689021: 100%|##########| 7/7 [00:07<00:00,  1.03s/it]
num_leaves, val_score: 0.689021:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012666 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.577872	valid's binary_logloss: 0.693632
Early stopping, best iteration is:
[18]	train's binary_logloss: 0.663683	valid's binary_logloss: 0.689631
num_leaves, val_score: 0.689021:   5%|5         | 1/20 [00:01<00:30,  1.61s/it][I 2020-09-27 04:49:06,444] Trial 7 finished with value: 0.689631097968691 and parameters: {'num_leaves': 168}. Best is trial 7 with value: 0.689631097968691.
num_leaves, val_score: 0.689021:   5%|5         | 1/20 [00:01<00:30,  1.61s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.017419 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.604202	valid's binary_logloss: 0.69167
Early stopping, best iteration is:
[38]	train's binary_logloss: 0.650476	valid's binary_logloss: 0.689618
num_leaves, val_score: 0.689021:  10%|#         | 2/20 [00:02<00:26,  1.47s/it][I 2020-09-27 04:49:07,566] Trial 8 finished with value: 0.689618324134334 and parameters: {'num_leaves': 122}. Best is trial 8 with value: 0.689618324134334.
num_leaves, val_score: 0.689021:  10%|#         | 2/20 [00:02<00:26,  1.47s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016515 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.685172	valid's binary_logloss: 0.689666
[200]	train's binary_logloss: 0.681255	valid's binary_logloss: 0.689678
Early stopping, best iteration is:
[156]	train's binary_logloss: 0.68289	valid's binary_logloss: 0.689534
num_leaves, val_score: 0.689021:  15%|#5        | 3/20 [00:03<00:22,  1.34s/it][I 2020-09-27 04:49:08,626] Trial 9 finished with value: 0.6895342400114649 and parameters: {'num_leaves': 6}. Best is trial 9 with value: 0.6895342400114649.
num_leaves, val_score: 0.689021:  15%|#5        | 3/20 [00:03<00:22,  1.34s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016558 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.535492	valid's binary_logloss: 0.695541
Early stopping, best iteration is:
[19]	train's binary_logloss: 0.649759	valid's binary_logloss: 0.690794
num_leaves, val_score: 0.689021:  20%|##        | 4/20 [00:05<00:23,  1.49s/it][I 2020-09-27 04:49:10,461] Trial 10 finished with value: 0.6907940845828584 and parameters: {'num_leaves': 253}. Best is trial 9 with value: 0.6895342400114649.
num_leaves, val_score: 0.689021:  20%|##        | 4/20 [00:05<00:23,  1.49s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.017448 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.67101	valid's binary_logloss: 0.689552
Early stopping, best iteration is:
[93]	train's binary_logloss: 0.672156	valid's binary_logloss: 0.68942
num_leaves, val_score: 0.689021:  25%|##5       | 5/20 [00:06<00:19,  1.31s/it][I 2020-09-27 04:49:11,362] Trial 11 finished with value: 0.6894197097930296 and parameters: {'num_leaves': 22}. Best is trial 11 with value: 0.6894197097930296.
num_leaves, val_score: 0.689021:  25%|##5       | 5/20 [00:06<00:19,  1.31s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.014207 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.690044	valid's binary_logloss: 0.690612
[200]	train's binary_logloss: 0.68885	valid's binary_logloss: 0.689962
[300]	train's binary_logloss: 0.688098	valid's binary_logloss: 0.689612
[400]	train's binary_logloss: 0.687569	valid's binary_logloss: 0.689487
[500]	train's binary_logloss: 0.68718	valid's binary_logloss: 0.689391
[600]	train's binary_logloss: 0.686881	valid's binary_logloss: 0.689318
[700]	train's binary_logloss: 0.68664	valid's binary_logloss: 0.689265
[800]	train's binary_logloss: 0.68644	valid's binary_logloss: 0.689264
Early stopping, best iteration is:
[710]	train's binary_logloss: 0.686618	valid's binary_logloss: 0.689244
num_leaves, val_score: 0.689021:  30%|###       | 6/20 [00:09<00:25,  1.86s/it][I 2020-09-27 04:49:14,484] Trial 12 finished with value: 0.689244475322615 and parameters: {'num_leaves': 2}. Best is trial 12 with value: 0.689244475322615.
num_leaves, val_score: 0.689021:  30%|###       | 6/20 [00:09<00:25,  1.86s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.015182 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.630223	valid's binary_logloss: 0.692152
Early stopping, best iteration is:
[21]	train's binary_logloss: 0.674331	valid's binary_logloss: 0.690049
num_leaves, val_score: 0.689021:  35%|###5      | 7/20 [00:10<00:20,  1.56s/it][I 2020-09-27 04:49:15,354] Trial 13 finished with value: 0.6900491534448284 and parameters: {'num_leaves': 80}. Best is trial 12 with value: 0.689244475322615.
num_leaves, val_score: 0.689021:  35%|###5      | 7/20 [00:10<00:20,  1.56s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.017787 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.543727	valid's binary_logloss: 0.696815
Early stopping, best iteration is:
[10]	train's binary_logloss: 0.670334	valid's binary_logloss: 0.691482
num_leaves, val_score: 0.689021:  40%|####      | 8/20 [00:11<00:17,  1.46s/it][I 2020-09-27 04:49:16,588] Trial 14 finished with value: 0.6914823998645343 and parameters: {'num_leaves': 234}. Best is trial 12 with value: 0.689244475322615.
num_leaves, val_score: 0.689021:  40%|####      | 8/20 [00:11<00:17,  1.46s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.015457 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.646446	valid's binary_logloss: 0.691884
Early stopping, best iteration is:
[17]	train's binary_logloss: 0.681433	valid's binary_logloss: 0.689734
num_leaves, val_score: 0.689021:  45%|####5     | 9/20 [00:13<00:15,  1.43s/it][I 2020-09-27 04:49:17,939] Trial 15 finished with value: 0.6897344485270362 and parameters: {'num_leaves': 55}. Best is trial 12 with value: 0.689244475322615.
num_leaves, val_score: 0.689021:  45%|####5     | 9/20 [00:13<00:15,  1.43s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.015412 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.570477	valid's binary_logloss: 0.695696
Early stopping, best iteration is:
[23]	train's binary_logloss: 0.654169	valid's binary_logloss: 0.690909
num_leaves, val_score: 0.689021:  50%|#####     | 10/20 [00:14<00:13,  1.34s/it][I 2020-09-27 04:49:19,067] Trial 16 finished with value: 0.6909092950799796 and parameters: {'num_leaves': 181}. Best is trial 12 with value: 0.689244475322615.
num_leaves, val_score: 0.689021:  50%|#####     | 10/20 [00:14<00:13,  1.34s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016435 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.623443	valid's binary_logloss: 0.691768
Early stopping, best iteration is:
[19]	train's binary_logloss: 0.674343	valid's binary_logloss: 0.690276
num_leaves, val_score: 0.689021:  55%|#####5    | 11/20 [00:15<00:10,  1.20s/it][I 2020-09-27 04:49:19,941] Trial 17 finished with value: 0.690276336147683 and parameters: {'num_leaves': 90}. Best is trial 12 with value: 0.689244475322615.
num_leaves, val_score: 0.689021:  55%|#####5    | 11/20 [00:15<00:10,  1.20s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016181 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.659158	valid's binary_logloss: 0.689655
Early stopping, best iteration is:
[35]	train's binary_logloss: 0.677405	valid's binary_logloss: 0.689049
num_leaves, val_score: 0.689021:  60%|######    | 12/20 [00:15<00:08,  1.07s/it][I 2020-09-27 04:49:20,717] Trial 18 finished with value: 0.6890491443530252 and parameters: {'num_leaves': 38}. Best is trial 18 with value: 0.6890491443530252.
num_leaves, val_score: 0.689021:  60%|######    | 12/20 [00:15<00:08,  1.07s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.023330 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.652371	valid's binary_logloss: 0.690333
Early stopping, best iteration is:
[39]	train's binary_logloss: 0.672949	valid's binary_logloss: 0.68919
num_leaves, val_score: 0.689021:  65%|######5   | 13/20 [00:17<00:08,  1.17s/it][I 2020-09-27 04:49:22,101] Trial 19 finished with value: 0.6891896975164848 and parameters: {'num_leaves': 47}. Best is trial 18 with value: 0.6890491443530252.
num_leaves, val_score: 0.689021:  65%|######5   | 13/20 [00:17<00:08,  1.17s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.017299 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.59964	valid's binary_logloss: 0.69374
Early stopping, best iteration is:
[14]	train's binary_logloss: 0.674012	valid's binary_logloss: 0.690104
num_leaves, val_score: 0.689021:  70%|#######   | 14/20 [00:18<00:06,  1.10s/it][I 2020-09-27 04:49:23,062] Trial 20 finished with value: 0.6901038597516463 and parameters: {'num_leaves': 130}. Best is trial 18 with value: 0.6890491443530252.
num_leaves, val_score: 0.689021:  70%|#######   | 14/20 [00:18<00:06,  1.10s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016669 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657163	valid's binary_logloss: 0.690788
Early stopping, best iteration is:
[43]	train's binary_logloss: 0.674068	valid's binary_logloss: 0.689599
num_leaves, val_score: 0.689021:  75%|#######5  | 15/20 [00:19<00:05,  1.01s/it][I 2020-09-27 04:49:23,851] Trial 21 finished with value: 0.6895994157631041 and parameters: {'num_leaves': 40}. Best is trial 18 with value: 0.6890491443530252.
num_leaves, val_score: 0.689021:  75%|#######5  | 15/20 [00:19<00:05,  1.01s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016387 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.632347	valid's binary_logloss: 0.692318
Early stopping, best iteration is:
[23]	train's binary_logloss: 0.673649	valid's binary_logloss: 0.689645
num_leaves, val_score: 0.689021:  80%|########  | 16/20 [00:19<00:03,  1.05it/s][I 2020-09-27 04:49:24,677] Trial 22 finished with value: 0.6896448463652016 and parameters: {'num_leaves': 76}. Best is trial 18 with value: 0.6890491443530252.
num_leaves, val_score: 0.689021:  80%|########  | 16/20 [00:19<00:03,  1.05it/s][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.020197 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.656137	valid's binary_logloss: 0.689636
Early stopping, best iteration is:
[96]	train's binary_logloss: 0.657272	valid's binary_logloss: 0.689437
num_leaves, val_score: 0.689021:  85%|########5 | 17/20 [00:21<00:03,  1.10s/it][I 2020-09-27 04:49:26,120] Trial 23 finished with value: 0.6894373899731695 and parameters: {'num_leaves': 42}. Best is trial 18 with value: 0.6890491443530252.
num_leaves, val_score: 0.689021:  85%|########5 | 17/20 [00:21<00:03,  1.10s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016258 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.608119	valid's binary_logloss: 0.691911
Early stopping, best iteration is:
[30]	train's binary_logloss: 0.659223	valid's binary_logloss: 0.689435
num_leaves, val_score: 0.689021:  90%|######### | 18/20 [00:22<00:02,  1.09s/it][I 2020-09-27 04:49:27,187] Trial 24 finished with value: 0.6894354954462946 and parameters: {'num_leaves': 116}. Best is trial 18 with value: 0.6890491443530252.
num_leaves, val_score: 0.689021:  90%|######### | 18/20 [00:22<00:02,  1.09s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.018058 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.68615	valid's binary_logloss: 0.689604
[200]	train's binary_logloss: 0.682919	valid's binary_logloss: 0.689378
Early stopping, best iteration is:
[158]	train's binary_logloss: 0.684169	valid's binary_logloss: 0.689219
num_leaves, val_score: 0.689021:  95%|#########5| 19/20 [00:23<00:01,  1.08s/it][I 2020-09-27 04:49:28,242] Trial 25 finished with value: 0.6892189395196165 and parameters: {'num_leaves': 5}. Best is trial 18 with value: 0.6890491443530252.
num_leaves, val_score: 0.689021:  95%|#########5| 19/20 [00:23<00:01,  1.08s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016101 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.651145	valid's binary_logloss: 0.690351
Early stopping, best iteration is:
[34]	train's binary_logloss: 0.674412	valid's binary_logloss: 0.689378
num_leaves, val_score: 0.689021: 100%|##########| 20/20 [00:24<00:00,  1.16s/it][I 2020-09-27 04:49:29,591] Trial 26 finished with value: 0.6893783362619839 and parameters: {'num_leaves': 49}. Best is trial 18 with value: 0.6890491443530252.
num_leaves, val_score: 0.689021: 100%|##########| 20/20 [00:24<00:00,  1.24s/it]
bagging, val_score: 0.689021:   0%|          | 0/10 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.013092 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664277	valid's binary_logloss: 0.68939
Early stopping, best iteration is:
[71]	train's binary_logloss: 0.670651	valid's binary_logloss: 0.688825
bagging, val_score: 0.688825:  10%|#         | 1/10 [00:01<00:09,  1.08s/it][I 2020-09-27 04:49:30,680] Trial 27 finished with value: 0.6888247862401992 and parameters: {'bagging_fraction': 0.958245225148382, 'bagging_freq': 2}. Best is trial 27 with value: 0.6888247862401992.
bagging, val_score: 0.688825:  10%|#         | 1/10 [00:01<00:09,  1.08s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.014041 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664031	valid's binary_logloss: 0.689087
[200]	train's binary_logloss: 0.64388	valid's binary_logloss: 0.689768
Early stopping, best iteration is:
[119]	train's binary_logloss: 0.659997	valid's binary_logloss: 0.688879
bagging, val_score: 0.688825:  20%|##        | 2/10 [00:02<00:09,  1.15s/it][I 2020-09-27 04:49:32,003] Trial 28 finished with value: 0.6888794888870088 and parameters: {'bagging_fraction': 0.9745012448452529, 'bagging_freq': 2}. Best is trial 27 with value: 0.6888247862401992.
bagging, val_score: 0.688825:  20%|##        | 2/10 [00:02<00:09,  1.15s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.015566 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664149	valid's binary_logloss: 0.689861
Early stopping, best iteration is:
[50]	train's binary_logloss: 0.675702	valid's binary_logloss: 0.689193
bagging, val_score: 0.688825:  30%|###       | 3/10 [00:03<00:08,  1.23s/it][I 2020-09-27 04:49:33,428] Trial 29 finished with value: 0.689193182327793 and parameters: {'bagging_fraction': 0.9667819838074765, 'bagging_freq': 2}. Best is trial 27 with value: 0.6888247862401992.
bagging, val_score: 0.688825:  30%|###       | 3/10 [00:03<00:08,  1.23s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.013341 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664296	valid's binary_logloss: 0.689547
Early stopping, best iteration is:
[60]	train's binary_logloss: 0.673228	valid's binary_logloss: 0.689074
bagging, val_score: 0.688825:  40%|####      | 4/10 [00:04<00:07,  1.17s/it][I 2020-09-27 04:49:34,451] Trial 30 finished with value: 0.6890740654309894 and parameters: {'bagging_fraction': 0.9690126597280045, 'bagging_freq': 2}. Best is trial 27 with value: 0.6888247862401992.
bagging, val_score: 0.688825:  40%|####      | 4/10 [00:04<00:07,  1.17s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016944 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664297	valid's binary_logloss: 0.690412
Early stopping, best iteration is:
[54]	train's binary_logloss: 0.674733	valid's binary_logloss: 0.68958
bagging, val_score: 0.688825:  50%|#####     | 5/10 [00:05<00:05,  1.10s/it][I 2020-09-27 04:49:35,398] Trial 31 finished with value: 0.6895802218638041 and parameters: {'bagging_fraction': 0.9943804516504096, 'bagging_freq': 4}. Best is trial 27 with value: 0.6888247862401992.
bagging, val_score: 0.688825:  50%|#####     | 5/10 [00:05<00:05,  1.10s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.014817 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664664	valid's binary_logloss: 0.689957
Early stopping, best iteration is:
[37]	train's binary_logloss: 0.679397	valid's binary_logloss: 0.689326
bagging, val_score: 0.688825:  60%|######    | 6/10 [00:06<00:04,  1.07s/it][I 2020-09-27 04:49:36,386] Trial 32 finished with value: 0.6893263473853064 and parameters: {'bagging_fraction': 0.6476701297352184, 'bagging_freq': 1}. Best is trial 27 with value: 0.6888247862401992.
bagging, val_score: 0.688825:  60%|######    | 6/10 [00:06<00:04,  1.07s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.029663 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
bagging, val_score: 0.688719:  70%|#######   | 7/10 [00:08<00:03,  1.16s/it][I 2020-09-27 04:49:37,766] Trial 33 finished with value: 0.6887188408517487 and parameters: {'bagging_fraction': 0.8287757616049252, 'bagging_freq': 7}. Best is trial 33 with value: 0.6887188408517487.
bagging, val_score: 0.688719:  70%|#######   | 7/10 [00:08<00:03,  1.16s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012644 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.66436	valid's binary_logloss: 0.689716
Early stopping, best iteration is:
[51]	train's binary_logloss: 0.675497	valid's binary_logloss: 0.689039
bagging, val_score: 0.688719:  80%|########  | 8/10 [00:09<00:02,  1.09s/it][I 2020-09-27 04:49:38,691] Trial 34 finished with value: 0.689039279019024 and parameters: {'bagging_fraction': 0.8282760453593162, 'bagging_freq': 7}. Best is trial 33 with value: 0.6887188408517487.
bagging, val_score: 0.688719:  80%|########  | 8/10 [00:09<00:02,  1.09s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016881 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.663999	valid's binary_logloss: 0.690093
Early stopping, best iteration is:
[92]	train's binary_logloss: 0.665835	valid's binary_logloss: 0.689992
bagging, val_score: 0.688719:  90%|######### | 9/10 [00:10<00:01,  1.10s/it][I 2020-09-27 04:49:39,803] Trial 35 finished with value: 0.6899923113836022 and parameters: {'bagging_fraction': 0.8214825248241318, 'bagging_freq': 4}. Best is trial 33 with value: 0.6887188408517487.
bagging, val_score: 0.688719:  90%|######### | 9/10 [00:10<00:01,  1.10s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.013803 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664239	valid's binary_logloss: 0.688911
[200]	train's binary_logloss: 0.644408	valid's binary_logloss: 0.690305
Early stopping, best iteration is:
[103]	train's binary_logloss: 0.663554	valid's binary_logloss: 0.688799
bagging, val_score: 0.688719: 100%|##########| 10/10 [00:11<00:00,  1.24s/it][I 2020-09-27 04:49:41,362] Trial 36 finished with value: 0.6887991231168928 and parameters: {'bagging_fraction': 0.8383907363062589, 'bagging_freq': 7}. Best is trial 33 with value: 0.6887188408517487.
bagging, val_score: 0.688719: 100%|##########| 10/10 [00:11<00:00,  1.18s/it]
feature_fraction_stage2, val_score: 0.688719:   0%|          | 0/6 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.013665 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
feature_fraction_stage2, val_score: 0.688719:  17%|#6        | 1/6 [00:00<00:04,  1.03it/s][I 2020-09-27 04:49:42,349] Trial 37 finished with value: 0.6887188408517487 and parameters: {'feature_fraction': 0.616}. Best is trial 37 with value: 0.6887188408517487.
feature_fraction_stage2, val_score: 0.688719:  17%|#6        | 1/6 [00:00<00:04,  1.03it/s][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001729 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664799	valid's binary_logloss: 0.689634
Early stopping, best iteration is:
[69]	train's binary_logloss: 0.671673	valid's binary_logloss: 0.689345
feature_fraction_stage2, val_score: 0.688719:  33%|###3      | 2/6 [00:02<00:03,  1.00it/s][I 2020-09-27 04:49:43,417] Trial 38 finished with value: 0.6893450614153805 and parameters: {'feature_fraction': 0.552}. Best is trial 37 with value: 0.6887188408517487.
feature_fraction_stage2, val_score: 0.688719:  33%|###3      | 2/6 [00:02<00:03,  1.00it/s][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001868 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664799	valid's binary_logloss: 0.689634
Early stopping, best iteration is:
[69]	train's binary_logloss: 0.671673	valid's binary_logloss: 0.689345
feature_fraction_stage2, val_score: 0.688719:  50%|#####     | 3/6 [00:03<00:03,  1.17s/it][I 2020-09-27 04:49:44,990] Trial 39 finished with value: 0.6893450614153805 and parameters: {'feature_fraction': 0.52}. Best is trial 37 with value: 0.6887188408517487.
feature_fraction_stage2, val_score: 0.688719:  50%|#####     | 3/6 [00:03<00:03,  1.17s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001814 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664285	valid's binary_logloss: 0.690298
Early stopping, best iteration is:
[64]	train's binary_logloss: 0.672325	valid's binary_logloss: 0.689397
feature_fraction_stage2, val_score: 0.688719:  67%|######6   | 4/6 [00:04<00:02,  1.14s/it][I 2020-09-27 04:49:46,067] Trial 40 finished with value: 0.6893965384793862 and parameters: {'feature_fraction': 0.584}. Best is trial 37 with value: 0.6887188408517487.
feature_fraction_stage2, val_score: 0.688719:  67%|######6   | 4/6 [00:04<00:02,  1.14s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.013983 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.663794	valid's binary_logloss: 0.689901
[200]	train's binary_logloss: 0.64382	valid's binary_logloss: 0.690825
Early stopping, best iteration is:
[115]	train's binary_logloss: 0.66065	valid's binary_logloss: 0.689797
feature_fraction_stage2, val_score: 0.688719:  83%|########3 | 5/6 [00:06<00:01,  1.42s/it][I 2020-09-27 04:49:48,122] Trial 41 finished with value: 0.6897973746145477 and parameters: {'feature_fraction': 0.6799999999999999}. Best is trial 37 with value: 0.6887188408517487.
feature_fraction_stage2, val_score: 0.688719:  83%|########3 | 5/6 [00:06<00:01,  1.42s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.031387 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.663746	valid's binary_logloss: 0.689073
[200]	train's binary_logloss: 0.643757	valid's binary_logloss: 0.691217
Early stopping, best iteration is:
[105]	train's binary_logloss: 0.662622	valid's binary_logloss: 0.689054
feature_fraction_stage2, val_score: 0.688719: 100%|##########| 6/6 [00:08<00:00,  1.44s/it][I 2020-09-27 04:49:49,607] Trial 42 finished with value: 0.6890541064139101 and parameters: {'feature_fraction': 0.6479999999999999}. Best is trial 37 with value: 0.6887188408517487.
feature_fraction_stage2, val_score: 0.688719: 100%|##########| 6/6 [00:08<00:00,  1.37s/it]
regularization_factors, val_score: 0.688719:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012472 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664985	valid's binary_logloss: 0.690223
[200]	train's binary_logloss: 0.646149	valid's binary_logloss: 0.690617
Early stopping, best iteration is:
[151]	train's binary_logloss: 0.655125	valid's binary_logloss: 0.690101
regularization_factors, val_score: 0.688719:   5%|5         | 1/20 [00:01<00:28,  1.50s/it][I 2020-09-27 04:49:51,126] Trial 43 finished with value: 0.6901011755065565 and parameters: {'lambda_l1': 1.1303470188424833, 'lambda_l2': 0.06375127708462985}. Best is trial 43 with value: 0.6901011755065565.
regularization_factors, val_score: 0.688719:   5%|5         | 1/20 [00:01<00:28,  1.50s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.014406 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  10%|#         | 2/20 [00:02<00:26,  1.47s/it][I 2020-09-27 04:49:52,538] Trial 44 finished with value: 0.6887188408527825 and parameters: {'lambda_l1': 1.8636898004930976e-08, 'lambda_l2': 6.192880162509011e-08}. Best is trial 44 with value: 0.6887188408527825.
regularization_factors, val_score: 0.688719:  10%|#         | 2/20 [00:02<00:26,  1.47s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016047 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  15%|#5        | 3/20 [00:03<00:22,  1.32s/it][I 2020-09-27 04:49:53,496] Trial 45 finished with value: 0.6887188408522336 and parameters: {'lambda_l1': 1.8439298966850396e-08, 'lambda_l2': 1.3511479384741456e-08}. Best is trial 45 with value: 0.6887188408522336.
regularization_factors, val_score: 0.688719:  15%|#5        | 3/20 [00:03<00:22,  1.32s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012081 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  20%|##        | 4/20 [00:04<00:19,  1.20s/it][I 2020-09-27 04:49:54,422] Trial 46 finished with value: 0.6887188408519217 and parameters: {'lambda_l1': 1.2530425311422802e-08, 'lambda_l2': 1.0756585467257088e-08}. Best is trial 46 with value: 0.6887188408519217.
regularization_factors, val_score: 0.688719:  20%|##        | 4/20 [00:04<00:19,  1.20s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.014442 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  25%|##5       | 5/20 [00:05<00:16,  1.13s/it][I 2020-09-27 04:49:55,370] Trial 47 finished with value: 0.6887188408519394 and parameters: {'lambda_l1': 1.2148997058183217e-08, 'lambda_l2': 1.076670603152709e-08}. Best is trial 46 with value: 0.6887188408519217.
regularization_factors, val_score: 0.688719:  25%|##5       | 5/20 [00:05<00:16,  1.13s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.015325 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  30%|###       | 6/20 [00:07<00:16,  1.21s/it][I 2020-09-27 04:49:56,764] Trial 48 finished with value: 0.6887188408519016 and parameters: {'lambda_l1': 1.0473302644937364e-08, 'lambda_l2': 1.4000255193861959e-08}. Best is trial 48 with value: 0.6887188408519016.
regularization_factors, val_score: 0.688719:  30%|###       | 6/20 [00:07<00:16,  1.21s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.017485 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  35%|###5      | 7/20 [00:08<00:14,  1.15s/it][I 2020-09-27 04:49:57,778] Trial 49 finished with value: 0.6887188408520548 and parameters: {'lambda_l1': 1.2933235345738932e-08, 'lambda_l2': 1.3210044579433778e-08}. Best is trial 48 with value: 0.6887188408519016.
regularization_factors, val_score: 0.688719:  35%|###5      | 7/20 [00:08<00:14,  1.15s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016188 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  40%|####      | 8/20 [00:09<00:13,  1.09s/it][I 2020-09-27 04:49:58,722] Trial 50 finished with value: 0.6887188410169663 and parameters: {'lambda_l1': 6.522515649363731e-06, 'lambda_l2': 5.399603910938485e-06}. Best is trial 48 with value: 0.6887188408519016.
regularization_factors, val_score: 0.688719:  40%|####      | 8/20 [00:09<00:13,  1.09s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.018393 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  45%|####5     | 9/20 [00:10<00:12,  1.18s/it][I 2020-09-27 04:50:00,120] Trial 51 finished with value: 0.6887188408522066 and parameters: {'lambda_l1': 1.7781053835321766e-08, 'lambda_l2': 1.2911207447672734e-08}. Best is trial 48 with value: 0.6887188408519016.
regularization_factors, val_score: 0.688719:  45%|####5     | 9/20 [00:10<00:12,  1.18s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012528 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  50%|#####     | 10/20 [00:11<00:11,  1.14s/it][I 2020-09-27 04:50:01,154] Trial 52 finished with value: 0.6887188408521625 and parameters: {'lambda_l1': 1.487838960944784e-08, 'lambda_l2': 1.2171741979090901e-08}. Best is trial 48 with value: 0.6887188408519016.
regularization_factors, val_score: 0.688719:  50%|#####     | 10/20 [00:11<00:11,  1.14s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.013835 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  55%|#####5    | 11/20 [00:12<00:09,  1.07s/it][I 2020-09-27 04:50:02,068] Trial 53 finished with value: 0.688718840880323 and parameters: {'lambda_l1': 1.2436063977852807e-06, 'lambda_l2': 5.416531879194707e-07}. Best is trial 48 with value: 0.6887188408519016.
regularization_factors, val_score: 0.688719:  55%|#####5    | 11/20 [00:12<00:09,  1.07s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.020591 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  60%|######    | 12/20 [00:13<00:08,  1.03s/it][I 2020-09-27 04:50:03,019] Trial 54 finished with value: 0.6887188408518048 and parameters: {'lambda_l1': 1.001927452586009e-08, 'lambda_l2': 1.0173335682514728e-08}. Best is trial 54 with value: 0.6887188408518048.
regularization_factors, val_score: 0.688719:  60%|######    | 12/20 [00:13<00:08,  1.03s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.013735 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  65%|######5   | 13/20 [00:14<00:07,  1.13s/it][I 2020-09-27 04:50:04,380] Trial 55 finished with value: 0.6887188408925199 and parameters: {'lambda_l1': 1.0246196926378488e-06, 'lambda_l2': 2.2586132878897466e-06}. Best is trial 54 with value: 0.6887188408518048.
regularization_factors, val_score: 0.688719:  65%|######5   | 13/20 [00:14<00:07,  1.13s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.017187 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664481	valid's binary_logloss: 0.689645
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675961	valid's binary_logloss: 0.689284
regularization_factors, val_score: 0.688719:  70%|#######   | 14/20 [00:15<00:06,  1.08s/it][I 2020-09-27 04:50:05,336] Trial 56 finished with value: 0.6892843278626516 and parameters: {'lambda_l1': 0.04091667144851904, 'lambda_l2': 1.0750374226083052e-08}. Best is trial 54 with value: 0.6887188408518048.
regularization_factors, val_score: 0.688719:  70%|#######   | 14/20 [00:15<00:06,  1.08s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.013760 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  75%|#######5  | 15/20 [00:16<00:05,  1.04s/it][I 2020-09-27 04:50:06,270] Trial 57 finished with value: 0.6887188408561253 and parameters: {'lambda_l1': 1.5233181813455253e-07, 'lambda_l2': 1.8743684852533287e-07}. Best is trial 54 with value: 0.6887188408518048.
regularization_factors, val_score: 0.688719:  75%|#######5  | 15/20 [00:16<00:05,  1.04s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.015130 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  80%|########  | 16/20 [00:19<00:06,  1.55s/it][I 2020-09-27 04:50:09,004] Trial 58 finished with value: 0.6887188422885628 and parameters: {'lambda_l1': 1.744760689790351e-07, 'lambda_l2': 0.00013088268252619333}. Best is trial 54 with value: 0.6887188408518048.
regularization_factors, val_score: 0.688719:  80%|########  | 16/20 [00:19<00:06,  1.55s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.018473 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664768	valid's binary_logloss: 0.689166
Early stopping, best iteration is:
[85]	train's binary_logloss: 0.667917	valid's binary_logloss: 0.688911
regularization_factors, val_score: 0.688719:  85%|########5 | 17/20 [00:20<00:04,  1.45s/it][I 2020-09-27 04:50:10,235] Trial 59 finished with value: 0.6889111184760159 and parameters: {'lambda_l1': 0.00016154773417025348, 'lambda_l2': 0.9174839426871738}. Best is trial 54 with value: 0.6887188408518048.
regularization_factors, val_score: 0.688719:  85%|########5 | 17/20 [00:20<00:04,  1.45s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.015601 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  90%|######### | 18/20 [00:21<00:02,  1.32s/it][I 2020-09-27 04:50:11,235] Trial 60 finished with value: 0.6887188408571195 and parameters: {'lambda_l1': 1.0344224061447487e-07, 'lambda_l2': 3.115824250207294e-07}. Best is trial 54 with value: 0.6887188408518048.
regularization_factors, val_score: 0.688719:  90%|######### | 18/20 [00:21<00:02,  1.32s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.016453 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719:  95%|#########5| 19/20 [00:23<00:01,  1.34s/it][I 2020-09-27 04:50:12,647] Trial 61 finished with value: 0.6887188408520247 and parameters: {'lambda_l1': 1.3726047963300931e-08, 'lambda_l2': 1.1171463229442362e-08}. Best is trial 54 with value: 0.6887188408518048.
regularization_factors, val_score: 0.688719:  95%|#########5| 19/20 [00:23<00:01,  1.34s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.003196 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664324	valid's binary_logloss: 0.689515
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.675893	valid's binary_logloss: 0.688719
regularization_factors, val_score: 0.688719: 100%|##########| 20/20 [00:24<00:00,  1.26s/it][I 2020-09-27 04:50:13,695] Trial 62 finished with value: 0.688718840851812 and parameters: {'lambda_l1': 1.0200361074017904e-08, 'lambda_l2': 1.1298563130408035e-08}. Best is trial 54 with value: 0.6887188408518048.
regularization_factors, val_score: 0.688719: 100%|##########| 20/20 [00:24<00:00,  1.20s/it]
min_data_in_leaf, val_score: 0.688719:   0%|          | 0/5 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.013241 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664243	valid's binary_logloss: 0.689415
Early stopping, best iteration is:
[56]	train's binary_logloss: 0.674277	valid's binary_logloss: 0.689398
min_data_in_leaf, val_score: 0.688719:  20%|##        | 1/5 [00:00<00:03,  1.06it/s][I 2020-09-27 04:50:14,651] Trial 63 finished with value: 0.6893981290180272 and parameters: {'min_child_samples': 25}. Best is trial 63 with value: 0.6893981290180272.
min_data_in_leaf, val_score: 0.688719:  20%|##        | 1/5 [00:00<00:03,  1.06it/s][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012878 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.663871	valid's binary_logloss: 0.689525
[200]	train's binary_logloss: 0.643739	valid's binary_logloss: 0.690786
Early stopping, best iteration is:
[124]	train's binary_logloss: 0.658769	valid's binary_logloss: 0.689265
min_data_in_leaf, val_score: 0.688719:  40%|####      | 2/5 [00:02<00:03,  1.17s/it][I 2020-09-27 04:50:16,344] Trial 64 finished with value: 0.6892650557557729 and parameters: {'min_child_samples': 5}. Best is trial 64 with value: 0.6892650557557729.
min_data_in_leaf, val_score: 0.688719:  40%|####      | 2/5 [00:02<00:03,  1.17s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.013201 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664176	valid's binary_logloss: 0.689954
Early stopping, best iteration is:
[55]	train's binary_logloss: 0.674447	valid's binary_logloss: 0.689227
min_data_in_leaf, val_score: 0.688719:  60%|######    | 3/5 [00:03<00:02,  1.14s/it][I 2020-09-27 04:50:17,422] Trial 65 finished with value: 0.6892266404784428 and parameters: {'min_child_samples': 10}. Best is trial 65 with value: 0.6892266404784428.
min_data_in_leaf, val_score: 0.688719:  60%|######    | 3/5 [00:03<00:02,  1.14s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.013886 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.664683	valid's binary_logloss: 0.689526
Early stopping, best iteration is:
[69]	train's binary_logloss: 0.67151	valid's binary_logloss: 0.689297
min_data_in_leaf, val_score: 0.688719:  80%|########  | 4/5 [00:04<00:01,  1.13s/it][I 2020-09-27 04:50:18,524] Trial 66 finished with value: 0.689296784041159 and parameters: {'min_child_samples': 50}. Best is trial 65 with value: 0.6892266404784428.
min_data_in_leaf, val_score: 0.688719:  80%|########  | 4/5 [00:04<00:01,  1.13s/it][LightGBM] [Info] Number of positive: 46608, number of negative: 46417
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.013526 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4688
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.501027 -> initscore=0.004106
[LightGBM] [Info] Start training from score 0.004106
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.665727	valid's binary_logloss: 0.689486
Early stopping, best iteration is:
[74]	train's binary_logloss: 0.67097	valid's binary_logloss: 0.689201
min_data_in_leaf, val_score: 0.688719: 100%|##########| 5/5 [00:06<00:00,  1.19s/it][I 2020-09-27 04:50:19,869] Trial 67 finished with value: 0.689200711916551 and parameters: {'min_child_samples': 100}. Best is trial 67 with value: 0.689200711916551.
min_data_in_leaf, val_score: 0.688719: 100%|##########| 5/5 [00:06<00:00,  1.24s/it]
Fold : 1
[I 2020-09-27 04:50:19,990] A new study created in memory with name: no-name-42491456-2d3a-4d80-ab38-ce544f64ff2c
feature_fraction, val_score: inf:   0%|          | 0/7 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001692 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.665024	valid's binary_logloss: 0.689794
Early stopping, best iteration is:
[70]	train's binary_logloss: 0.671608	valid's binary_logloss: 0.689532
feature_fraction, val_score: 0.689532:  14%|#4        | 1/7 [00:01<00:07,  1.22s/it][I 2020-09-27 04:50:21,234] Trial 0 finished with value: 0.689531715864851 and parameters: {'feature_fraction': 0.5}. Best is trial 0 with value: 0.689531715864851.
feature_fraction, val_score: 0.689532:  14%|#4        | 1/7 [00:01<00:07,  1.22s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.012878 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.663919	valid's binary_logloss: 0.690116
Early stopping, best iteration is:
[40]	train's binary_logloss: 0.678175	valid's binary_logloss: 0.689468
feature_fraction, val_score: 0.689468:  29%|##8       | 2/7 [00:01<00:05,  1.09s/it][I 2020-09-27 04:50:22,004] Trial 1 finished with value: 0.6894679163318443 and parameters: {'feature_fraction': 0.6}. Best is trial 1 with value: 0.6894679163318443.
feature_fraction, val_score: 0.689468:  29%|##8       | 2/7 [00:01<00:05,  1.09s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.011687 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.66358	valid's binary_logloss: 0.690887
Early stopping, best iteration is:
[35]	train's binary_logloss: 0.679357	valid's binary_logloss: 0.690079
feature_fraction, val_score: 0.689468:  43%|####2     | 3/7 [00:02<00:03,  1.01it/s][I 2020-09-27 04:50:22,771] Trial 2 finished with value: 0.6900787550751419 and parameters: {'feature_fraction': 0.7}. Best is trial 1 with value: 0.6894679163318443.
feature_fraction, val_score: 0.689468:  43%|####2     | 3/7 [00:02<00:03,  1.01it/s][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.003338 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.662262	valid's binary_logloss: 0.690303
Early stopping, best iteration is:
[93]	train's binary_logloss: 0.664122	valid's binary_logloss: 0.689979
feature_fraction, val_score: 0.689468:  57%|#####7    | 4/7 [00:04<00:03,  1.18s/it][I 2020-09-27 04:50:24,375] Trial 3 finished with value: 0.6899793275720499 and parameters: {'feature_fraction': 1.0}. Best is trial 1 with value: 0.6894679163318443.
feature_fraction, val_score: 0.689468:  57%|#####7    | 4/7 [00:04<00:03,  1.18s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.021087 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.662515	valid's binary_logloss: 0.689981
Early stopping, best iteration is:
[51]	train's binary_logloss: 0.674455	valid's binary_logloss: 0.689748
feature_fraction, val_score: 0.689468:  71%|#######1  | 5/7 [00:05<00:02,  1.10s/it][I 2020-09-27 04:50:25,314] Trial 4 finished with value: 0.6897477331954484 and parameters: {'feature_fraction': 0.8999999999999999}. Best is trial 1 with value: 0.6894679163318443.
feature_fraction, val_score: 0.689468:  71%|#######1  | 5/7 [00:05<00:02,  1.10s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001352 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.666503	valid's binary_logloss: 0.689198
[200]	train's binary_logloss: 0.647913	valid's binary_logloss: 0.690111
Early stopping, best iteration is:
[141]	train's binary_logloss: 0.658411	valid's binary_logloss: 0.689014
feature_fraction, val_score: 0.689014:  86%|########5 | 6/7 [00:06<00:01,  1.12s/it][I 2020-09-27 04:50:26,470] Trial 5 finished with value: 0.6890136743093267 and parameters: {'feature_fraction': 0.4}. Best is trial 5 with value: 0.6890136743093267.
feature_fraction, val_score: 0.689014:  86%|########5 | 6/7 [00:06<00:01,  1.12s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.018804 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.66328	valid's binary_logloss: 0.690498
Early stopping, best iteration is:
[29]	train's binary_logloss: 0.681042	valid's binary_logloss: 0.689914
feature_fraction, val_score: 0.689014: 100%|##########| 7/7 [00:07<00:00,  1.02s/it][I 2020-09-27 04:50:27,266] Trial 6 finished with value: 0.6899136884834198 and parameters: {'feature_fraction': 0.8}. Best is trial 5 with value: 0.6890136743093267.
feature_fraction, val_score: 0.689014: 100%|##########| 7/7 [00:07<00:00,  1.04s/it]
num_leaves, val_score: 0.689014:   0%|          | 0/20 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001398 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.59502	valid's binary_logloss: 0.693529
Early stopping, best iteration is:
[25]	train's binary_logloss: 0.660259	valid's binary_logloss: 0.690335
num_leaves, val_score: 0.689014:   5%|5         | 1/20 [00:01<00:30,  1.59s/it][I 2020-09-27 04:50:28,875] Trial 7 finished with value: 0.6903350331651174 and parameters: {'num_leaves': 152}. Best is trial 7 with value: 0.6903350331651174.
num_leaves, val_score: 0.689014:   5%|5         | 1/20 [00:01<00:30,  1.59s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001113 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.621859	valid's binary_logloss: 0.693386
Early stopping, best iteration is:
[24]	train's binary_logloss: 0.669901	valid's binary_logloss: 0.690519
num_leaves, val_score: 0.689014:  10%|#         | 2/20 [00:02<00:25,  1.40s/it][I 2020-09-27 04:50:29,827] Trial 8 finished with value: 0.6905190321264499 and parameters: {'num_leaves': 103}. Best is trial 7 with value: 0.6903350331651174.
num_leaves, val_score: 0.689014:  10%|#         | 2/20 [00:02<00:25,  1.40s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001303 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.607653	valid's binary_logloss: 0.691023
Early stopping, best iteration is:
[31]	train's binary_logloss: 0.658737	valid's binary_logloss: 0.690547
num_leaves, val_score: 0.689014:  15%|#5        | 3/20 [00:03<00:22,  1.30s/it][I 2020-09-27 04:50:30,899] Trial 9 finished with value: 0.6905470890370582 and parameters: {'num_leaves': 128}. Best is trial 7 with value: 0.6903350331651174.
num_leaves, val_score: 0.689014:  15%|#5        | 3/20 [00:03<00:22,  1.30s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001257 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.68554	valid's binary_logloss: 0.689548
[200]	train's binary_logloss: 0.681807	valid's binary_logloss: 0.689004
[300]	train's binary_logloss: 0.678526	valid's binary_logloss: 0.688996
Early stopping, best iteration is:
[221]	train's binary_logloss: 0.681102	valid's binary_logloss: 0.688886
num_leaves, val_score: 0.688886:  20%|##        | 4/20 [00:05<00:23,  1.44s/it][I 2020-09-27 04:50:32,672] Trial 10 finished with value: 0.6888861673616007 and parameters: {'num_leaves': 6}. Best is trial 10 with value: 0.6888861673616007.
num_leaves, val_score: 0.688886:  20%|##        | 4/20 [00:05<00:23,  1.44s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001441 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.674935	valid's binary_logloss: 0.689151
[200]	train's binary_logloss: 0.663169	valid's binary_logloss: 0.689994
Early stopping, best iteration is:
[122]	train's binary_logloss: 0.672063	valid's binary_logloss: 0.689042
num_leaves, val_score: 0.688886:  25%|##5       | 5/20 [00:06<00:19,  1.33s/it][I 2020-09-27 04:50:33,728] Trial 11 finished with value: 0.6890424580437142 and parameters: {'num_leaves': 19}. Best is trial 10 with value: 0.6888861673616007.
num_leaves, val_score: 0.688886:  25%|##5       | 5/20 [00:06<00:19,  1.33s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001371 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.688532	valid's binary_logloss: 0.690109
[200]	train's binary_logloss: 0.686793	valid's binary_logloss: 0.689219
[300]	train's binary_logloss: 0.685512	valid's binary_logloss: 0.68876
[400]	train's binary_logloss: 0.684466	valid's binary_logloss: 0.688694
[500]	train's binary_logloss: 0.68348	valid's binary_logloss: 0.6886
Early stopping, best iteration is:
[438]	train's binary_logloss: 0.684084	valid's binary_logloss: 0.68855
num_leaves, val_score: 0.688550:  30%|###       | 6/20 [00:08<00:23,  1.66s/it][I 2020-09-27 04:50:36,153] Trial 12 finished with value: 0.6885499863950102 and parameters: {'num_leaves': 3}. Best is trial 12 with value: 0.6885499863950102.
num_leaves, val_score: 0.688550:  30%|###       | 6/20 [00:08<00:23,  1.66s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001342 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.68554	valid's binary_logloss: 0.689548
[200]	train's binary_logloss: 0.681807	valid's binary_logloss: 0.689004
[300]	train's binary_logloss: 0.678526	valid's binary_logloss: 0.688996
Early stopping, best iteration is:
[221]	train's binary_logloss: 0.681102	valid's binary_logloss: 0.688886
num_leaves, val_score: 0.688550:  35%|###5      | 7/20 [00:10<00:20,  1.54s/it][I 2020-09-27 04:50:37,431] Trial 13 finished with value: 0.6888861673616007 and parameters: {'num_leaves': 6}. Best is trial 12 with value: 0.6885499863950102.
num_leaves, val_score: 0.688550:  35%|###5      | 7/20 [00:10<00:20,  1.54s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000758 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.657781	valid's binary_logloss: 0.68976
Early stopping, best iteration is:
[90]	train's binary_logloss: 0.660403	valid's binary_logloss: 0.6895
num_leaves, val_score: 0.688550:  40%|####      | 8/20 [00:11<00:16,  1.36s/it][I 2020-09-27 04:50:38,375] Trial 14 finished with value: 0.6895000449518462 and parameters: {'num_leaves': 43}. Best is trial 12 with value: 0.6885499863950102.
num_leaves, val_score: 0.688550:  40%|####      | 8/20 [00:11<00:16,  1.36s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000790 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.64493	valid's binary_logloss: 0.691503
Early stopping, best iteration is:
[44]	train's binary_logloss: 0.667465	valid's binary_logloss: 0.690163
num_leaves, val_score: 0.688550:  45%|####5     | 9/20 [00:11<00:13,  1.20s/it][I 2020-09-27 04:50:39,181] Trial 15 finished with value: 0.6901631360080164 and parameters: {'num_leaves': 64}. Best is trial 12 with value: 0.6885499863950102.
num_leaves, val_score: 0.688550:  45%|####5     | 9/20 [00:11<00:13,  1.20s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000739 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.548818	valid's binary_logloss: 0.697085
Early stopping, best iteration is:
[20]	train's binary_logloss: 0.653082	valid's binary_logloss: 0.690873
num_leaves, val_score: 0.688550:  50%|#####     | 10/20 [00:13<00:12,  1.22s/it][I 2020-09-27 04:50:40,456] Trial 16 finished with value: 0.6908727065690785 and parameters: {'num_leaves': 247}. Best is trial 12 with value: 0.6885499863950102.
num_leaves, val_score: 0.688550:  50%|#####     | 10/20 [00:13<00:12,  1.22s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000719 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.690028	valid's binary_logloss: 0.690957
[200]	train's binary_logloss: 0.688879	valid's binary_logloss: 0.690086
[300]	train's binary_logloss: 0.688169	valid's binary_logloss: 0.689501
[400]	train's binary_logloss: 0.687668	valid's binary_logloss: 0.689075
[500]	train's binary_logloss: 0.687288	valid's binary_logloss: 0.688845
[600]	train's binary_logloss: 0.686991	valid's binary_logloss: 0.688707
[700]	train's binary_logloss: 0.68675	valid's binary_logloss: 0.688619
[800]	train's binary_logloss: 0.686547	valid's binary_logloss: 0.688562
[900]	train's binary_logloss: 0.686373	valid's binary_logloss: 0.688526
[1000]	train's binary_logloss: 0.686221	valid's binary_logloss: 0.688477
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.686221	valid's binary_logloss: 0.688477
num_leaves, val_score: 0.688477:  55%|#####5    | 11/20 [00:16<00:15,  1.75s/it][I 2020-09-27 04:50:43,436] Trial 17 finished with value: 0.6884769144658212 and parameters: {'num_leaves': 2}. Best is trial 17 with value: 0.6884769144658212.
num_leaves, val_score: 0.688477:  55%|#####5    | 11/20 [00:16<00:15,  1.75s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000850 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.573685	valid's binary_logloss: 0.696971
Early stopping, best iteration is:
[20]	train's binary_logloss: 0.660038	valid's binary_logloss: 0.691352
num_leaves, val_score: 0.688477:  60%|######    | 12/20 [00:17<00:12,  1.57s/it][I 2020-09-27 04:50:44,593] Trial 18 finished with value: 0.6913522035050282 and parameters: {'num_leaves': 195}. Best is trial 17 with value: 0.6884769144658212.
num_leaves, val_score: 0.688477:  60%|######    | 12/20 [00:17<00:12,  1.57s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000815 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.64326	valid's binary_logloss: 0.690631
Early stopping, best iteration is:
[71]	train's binary_logloss: 0.654351	valid's binary_logloss: 0.690055
num_leaves, val_score: 0.688477:  65%|######5   | 13/20 [00:18<00:09,  1.39s/it][I 2020-09-27 04:50:45,549] Trial 19 finished with value: 0.6900553800103405 and parameters: {'num_leaves': 66}. Best is trial 17 with value: 0.6884769144658212.
num_leaves, val_score: 0.688477:  65%|######5   | 13/20 [00:18<00:09,  1.39s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000857 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.68554	valid's binary_logloss: 0.689548
[200]	train's binary_logloss: 0.681807	valid's binary_logloss: 0.689004
[300]	train's binary_logloss: 0.678526	valid's binary_logloss: 0.688996
Early stopping, best iteration is:
[221]	train's binary_logloss: 0.681102	valid's binary_logloss: 0.688886
num_leaves, val_score: 0.688477:  70%|#######   | 14/20 [00:19<00:07,  1.30s/it][I 2020-09-27 04:50:46,632] Trial 20 finished with value: 0.6888861673616007 and parameters: {'num_leaves': 6}. Best is trial 17 with value: 0.6884769144658212.
num_leaves, val_score: 0.688477:  70%|#######   | 14/20 [00:19<00:07,  1.30s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000852 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.688532	valid's binary_logloss: 0.690109
[200]	train's binary_logloss: 0.686793	valid's binary_logloss: 0.689219
[300]	train's binary_logloss: 0.685512	valid's binary_logloss: 0.68876
[400]	train's binary_logloss: 0.684466	valid's binary_logloss: 0.688694
[500]	train's binary_logloss: 0.68348	valid's binary_logloss: 0.6886
Early stopping, best iteration is:
[438]	train's binary_logloss: 0.684084	valid's binary_logloss: 0.68855
num_leaves, val_score: 0.688477:  75%|#######5  | 15/20 [00:21<00:07,  1.42s/it][I 2020-09-27 04:50:48,340] Trial 21 finished with value: 0.6885499863950102 and parameters: {'num_leaves': 3}. Best is trial 17 with value: 0.6884769144658212.
num_leaves, val_score: 0.688477:  75%|#######5  | 15/20 [00:21<00:07,  1.42s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000875 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.660592	valid's binary_logloss: 0.690578
Early stopping, best iteration is:
[41]	train's binary_logloss: 0.676299	valid's binary_logloss: 0.690019
num_leaves, val_score: 0.688477:  80%|########  | 16/20 [00:21<00:04,  1.22s/it][I 2020-09-27 04:50:49,088] Trial 22 finished with value: 0.6900188884483167 and parameters: {'num_leaves': 39}. Best is trial 17 with value: 0.6884769144658212.
num_leaves, val_score: 0.688477:  80%|########  | 16/20 [00:21<00:04,  1.22s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000823 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.658652	valid's binary_logloss: 0.689977
Early stopping, best iteration is:
[49]	train's binary_logloss: 0.672748	valid's binary_logloss: 0.689541
num_leaves, val_score: 0.688477:  85%|########5 | 17/20 [00:22<00:03,  1.09s/it][I 2020-09-27 04:50:49,865] Trial 23 finished with value: 0.6895412052804837 and parameters: {'num_leaves': 42}. Best is trial 17 with value: 0.6884769144658212.
num_leaves, val_score: 0.688477:  85%|########5 | 17/20 [00:22<00:03,  1.09s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000785 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.688532	valid's binary_logloss: 0.690109
[200]	train's binary_logloss: 0.686793	valid's binary_logloss: 0.689219
[300]	train's binary_logloss: 0.685512	valid's binary_logloss: 0.68876
[400]	train's binary_logloss: 0.684466	valid's binary_logloss: 0.688694
[500]	train's binary_logloss: 0.68348	valid's binary_logloss: 0.6886
Early stopping, best iteration is:
[438]	train's binary_logloss: 0.684084	valid's binary_logloss: 0.68855
num_leaves, val_score: 0.688477:  90%|######### | 18/20 [00:24<00:02,  1.28s/it][I 2020-09-27 04:50:51,590] Trial 24 finished with value: 0.6885499863950102 and parameters: {'num_leaves': 3}. Best is trial 17 with value: 0.6884769144658212.
num_leaves, val_score: 0.688477:  90%|######### | 18/20 [00:24<00:02,  1.28s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000830 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.633629	valid's binary_logloss: 0.690048
Early stopping, best iteration is:
[41]	train's binary_logloss: 0.66326	valid's binary_logloss: 0.689286
num_leaves, val_score: 0.688477:  95%|#########5| 19/20 [00:25<00:01,  1.16s/it][I 2020-09-27 04:50:52,488] Trial 25 finished with value: 0.689285779471649 and parameters: {'num_leaves': 81}. Best is trial 17 with value: 0.6884769144658212.
num_leaves, val_score: 0.688477:  95%|#########5| 19/20 [00:25<00:01,  1.16s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.001270 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.669687	valid's binary_logloss: 0.68995
Early stopping, best iteration is:
[70]	train's binary_logloss: 0.674818	valid's binary_logloss: 0.689845
num_leaves, val_score: 0.688477: 100%|##########| 20/20 [00:25<00:00,  1.04s/it][I 2020-09-27 04:50:53,247] Trial 26 finished with value: 0.6898451259704299 and parameters: {'num_leaves': 26}. Best is trial 17 with value: 0.6884769144658212.
num_leaves, val_score: 0.688477: 100%|##########| 20/20 [00:25<00:00,  1.30s/it]
bagging, val_score: 0.688477:   0%|          | 0/10 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000794 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.689916	valid's binary_logloss: 0.690872
[200]	train's binary_logloss: 0.688688	valid's binary_logloss: 0.689969
[300]	train's binary_logloss: 0.687926	valid's binary_logloss: 0.689254
[400]	train's binary_logloss: 0.687375	valid's binary_logloss: 0.688976
[500]	train's binary_logloss: 0.686964	valid's binary_logloss: 0.688668
[600]	train's binary_logloss: 0.686637	valid's binary_logloss: 0.688603
[700]	train's binary_logloss: 0.686361	valid's binary_logloss: 0.688503
[800]	train's binary_logloss: 0.686138	valid's binary_logloss: 0.688427
[900]	train's binary_logloss: 0.68594	valid's binary_logloss: 0.68828
Early stopping, best iteration is:
[893]	train's binary_logloss: 0.685953	valid's binary_logloss: 0.688266
bagging, val_score: 0.688266:  10%|#         | 1/10 [00:03<00:29,  3.27s/it][I 2020-09-27 04:50:56,524] Trial 27 finished with value: 0.6882656720836928 and parameters: {'bagging_fraction': 0.9056687704253129, 'bagging_freq': 1}. Best is trial 27 with value: 0.6882656720836928.
bagging, val_score: 0.688266:  10%|#         | 1/10 [00:03<00:29,  3.27s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000821 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.689949	valid's binary_logloss: 0.690893
[200]	train's binary_logloss: 0.688731	valid's binary_logloss: 0.689897
[300]	train's binary_logloss: 0.687963	valid's binary_logloss: 0.689438
[400]	train's binary_logloss: 0.687422	valid's binary_logloss: 0.688923
[500]	train's binary_logloss: 0.687012	valid's binary_logloss: 0.688674
[600]	train's binary_logloss: 0.686692	valid's binary_logloss: 0.688589
[700]	train's binary_logloss: 0.686428	valid's binary_logloss: 0.688421
[800]	train's binary_logloss: 0.686204	valid's binary_logloss: 0.688439
[900]	train's binary_logloss: 0.686006	valid's binary_logloss: 0.688398
Early stopping, best iteration is:
[868]	train's binary_logloss: 0.686066	valid's binary_logloss: 0.68835
bagging, val_score: 0.688266:  20%|##        | 2/10 [00:06<00:25,  3.24s/it][I 2020-09-27 04:50:59,688] Trial 28 finished with value: 0.68834951195225 and parameters: {'bagging_fraction': 0.9341933335925237, 'bagging_freq': 1}. Best is trial 27 with value: 0.6882656720836928.
bagging, val_score: 0.688266:  20%|##        | 2/10 [00:06<00:25,  3.24s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000760 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.689958	valid's binary_logloss: 0.690901
[200]	train's binary_logloss: 0.688753	valid's binary_logloss: 0.689921
[300]	train's binary_logloss: 0.687991	valid's binary_logloss: 0.68943
[400]	train's binary_logloss: 0.687457	valid's binary_logloss: 0.688954
[500]	train's binary_logloss: 0.687054	valid's binary_logloss: 0.688639
[600]	train's binary_logloss: 0.686734	valid's binary_logloss: 0.688575
[700]	train's binary_logloss: 0.686471	valid's binary_logloss: 0.688404
[800]	train's binary_logloss: 0.686247	valid's binary_logloss: 0.688399
[900]	train's binary_logloss: 0.686054	valid's binary_logloss: 0.688395
[1000]	train's binary_logloss: 0.685884	valid's binary_logloss: 0.688386
Did not meet early stopping. Best iteration is:
[1000]	train's binary_logloss: 0.685884	valid's binary_logloss: 0.688386
bagging, val_score: 0.688266:  30%|###       | 3/10 [00:09<00:22,  3.22s/it][I 2020-09-27 04:51:02,884] Trial 29 finished with value: 0.6883861307718222 and parameters: {'bagging_fraction': 0.9472695270258076, 'bagging_freq': 1}. Best is trial 27 with value: 0.6882656720836928.
bagging, val_score: 0.688266:  30%|###       | 3/10 [00:09<00:22,  3.22s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000818 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.689959	valid's binary_logloss: 0.690908
[200]	train's binary_logloss: 0.688754	valid's binary_logloss: 0.689942
[300]	train's binary_logloss: 0.68799	valid's binary_logloss: 0.689383
[400]	train's binary_logloss: 0.687454	valid's binary_logloss: 0.688954
[500]	train's binary_logloss: 0.687052	valid's binary_logloss: 0.688747
[600]	train's binary_logloss: 0.686731	valid's binary_logloss: 0.68862
[700]	train's binary_logloss: 0.68647	valid's binary_logloss: 0.688482
[800]	train's binary_logloss: 0.686248	valid's binary_logloss: 0.68848
Early stopping, best iteration is:
[725]	train's binary_logloss: 0.686411	valid's binary_logloss: 0.68843
bagging, val_score: 0.688266:  40%|####      | 4/10 [00:12<00:18,  3.05s/it][I 2020-09-27 04:51:05,526] Trial 30 finished with value: 0.6884302974689744 and parameters: {'bagging_fraction': 0.9478078952145158, 'bagging_freq': 1}. Best is trial 27 with value: 0.6882656720836928.
bagging, val_score: 0.688266:  40%|####      | 4/10 [00:12<00:18,  3.05s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000815 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.689962	valid's binary_logloss: 0.690903
[200]	train's binary_logloss: 0.688757	valid's binary_logloss: 0.690012
[300]	train's binary_logloss: 0.688001	valid's binary_logloss: 0.689453
[400]	train's binary_logloss: 0.687468	valid's binary_logloss: 0.688994
[500]	train's binary_logloss: 0.687058	valid's binary_logloss: 0.688733
[600]	train's binary_logloss: 0.686738	valid's binary_logloss: 0.688599
[700]	train's binary_logloss: 0.686473	valid's binary_logloss: 0.688496
[800]	train's binary_logloss: 0.686255	valid's binary_logloss: 0.688458
[900]	train's binary_logloss: 0.686061	valid's binary_logloss: 0.688445
Early stopping, best iteration is:
[867]	train's binary_logloss: 0.686119	valid's binary_logloss: 0.688391
bagging, val_score: 0.688266:  50%|#####     | 5/10 [00:15<00:15,  3.08s/it][I 2020-09-27 04:51:08,673] Trial 31 finished with value: 0.6883910883412938 and parameters: {'bagging_fraction': 0.9500225140884224, 'bagging_freq': 1}. Best is trial 27 with value: 0.6882656720836928.
bagging, val_score: 0.688266:  50%|#####     | 5/10 [00:15<00:15,  3.08s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000836 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.689952	valid's binary_logloss: 0.690877
[200]	train's binary_logloss: 0.688739	valid's binary_logloss: 0.689932
[300]	train's binary_logloss: 0.687981	valid's binary_logloss: 0.689456
[400]	train's binary_logloss: 0.687443	valid's binary_logloss: 0.688946
[500]	train's binary_logloss: 0.687035	valid's binary_logloss: 0.688722
[600]	train's binary_logloss: 0.686712	valid's binary_logloss: 0.688567
[700]	train's binary_logloss: 0.686447	valid's binary_logloss: 0.688364
[800]	train's binary_logloss: 0.686224	valid's binary_logloss: 0.688372
Early stopping, best iteration is:
[701]	train's binary_logloss: 0.686445	valid's binary_logloss: 0.688354
bagging, val_score: 0.688266:  60%|######    | 6/10 [00:18<00:11,  2.94s/it][I 2020-09-27 04:51:11,277] Trial 32 finished with value: 0.6883544364816402 and parameters: {'bagging_fraction': 0.939276221019466, 'bagging_freq': 1}. Best is trial 27 with value: 0.6882656720836928.
bagging, val_score: 0.688266:  60%|######    | 6/10 [00:18<00:11,  2.94s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000816 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.689945	valid's binary_logloss: 0.690873
[200]	train's binary_logloss: 0.688736	valid's binary_logloss: 0.68993
[300]	train's binary_logloss: 0.687975	valid's binary_logloss: 0.689441
[400]	train's binary_logloss: 0.687436	valid's binary_logloss: 0.688977
[500]	train's binary_logloss: 0.687029	valid's binary_logloss: 0.68869
[600]	train's binary_logloss: 0.68671	valid's binary_logloss: 0.688599
[700]	train's binary_logloss: 0.686444	valid's binary_logloss: 0.688457
[800]	train's binary_logloss: 0.686223	valid's binary_logloss: 0.688493
Early stopping, best iteration is:
[700]	train's binary_logloss: 0.686444	valid's binary_logloss: 0.688457
bagging, val_score: 0.688266:  70%|#######   | 7/10 [00:20<00:08,  2.83s/it][I 2020-09-27 04:51:13,857] Trial 33 finished with value: 0.6884573747067545 and parameters: {'bagging_fraction': 0.9386192260753653, 'bagging_freq': 1}. Best is trial 27 with value: 0.6882656720836928.
bagging, val_score: 0.688266:  70%|#######   | 7/10 [00:20<00:08,  2.83s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000907 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.689934	valid's binary_logloss: 0.690812
[200]	train's binary_logloss: 0.688717	valid's binary_logloss: 0.689892
[300]	train's binary_logloss: 0.687955	valid's binary_logloss: 0.689321
[400]	train's binary_logloss: 0.687414	valid's binary_logloss: 0.688937
[500]	train's binary_logloss: 0.686999	valid's binary_logloss: 0.688627
[600]	train's binary_logloss: 0.686678	valid's binary_logloss: 0.68863
[700]	train's binary_logloss: 0.686413	valid's binary_logloss: 0.688396
[800]	train's binary_logloss: 0.68619	valid's binary_logloss: 0.688375
[900]	train's binary_logloss: 0.685991	valid's binary_logloss: 0.688321
Early stopping, best iteration is:
[867]	train's binary_logloss: 0.686053	valid's binary_logloss: 0.688263
bagging, val_score: 0.688263:  80%|########  | 8/10 [00:23<00:05,  2.91s/it][I 2020-09-27 04:51:16,952] Trial 34 finished with value: 0.6882634163638113 and parameters: {'bagging_fraction': 0.9270871288128361, 'bagging_freq': 1}. Best is trial 34 with value: 0.6882634163638113.
bagging, val_score: 0.688263:  80%|########  | 8/10 [00:23<00:05,  2.91s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000826 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.689852	valid's binary_logloss: 0.690627
[200]	train's binary_logloss: 0.688571	valid's binary_logloss: 0.689877
[300]	train's binary_logloss: 0.687776	valid's binary_logloss: 0.68924
[400]	train's binary_logloss: 0.687205	valid's binary_logloss: 0.688859
[500]	train's binary_logloss: 0.686782	valid's binary_logloss: 0.68858
[600]	train's binary_logloss: 0.686448	valid's binary_logloss: 0.688462
[700]	train's binary_logloss: 0.686171	valid's binary_logloss: 0.688201
Early stopping, best iteration is:
[699]	train's binary_logloss: 0.686173	valid's binary_logloss: 0.688186
bagging, val_score: 0.688186:  90%|######### | 9/10 [00:26<00:02,  2.83s/it][I 2020-09-27 04:51:19,581] Trial 35 finished with value: 0.6881860644541565 and parameters: {'bagging_fraction': 0.797756062798586, 'bagging_freq': 1}. Best is trial 35 with value: 0.6881860644541565.
bagging, val_score: 0.688186:  90%|######### | 9/10 [00:26<00:02,  2.83s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000754 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.689842	valid's binary_logloss: 0.690437
[200]	train's binary_logloss: 0.688572	valid's binary_logloss: 0.689737
[300]	train's binary_logloss: 0.687777	valid's binary_logloss: 0.68909
[400]	train's binary_logloss: 0.687207	valid's binary_logloss: 0.688463
[500]	train's binary_logloss: 0.686786	valid's binary_logloss: 0.688389
Early stopping, best iteration is:
[463]	train's binary_logloss: 0.686927	valid's binary_logloss: 0.688299
bagging, val_score: 0.688186: 100%|##########| 10/10 [00:28<00:00,  2.52s/it][I 2020-09-27 04:51:21,392] Trial 36 finished with value: 0.6882991425357883 and parameters: {'bagging_fraction': 0.7512163390719674, 'bagging_freq': 4}. Best is trial 35 with value: 0.6881860644541565.
bagging, val_score: 0.688186: 100%|##########| 10/10 [00:28<00:00,  2.81s/it]
feature_fraction_stage2, val_score: 0.688186:   0%|          | 0/3 [00:00<?, ?it/s][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000836 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.689843	valid's binary_logloss: 0.690616
[200]	train's binary_logloss: 0.688551	valid's binary_logloss: 0.689846
[300]	train's binary_logloss: 0.687748	valid's binary_logloss: 0.689146
[400]	train's binary_logloss: 0.687179	valid's binary_logloss: 0.688875
[500]	train's binary_logloss: 0.686761	valid's binary_logloss: 0.688605
[600]	train's binary_logloss: 0.686432	valid's binary_logloss: 0.68842
[700]	train's binary_logloss: 0.686158	valid's binary_logloss: 0.688213
Early stopping, best iteration is:
[699]	train's binary_logloss: 0.686161	valid's binary_logloss: 0.68819
feature_fraction_stage2, val_score: 0.688186:  33%|###3      | 1/3 [00:02<00:05,  2.77s/it][I 2020-09-27 04:51:24,172] Trial 37 finished with value: 0.6881903702314308 and parameters: {'feature_fraction': 0.41600000000000004}. Best is trial 37 with value: 0.6881903702314308.
feature_fraction_stage2, val_score: 0.688186:  33%|###3      | 1/3 [00:02<00:05,  2.77s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000871 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.689828	valid's binary_logloss: 0.690608
[200]	train's binary_logloss: 0.688553	valid's binary_logloss: 0.689888
[300]	train's binary_logloss: 0.687744	valid's binary_logloss: 0.689142
[400]	train's binary_logloss: 0.687173	valid's binary_logloss: 0.688816
[500]	train's binary_logloss: 0.686758	valid's binary_logloss: 0.688522
[600]	train's binary_logloss: 0.686423	valid's binary_logloss: 0.688429
[700]	train's binary_logloss: 0.686151	valid's binary_logloss: 0.688287
Early stopping, best iteration is:
[697]	train's binary_logloss: 0.686159	valid's binary_logloss: 0.688254
feature_fraction_stage2, val_score: 0.688186:  67%|######6   | 2/3 [00:05<00:02,  2.75s/it][I 2020-09-27 04:51:26,877] Trial 38 finished with value: 0.6882536571755299 and parameters: {'feature_fraction': 0.44800000000000006}. Best is trial 37 with value: 0.6881903702314308.
feature_fraction_stage2, val_score: 0.688186:  67%|######6   | 2/3 [00:05<00:02,  2.75s/it][LightGBM] [Info] Number of positive: 46746, number of negative: 46279
[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.000812 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 4689
[LightGBM] [Info] Number of data points in the train set: 93025, number of used features: 26
[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.502510 -> initscore=0.010040
[LightGBM] [Info] Start training from score 0.010040
Training until validation scores don't improve for 100 rounds
[100]	train's binary_logloss: 0.689828	valid's binary_logloss: 0.690608
[200]	train's binary_logloss: 0.688553	valid's binary_logloss: 0.689888
[300]	train's binary_logloss: 0.687744	valid's binary_logloss: 0.689142
[400]	train's binary_logloss: 0.687173	valid's binary_logloss: 0.688816
[500]	train's binary_logloss: 0.686758	valid's binary_logloss: 0.688522
[600]	train's binary_logloss: 0.686423	valid's binary_logloss: 0.688429
[700]	train's binary_logloss: 0.686151	valid's binary_logloss: 0.688287
Early stopping, best iteration is:
[697]	train's binary_logloss: 0.686159	valid's binary_logloss: 0.688254
feature_fraction_stage2, val_score: 0.688186: 100%|##########| 3/3 [00:08<00:00,  2