Competition/Kaggle
[kaggle][필사] Costa Rican Household Proverty (3)
bisi
2020. 9. 22. 12:44
이번 주제는 Costa Rican Household Proverty 로,
목표는 미주 개발 은행(Inter-American Development Bank)의 세계에서 가장 빈곤 한 일부 가정의 소득 자격을 예측을 하는 것이다.
보통 세계 최 빈곤층은 그들의 자격을 증명하기가 어려운데, 라틴 아메리카는 알고리즘을 통해 소득자격을 확인한다.
예를 들어 프록시 수단 테스트(PMT)을 통해 벽과 천장의 재료 또는 집에서 발견 된 자산과 같은 가족의 관찰 가능한 가구 속성을 고려하는 것이다.
이를 바탕으로 다양한 feature가 제공 되었는데, LGBMClassifier를 사용하여 소득 자격을 예측해본다.
이번 필사는 이유한님의 코드를 참고하였다.
목록
Costa Rican Household Proverty (1)
더보기
1. Check datasets
1.1 Read datasets
1.2 Make description df
1.3 Check null data
1.4 Fill missing values
Costa Rican Household Proverty (2)
더보기
2. Feature Engineering
2.1 Object features
2.1.1 dependency
2.1.2 edjefe
2.1.3 edjefa
2.1.4 roof and elecricity
2.2 카테고리 변수 추출
2.3 연속형 변수를 사용하여 새로운 변수 생성
2.3.1 연속형 변수 컬럼 추출
2.3.2 새로운 변수 생성
2.3.3 가족 변수의 대출 비율
2.3.4 가족 변수의 방 비율
2.3.5 가족 변수의 침대 비율
2.3.6 가족 변수의 태블릿 보유 비율
2.3.7 가족 변수의 핸드폰 보유 비율
2.3.8 가족 변수의 학창 시절의 몇년뒤
2.3.9 Rich features
2.4 집합 변수
Costa Rican Household Proverty (3)
더보기
3. Feature Selection Using shap
4. Model Development
4.1 LGB를 통한 예측 및 변수 중요도 생성
4.2 랜덤하게 찾기 (Randomized Search)
Costa Rican Household Poverty Level Prediction
3. Feature Selection Using Shap
In [94]:
# 중복되는 열이 있으므로, 중복되는 열을 삭제하고 binary category 변수를 생성한다.
# train = train.T.drop_duplicates().T
binary_cat_features = [col for col in train.columns if train[col].value_counts().shape[0] == 2]
object_features = ['edjefe', 'edjefa']
categorical_feats = binary_cat_features + object_features
In [95]:
def evaluate_macroF1_lgb(truth, predictions):
# this follows the discussion in https://github.com/Microsoft/LightGBM/issues/1483
pred_labels = predictions.reshape(len(np.unique(truth)),-1).argmax(axis=0)
f1 = f1_score(truth, pred_labels, average='macro')
return ('macroF1', f1, True)
In [96]:
y = train['Target']
train.drop(columns=['Target'], inplace=True)
In [97]:
def print_execution_time(start):
end = time.time()
hours, rem = divmod(end-start, 3600)
minutes, seconds = divmod(rem, 60)
print('*'*20, "Execution ended in {:0>2}h {:0>2}m {:05.2f}s".format(int(hours),int(minutes),seconds), '*'*20)
In [98]:
# LGBMClassifier https://lightgbm.readthedocs.io/en/latest/Parameters-Tuning.html
# def extract_good_features_using_shap_LGB(params, SEED):
# clf = lgb.LGBMClassifier(objective='multiclass',
# random_state=1989,
# max_depth=params['max_depth'],#max_depth:트리의 최대 깊이
# learning_rate=params['learning_rate'],#learning_rate : 학습률
# silent=True,
# metric='multi_logloss',
# n_jobs=-1, n_estimators =10000,#n_estimators : 반복하려는 트리의 개수
# class_weight = 'balanced',
# colsample_bytree=params['colsample_bytree'],
# min_split_gain=params['min_split_gain'],
# bagging_freq=params['bagging_freq'],
# min_child_weight=params['min_child_weight'],#min_child_samples : 리프 노드가 되기 위한 최소한의 샘플 데이터 수
# num_leaves=params['num_leaves'],#num_leaves:하나의 트리가 가질수 있는 최대의 리프 개수
# subsample=params['subsample'],
# reg_alpha=params['reg_alpha'],#reg_alpha:L2regularization
# reg_lambda=params['reg_lambda'],#reg_lambda:L1regularization
# num_class=len(np.unique(y)),
# bagging_seed=SEED,
# seed=SEED
# )
# kfold = 5
# kf = StratifiedKFold(n_splits=kfold, shuffle=True)
# feat_importance_df = pd.DataFrame()
# for i, (train_index, test_index) in enumerate(kf.split(train, y)): # y : Target
# print('='*30,'{} of {} folds'.format(i+1, kfold), '='*30)
# start = time.time()
# X_train, X_val = train.iloc[train_index], train_iloc[test_index]
# y_train, y_val = y.iloc[train_index], y.iloc[testst_index]
# clf.fit(X_train, y_train, eval_set=[(X_train, y_train), (X_val, y_val)], eval_metric=evaluate_macroF1_lgb,
# categorical_feature=categorical_feats, early_stopping_rounds=500, verbose=500)
# shap_values=shap.TreeExplainer(clf.booster_).shap_values(X_train)
# fold_importance_df = pd.DataFrame()
# fold_importance_df['feature'] = X_train.columns
# fold_importance_df['shap_values'] = abs(np.array(shap_values)[:,:].mean(1).mean(0))
# fold_importance_df['feat_imp'] = clf.feature_importances_
# feat_importance_df = pd.concat([feat_importance_df, fold_importance_df])
# print_execution_time(start)
# feat_importance_df_shap = feat_importance_df.groupby('feature').mean().sort_values('shap_values', ascendig=False).reset_index()
# return feat_importance_df_shap
In [99]:
def extract_good_features_using_shap_LGB(params, SEED):
clf = lgb.LGBMClassifier(objective='multiclass',
random_state=1989,
max_depth=params['max_depth'],
learning_rate=params['learning_rate'],
silent=True,
metric='multi_logloss',
n_jobs=-1, n_estimators=10000,
class_weight='balanced',
colsample_bytree = params['colsample_bytree'],
min_split_gain= params['min_split_gain'],
bagging_freq = params['bagging_freq'],
min_child_weight=params['min_child_weight'],
num_leaves = params['num_leaves'],
subsample = params['subsample'],
reg_alpha= params['reg_alpha'],
reg_lambda= params['reg_lambda'],
num_class=len(np.unique(y)),
bagging_seed=SEED,
seed=SEED,
)
kfold = 5
kf = StratifiedKFold(n_splits=kfold, shuffle=True)
feat_importance_df = pd.DataFrame()
for i, (train_index, test_index) in enumerate(kf.split(train, y)):
print('='*30, '{} of {} folds'.format(i+1, kfold), '='*30)
start = time.time()
X_train, X_val = train.iloc[train_index], train.iloc[test_index]
y_train, y_val = y.iloc[train_index], y.iloc[test_index]
clf.fit(X_train, y_train, eval_set=[(X_train, y_train), (X_val, y_val)], eval_metric=evaluate_macroF1_lgb, categorical_feature=categorical_feats,
early_stopping_rounds=500, verbose=500)
shap_values = shap.TreeExplainer(clf.booster_).shap_values(X_train)
fold_importance_df = pd.DataFrame()
fold_importance_df['feature'] = X_train.columns
fold_importance_df['shap_values'] = abs(np.array(shap_values)[:, :].mean(1).mean(0))
fold_importance_df['feat_imp'] = clf.feature_importances_
feat_importance_df = pd.concat([feat_importance_df, fold_importance_df])
print_execution_time(start)
feat_importance_df_shap = feat_importance_df.groupby('feature').mean().sort_values('shap_values', ascending=False).reset_index()
# feat_importance_df_shap['shap_cumsum'] = feat_importance_df_shap['shap_values'].cumsum() / feat_importance_df_shap['shap_values'].sum()
# good_features = feat_importance_df_shap.loc[feat_importance_df_shap['shap_cumsum'] < 0.999].feature
return feat_importance_df_shap
In [100]:
total_shap_df = pd.DataFrame()
NUM_ITERATIONS = 50
for SEED in range(NUM_ITERATIONS):
print('#'*40, '{} of {} iterations'.format(SEED+1, NUM_ITERATIONS), '#' * 40)
params = {'max_depth': np.random.choice([5, 6, 7, 8, 10, 12, -1]),
'learning_rate': np.random.rand() * 0.02,
'colsample_bytree': np.random.rand() * (1 - 0.5) + 0.5,
'subsample': np.random.rand() * (1 - 0.5) + 0.5,
'min_split_gain': np.random.rand() * 0.2,
'num_leaves': np.random.choice([32, 48, 64]),
'reg_alpha': np.random.rand() * 2,
'reg_lambda': np.random.rand() *2,
'bagging_freq': np.random.randint(4) +1,
'min_child_weight': np.random.randint(100) + 20
}
temp_shap_df = extract_good_features_using_shap_LGB(params, SEED)
total_shap_df = pd.concat([total_shap_df, temp_shap_df])
######################################## 1 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.00485 training's macroF1: 0.557042 valid_1's multi_logloss: 1.04836 valid_1's macroF1: 0.377357
Early stopping, best iteration is:
[49] training's multi_logloss: 1.30129 training's macroF1: 0.487104 valid_1's multi_logloss: 1.27716 valid_1's macroF1: 0.387977
******************** Execution ended in 00h 00m 20.92s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.00997 training's macroF1: 0.560006 valid_1's multi_logloss: 1.08823 valid_1's macroF1: 0.38931
[1000] training's multi_logloss: 0.895372 training's macroF1: 0.618211 valid_1's multi_logloss: 1.07981 valid_1's macroF1: 0.397377
[1500] training's multi_logloss: 0.818834 training's macroF1: 0.662662 valid_1's multi_logloss: 1.07889 valid_1's macroF1: 0.390747
Early stopping, best iteration is:
[1032] training's multi_logloss: 0.889825 training's macroF1: 0.620445 valid_1's multi_logloss: 1.07929 valid_1's macroF1: 0.401916
******************** Execution ended in 00h 00m 57.44s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.01797 training's macroF1: 0.555856 valid_1's multi_logloss: 1.01584 valid_1's macroF1: 0.454784
[1000] training's multi_logloss: 0.903931 training's macroF1: 0.605618 valid_1's multi_logloss: 0.986396 valid_1's macroF1: 0.459385
Early stopping, best iteration is:
[701] training's multi_logloss: 0.964902 training's macroF1: 0.578769 valid_1's multi_logloss: 0.996749 valid_1's macroF1: 0.471793
******************** Execution ended in 00h 00m 46.32s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.02129 training's macroF1: 0.561591 valid_1's multi_logloss: 1.0278 valid_1's macroF1: 0.422317
Early stopping, best iteration is:
[388] training's multi_logloss: 1.06019 training's macroF1: 0.550715 valid_1's multi_logloss: 1.04633 valid_1's macroF1: 0.436703
******************** Execution ended in 00h 00m 34.18s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.01423 training's macroF1: 0.56672 valid_1's multi_logloss: 1.07017 valid_1's macroF1: 0.411672
Early stopping, best iteration is:
[462] training's multi_logloss: 1.02643 training's macroF1: 0.56065 valid_1's multi_logloss: 1.07366 valid_1's macroF1: 0.416939
******************** Execution ended in 00h 00m 37.52s ********************
######################################## 2 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15349 training's macroF1: 0.494975 valid_1's multi_logloss: 1.12155 valid_1's macroF1: 0.367781
[1000] training's multi_logloss: 1.06031 training's macroF1: 0.530256 valid_1's multi_logloss: 1.0632 valid_1's macroF1: 0.374037
Early stopping, best iteration is:
[562] training's multi_logloss: 1.13825 training's macroF1: 0.5007 valid_1's multi_logloss: 1.1094 valid_1's macroF1: 0.381025
******************** Execution ended in 00h 01m 06.75s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15701 training's macroF1: 0.492258 valid_1's multi_logloss: 1.13662 valid_1's macroF1: 0.417573
[1000] training's multi_logloss: 1.06583 training's macroF1: 0.52717 valid_1's multi_logloss: 1.08027 valid_1's macroF1: 0.420325
Early stopping, best iteration is:
[782] training's multi_logloss: 1.09846 training's macroF1: 0.51704 valid_1's multi_logloss: 1.09675 valid_1's macroF1: 0.432322
******************** Execution ended in 00h 01m 20.97s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15252 training's macroF1: 0.495734 valid_1's multi_logloss: 1.14348 valid_1's macroF1: 0.40653
[1000] training's multi_logloss: 1.05594 training's macroF1: 0.523409 valid_1's multi_logloss: 1.09 valid_1's macroF1: 0.390093
Early stopping, best iteration is:
[722] training's multi_logloss: 1.10135 training's macroF1: 0.516633 valid_1's multi_logloss: 1.1108 valid_1's macroF1: 0.414585
******************** Execution ended in 00h 01m 15.52s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15287 training's macroF1: 0.494296 valid_1's multi_logloss: 1.13193 valid_1's macroF1: 0.384456
[1000] training's multi_logloss: 1.05863 training's macroF1: 0.530018 valid_1's multi_logloss: 1.07259 valid_1's macroF1: 0.390425
[1500] training's multi_logloss: 1.00248 training's macroF1: 0.557114 valid_1's multi_logloss: 1.05278 valid_1's macroF1: 0.385397
Early stopping, best iteration is:
[1167] training's multi_logloss: 1.03744 training's macroF1: 0.540135 valid_1's multi_logloss: 1.06327 valid_1's macroF1: 0.39476
******************** Execution ended in 00h 01m 43.83s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.162 training's macroF1: 0.494477 valid_1's multi_logloss: 1.1371 valid_1's macroF1: 0.421501
[1000] training's multi_logloss: 1.06962 training's macroF1: 0.518801 valid_1's multi_logloss: 1.07597 valid_1's macroF1: 0.424294
Early stopping, best iteration is:
[606] training's multi_logloss: 1.137 training's macroF1: 0.502814 valid_1's multi_logloss: 1.11745 valid_1's macroF1: 0.431822
******************** Execution ended in 00h 01m 13.06s ********************
######################################## 3 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.809865 training's macroF1: 0.70056 valid_1's multi_logloss: 0.995402 valid_1's macroF1: 0.444196
Early stopping, best iteration is:
[490] training's multi_logloss: 0.81464 training's macroF1: 0.696039 valid_1's multi_logloss: 0.995938 valid_1's macroF1: 0.451588
******************** Execution ended in 00h 01m 05.31s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.825917 training's macroF1: 0.681596 valid_1's multi_logloss: 1.02221 valid_1's macroF1: 0.442533
Early stopping, best iteration is:
[310] training's multi_logloss: 0.929943 training's macroF1: 0.635213 valid_1's multi_logloss: 1.04201 valid_1's macroF1: 0.454028
******************** Execution ended in 00h 00m 53.24s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.827298 training's macroF1: 0.669178 valid_1's multi_logloss: 0.968586 valid_1's macroF1: 0.441924
Early stopping, best iteration is:
[344] training's multi_logloss: 0.910514 training's macroF1: 0.634273 valid_1's multi_logloss: 0.98736 valid_1's macroF1: 0.45069
******************** Execution ended in 00h 00m 56.89s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.820977 training's macroF1: 0.671463 valid_1's multi_logloss: 1.03319 valid_1's macroF1: 0.380203
Early stopping, best iteration is:
[3] training's multi_logloss: 1.37461 training's macroF1: 0.495498 valid_1's multi_logloss: 1.37252 valid_1's macroF1: 0.394692
******************** Execution ended in 00h 00m 33.10s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.814642 training's macroF1: 0.685804 valid_1's multi_logloss: 1.06496 valid_1's macroF1: 0.421855
[1000] training's multi_logloss: 0.647217 training's macroF1: 0.759141 valid_1's multi_logloss: 1.06244 valid_1's macroF1: 0.414225
Early stopping, best iteration is:
[521] training's multi_logloss: 0.805173 training's macroF1: 0.693018 valid_1's multi_logloss: 1.06418 valid_1's macroF1: 0.422102
******************** Execution ended in 00h 01m 10.16s ********************
######################################## 4 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.01083 training's macroF1: 0.537277 valid_1's multi_logloss: 1.05847 valid_1's macroF1: 0.40936
Early stopping, best iteration is:
[379] training's multi_logloss: 1.04687 training's macroF1: 0.523113 valid_1's multi_logloss: 1.0641 valid_1's macroF1: 0.428405
******************** Execution ended in 00h 00m 44.50s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.01284 training's macroF1: 0.543027 valid_1's multi_logloss: 1.02977 valid_1's macroF1: 0.423176
[1000] training's multi_logloss: 0.913201 training's macroF1: 0.596816 valid_1's multi_logloss: 1.01665 valid_1's macroF1: 0.426669
[1500] training's multi_logloss: 0.845772 training's macroF1: 0.628095 valid_1's multi_logloss: 1.01106 valid_1's macroF1: 0.427882
Early stopping, best iteration is:
[1252] training's multi_logloss: 0.876021 training's macroF1: 0.615787 valid_1's multi_logloss: 1.01373 valid_1's macroF1: 0.434092
******************** Execution ended in 00h 01m 22.45s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.01234 training's macroF1: 0.548258 valid_1's multi_logloss: 1.08506 valid_1's macroF1: 0.369514
[1000] training's multi_logloss: 0.911285 training's macroF1: 0.607829 valid_1's multi_logloss: 1.07324 valid_1's macroF1: 0.382539
Early stopping, best iteration is:
[859] training's multi_logloss: 0.933664 training's macroF1: 0.590257 valid_1's multi_logloss: 1.07817 valid_1's macroF1: 0.38561
******************** Execution ended in 00h 01m 04.50s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.01314 training's macroF1: 0.539969 valid_1's multi_logloss: 1.07433 valid_1's macroF1: 0.399989
[1000] training's multi_logloss: 0.912796 training's macroF1: 0.59641 valid_1's multi_logloss: 1.06692 valid_1's macroF1: 0.40821
Early stopping, best iteration is:
[675] training's multi_logloss: 0.971183 training's macroF1: 0.565023 valid_1's multi_logloss: 1.07018 valid_1's macroF1: 0.413991
******************** Execution ended in 00h 00m 56.07s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.00459 training's macroF1: 0.544863 valid_1's multi_logloss: 1.04974 valid_1's macroF1: 0.401321
Early stopping, best iteration is:
[177] training's multi_logloss: 1.1427 training's macroF1: 0.493032 valid_1's multi_logloss: 1.10439 valid_1's macroF1: 0.420533
******************** Execution ended in 00h 00m 33.07s ********************
######################################## 5 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.992275 training's macroF1: 0.554661 valid_1's multi_logloss: 1.04224 valid_1's macroF1: 0.350008
Early stopping, best iteration is:
[42] training's multi_logloss: 1.28725 training's macroF1: 0.456181 valid_1's multi_logloss: 1.25924 valid_1's macroF1: 0.369595
******************** Execution ended in 00h 00m 36.03s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.998066 training's macroF1: 0.554626 valid_1's multi_logloss: 1.07297 valid_1's macroF1: 0.394617
Early stopping, best iteration is:
[177] training's multi_logloss: 1.12975 training's macroF1: 0.496828 valid_1's multi_logloss: 1.12768 valid_1's macroF1: 0.411119
******************** Execution ended in 00h 00m 40.77s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.991497 training's macroF1: 0.547388 valid_1's multi_logloss: 1.05748 valid_1's macroF1: 0.413745
[1000] training's multi_logloss: 0.888162 training's macroF1: 0.598206 valid_1's multi_logloss: 1.05411 valid_1's macroF1: 0.420612
Early stopping, best iteration is:
[947] training's multi_logloss: 0.896833 training's macroF1: 0.590849 valid_1's multi_logloss: 1.05515 valid_1's macroF1: 0.423638
******************** Execution ended in 00h 01m 25.98s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.00452 training's macroF1: 0.542783 valid_1's multi_logloss: 1.05158 valid_1's macroF1: 0.414599
Early stopping, best iteration is:
[372] training's multi_logloss: 1.04301 training's macroF1: 0.52251 valid_1's multi_logloss: 1.06469 valid_1's macroF1: 0.423675
******************** Execution ended in 00h 00m 52.88s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.00984 training's macroF1: 0.544674 valid_1's multi_logloss: 1.07357 valid_1's macroF1: 0.410082
Early stopping, best iteration is:
[316] training's multi_logloss: 1.06984 training's macroF1: 0.515931 valid_1's multi_logloss: 1.08716 valid_1's macroF1: 0.424683
******************** Execution ended in 00h 00m 49.64s ********************
######################################## 6 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.14213 training's macroF1: 0.475743 valid_1's multi_logloss: 1.07646 valid_1's macroF1: 0.428338
[1000] training's multi_logloss: 1.08016 training's macroF1: 0.502444 valid_1's multi_logloss: 1.0567 valid_1's macroF1: 0.421571
Early stopping, best iteration is:
[693] training's multi_logloss: 1.11319 training's macroF1: 0.490205 valid_1's multi_logloss: 1.06397 valid_1's macroF1: 0.436966
******************** Execution ended in 00h 00m 42.05s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.13654 training's macroF1: 0.476825 valid_1's multi_logloss: 1.10079 valid_1's macroF1: 0.38883
Early stopping, best iteration is:
[138] training's multi_logloss: 1.25913 training's macroF1: 0.449623 valid_1's multi_logloss: 1.21325 valid_1's macroF1: 0.400343
******************** Execution ended in 00h 00m 22.95s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.12307 training's macroF1: 0.486005 valid_1's multi_logloss: 1.10988 valid_1's macroF1: 0.397362
Early stopping, best iteration is:
[257] training's multi_logloss: 1.18855 training's macroF1: 0.473953 valid_1's multi_logloss: 1.15157 valid_1's macroF1: 0.416235
******************** Execution ended in 00h 00m 27.46s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.12768 training's macroF1: 0.488627 valid_1's multi_logloss: 1.12428 valid_1's macroF1: 0.393269
[1000] training's multi_logloss: 1.06535 training's macroF1: 0.499936 valid_1's multi_logloss: 1.11425 valid_1's macroF1: 0.389314
Early stopping, best iteration is:
[606] training's multi_logloss: 1.11056 training's macroF1: 0.496642 valid_1's multi_logloss: 1.11785 valid_1's macroF1: 0.394677
******************** Execution ended in 00h 00m 39.21s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.13176 training's macroF1: 0.472903 valid_1's multi_logloss: 1.07774 valid_1's macroF1: 0.372848
[1000] training's multi_logloss: 1.07215 training's macroF1: 0.502397 valid_1's multi_logloss: 1.05157 valid_1's macroF1: 0.392767
[1500] training's multi_logloss: 1.03591 training's macroF1: 0.521544 valid_1's multi_logloss: 1.04452 valid_1's macroF1: 0.399919
Early stopping, best iteration is:
[1367] training's multi_logloss: 1.04458 training's macroF1: 0.517019 valid_1's multi_logloss: 1.04357 valid_1's macroF1: 0.409965
******************** Execution ended in 00h 01m 04.88s ********************
######################################## 7 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.05164 training's macroF1: 0.550302 valid_1's multi_logloss: 1.06148 valid_1's macroF1: 0.383299
[1000] training's multi_logloss: 0.930645 training's macroF1: 0.601107 valid_1's multi_logloss: 1.00967 valid_1's macroF1: 0.406095
[1500] training's multi_logloss: 0.8504 training's macroF1: 0.643756 valid_1's multi_logloss: 0.993597 valid_1's macroF1: 0.400182
Early stopping, best iteration is:
[1138] training's multi_logloss: 0.905869 training's macroF1: 0.610902 valid_1's multi_logloss: 1.00348 valid_1's macroF1: 0.41111
******************** Execution ended in 00h 01m 56.51s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.0452 training's macroF1: 0.577034 valid_1's multi_logloss: 1.07369 valid_1's macroF1: 0.406377
Early stopping, best iteration is:
[7] training's multi_logloss: 1.37621 training's macroF1: 0.477851 valid_1's multi_logloss: 1.37415 valid_1's macroF1: 0.424352
******************** Execution ended in 00h 00m 36.99s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.04699 training's macroF1: 0.580629 valid_1's multi_logloss: 1.06364 valid_1's macroF1: 0.403479
[1000] training's multi_logloss: 0.9211 training's macroF1: 0.614351 valid_1's multi_logloss: 1.01444 valid_1's macroF1: 0.416426
[1500] training's multi_logloss: 0.838359 training's macroF1: 0.646375 valid_1's multi_logloss: 0.998619 valid_1's macroF1: 0.417353
Early stopping, best iteration is:
[1171] training's multi_logloss: 0.889932 training's macroF1: 0.628065 valid_1's multi_logloss: 1.00701 valid_1's macroF1: 0.424089
******************** Execution ended in 00h 02m 03.77s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.04777 training's macroF1: 0.566903 valid_1's multi_logloss: 1.10848 valid_1's macroF1: 0.421771
[1000] training's multi_logloss: 0.92302 training's macroF1: 0.607525 valid_1's multi_logloss: 1.07036 valid_1's macroF1: 0.403037
Early stopping, best iteration is:
[623] training's multi_logloss: 1.00933 training's macroF1: 0.57663 valid_1's multi_logloss: 1.09255 valid_1's macroF1: 0.424716
******************** Execution ended in 00h 01m 20.67s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.04324 training's macroF1: 0.569833 valid_1's multi_logloss: 1.08703 valid_1's macroF1: 0.394684
Early stopping, best iteration is:
[1] training's multi_logloss: 1.38485 training's macroF1: 0.44307 valid_1's multi_logloss: 1.38456 valid_1's macroF1: 0.424377
******************** Execution ended in 00h 00m 36.32s ********************
######################################## 8 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.16149 training's macroF1: 0.543075 valid_1's multi_logloss: 1.17371 valid_1's macroF1: 0.387537
[1000] training's multi_logloss: 1.04611 training's macroF1: 0.574457 valid_1's multi_logloss: 1.10192 valid_1's macroF1: 0.386237
Early stopping, best iteration is:
[573] training's multi_logloss: 1.14036 training's macroF1: 0.548412 valid_1's multi_logloss: 1.15857 valid_1's macroF1: 0.400353
******************** Execution ended in 00h 01m 35.76s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.16426 training's macroF1: 0.534868 valid_1's multi_logloss: 1.17475 valid_1's macroF1: 0.398161
[1000] training's multi_logloss: 1.04831 training's macroF1: 0.567904 valid_1's multi_logloss: 1.10382 valid_1's macroF1: 0.416745
[1500] training's multi_logloss: 0.974683 training's macroF1: 0.595193 valid_1's multi_logloss: 1.07643 valid_1's macroF1: 0.412397
[2000] training's multi_logloss: 0.920059 training's macroF1: 0.61623 valid_1's multi_logloss: 1.06484 valid_1's macroF1: 0.417813
Early stopping, best iteration is:
[1752] training's multi_logloss: 0.94565 training's macroF1: 0.606602 valid_1's multi_logloss: 1.06978 valid_1's macroF1: 0.425967
******************** Execution ended in 00h 03m 17.94s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.16205 training's macroF1: 0.528716 valid_1's multi_logloss: 1.14379 valid_1's macroF1: 0.402361
Early stopping, best iteration is:
[1] training's multi_logloss: 1.3856 training's macroF1: 0.449768 valid_1's multi_logloss: 1.38543 valid_1's macroF1: 0.416367
******************** Execution ended in 00h 00m 46.12s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15768 training's macroF1: 0.532378 valid_1's multi_logloss: 1.16844 valid_1's macroF1: 0.398
[1000] training's multi_logloss: 1.03994 training's macroF1: 0.570291 valid_1's multi_logloss: 1.09988 valid_1's macroF1: 0.399108
Early stopping, best iteration is:
[679] training's multi_logloss: 1.10844 training's macroF1: 0.546831 valid_1's multi_logloss: 1.13583 valid_1's macroF1: 0.414
******************** Execution ended in 00h 01m 47.68s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.16801 training's macroF1: 0.536208 valid_1's multi_logloss: 1.15321 valid_1's macroF1: 0.459141
[1000] training's multi_logloss: 1.05757 training's macroF1: 0.564505 valid_1's multi_logloss: 1.0696 valid_1's macroF1: 0.458623
[1500] training's multi_logloss: 0.984854 training's macroF1: 0.591915 valid_1's multi_logloss: 1.03504 valid_1's macroF1: 0.472119
[2000] training's multi_logloss: 0.93066 training's macroF1: 0.610184 valid_1's multi_logloss: 1.01801 valid_1's macroF1: 0.462045
Early stopping, best iteration is:
[1510] training's multi_logloss: 0.983687 training's macroF1: 0.5917 valid_1's multi_logloss: 1.03466 valid_1's macroF1: 0.477436
******************** Execution ended in 00h 02m 56.35s ********************
######################################## 9 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.960583 training's macroF1: 0.579914 valid_1's multi_logloss: 1.03474 valid_1's macroF1: 0.39926
Early stopping, best iteration is:
[408] training's multi_logloss: 0.994213 training's macroF1: 0.568429 valid_1's multi_logloss: 1.04096 valid_1's macroF1: 0.40628
******************** Execution ended in 00h 00m 48.61s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.970497 training's macroF1: 0.583039 valid_1's multi_logloss: 1.06043 valid_1's macroF1: 0.417215
Early stopping, best iteration is:
[167] training's multi_logloss: 1.14265 training's macroF1: 0.524687 valid_1's multi_logloss: 1.13427 valid_1's macroF1: 0.429021
******************** Execution ended in 00h 00m 34.70s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.968217 training's macroF1: 0.586098 valid_1's multi_logloss: 1.04779 valid_1's macroF1: 0.409036
Early stopping, best iteration is:
[217] training's multi_logloss: 1.10033 training's macroF1: 0.537756 valid_1's multi_logloss: 1.10205 valid_1's macroF1: 0.429907
******************** Execution ended in 00h 00m 38.41s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.970681 training's macroF1: 0.593783 valid_1's multi_logloss: 1.0127 valid_1's macroF1: 0.432998
Early stopping, best iteration is:
[300] training's multi_logloss: 1.05387 training's macroF1: 0.561329 valid_1's multi_logloss: 1.03988 valid_1's macroF1: 0.447737
******************** Execution ended in 00h 00m 42.99s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.979745 training's macroF1: 0.572951 valid_1's multi_logloss: 1.019 valid_1's macroF1: 0.423471
[1000] training's multi_logloss: 0.854143 training's macroF1: 0.645413 valid_1's multi_logloss: 0.988961 valid_1's macroF1: 0.449642
[1500] training's multi_logloss: 0.770033 training's macroF1: 0.694107 valid_1's multi_logloss: 0.981734 valid_1's macroF1: 0.4499
[2000] training's multi_logloss: 0.706969 training's macroF1: 0.728366 valid_1's multi_logloss: 0.978016 valid_1's macroF1: 0.459857
Early stopping, best iteration is:
[1715] training's multi_logloss: 0.74066 training's macroF1: 0.709213 valid_1's multi_logloss: 0.978958 valid_1's macroF1: 0.469395
******************** Execution ended in 00h 01m 52.54s ********************
######################################## 10 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.05739 training's macroF1: 0.530952 valid_1's multi_logloss: 1.07139 valid_1's macroF1: 0.391319
Early stopping, best iteration is:
[333] training's multi_logloss: 1.11348 training's macroF1: 0.51335 valid_1's multi_logloss: 1.09858 valid_1's macroF1: 0.412825
******************** Execution ended in 00h 00m 52.83s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.06447 training's macroF1: 0.52675 valid_1's multi_logloss: 1.05993 valid_1's macroF1: 0.414848
[1000] training's multi_logloss: 0.96818 training's macroF1: 0.561465 valid_1's multi_logloss: 1.02944 valid_1's macroF1: 0.418391
Early stopping, best iteration is:
[713] training's multi_logloss: 1.01657 training's macroF1: 0.544651 valid_1's multi_logloss: 1.0394 valid_1's macroF1: 0.426892
******************** Execution ended in 00h 01m 18.15s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.04849 training's macroF1: 0.543154 valid_1's multi_logloss: 1.08344 valid_1's macroF1: 0.376857
Early stopping, best iteration is:
[335] training's multi_logloss: 1.10386 training's macroF1: 0.523546 valid_1's multi_logloss: 1.10624 valid_1's macroF1: 0.392807
******************** Execution ended in 00h 00m 52.42s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.04849 training's macroF1: 0.536687 valid_1's multi_logloss: 1.08009 valid_1's macroF1: 0.41184
[1000] training's multi_logloss: 0.947811 training's macroF1: 0.58389 valid_1's multi_logloss: 1.06999 valid_1's macroF1: 0.411491
Early stopping, best iteration is:
[692] training's multi_logloss: 1.0028 training's macroF1: 0.55583 valid_1's multi_logloss: 1.07183 valid_1's macroF1: 0.425535
******************** Execution ended in 00h 01m 13.43s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.04776 training's macroF1: 0.525231 valid_1's multi_logloss: 1.05478 valid_1's macroF1: 0.394999
[1000] training's multi_logloss: 0.953525 training's macroF1: 0.570049 valid_1's multi_logloss: 1.03083 valid_1's macroF1: 0.387886
Early stopping, best iteration is:
[600] training's multi_logloss: 1.02328 training's macroF1: 0.536102 valid_1's multi_logloss: 1.0459 valid_1's macroF1: 0.406943
******************** Execution ended in 00h 01m 09.34s ********************
######################################## 11 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.979067 training's macroF1: 0.580296 valid_1's multi_logloss: 1.03406 valid_1's macroF1: 0.41965
[1000] training's multi_logloss: 0.85012 training's macroF1: 0.644015 valid_1's multi_logloss: 1.00996 valid_1's macroF1: 0.42395
Early stopping, best iteration is:
[550] training's multi_logloss: 0.962447 training's macroF1: 0.589368 valid_1's multi_logloss: 1.02926 valid_1's macroF1: 0.432322
******************** Execution ended in 00h 00m 55.44s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.975745 training's macroF1: 0.581412 valid_1's multi_logloss: 1.01378 valid_1's macroF1: 0.414806
Early stopping, best iteration is:
[414] training's multi_logloss: 1.00842 training's macroF1: 0.567243 valid_1's multi_logloss: 1.02664 valid_1's macroF1: 0.422131
******************** Execution ended in 00h 00m 48.16s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.976941 training's macroF1: 0.591674 valid_1's multi_logloss: 1.05171 valid_1's macroF1: 0.420491
Early stopping, best iteration is:
[154] training's multi_logloss: 1.16688 training's macroF1: 0.531293 valid_1's multi_logloss: 1.15214 valid_1's macroF1: 0.441034
******************** Execution ended in 00h 00m 33.84s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.971184 training's macroF1: 0.583924 valid_1's multi_logloss: 1.03994 valid_1's macroF1: 0.402042
Early stopping, best iteration is:
[361] training's multi_logloss: 1.02742 training's macroF1: 0.568314 valid_1's multi_logloss: 1.05436 valid_1's macroF1: 0.411159
******************** Execution ended in 00h 00m 45.24s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.980007 training's macroF1: 0.592463 valid_1's multi_logloss: 1.08213 valid_1's macroF1: 0.418053
[1000] training's multi_logloss: 0.851693 training's macroF1: 0.660205 valid_1's multi_logloss: 1.06363 valid_1's macroF1: 0.428683
Early stopping, best iteration is:
[848] training's multi_logloss: 0.884041 training's macroF1: 0.637129 valid_1's multi_logloss: 1.06712 valid_1's macroF1: 0.429805
******************** Execution ended in 00h 01m 10.30s ********************
######################################## 12 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.29246 training's macroF1: 0.475911 valid_1's multi_logloss: 1.27084 valid_1's macroF1: 0.348424
Early stopping, best iteration is:
[80] training's multi_logloss: 1.36876 training's macroF1: 0.450804 valid_1's multi_logloss: 1.36327 valid_1's macroF1: 0.354663
******************** Execution ended in 00h 00m 39.81s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.29905 training's macroF1: 0.465427 valid_1's multi_logloss: 1.28232 valid_1's macroF1: 0.37384
Early stopping, best iteration is:
[158] training's multi_logloss: 1.35494 training's macroF1: 0.456206 valid_1's multi_logloss: 1.34719 valid_1's macroF1: 0.384269
******************** Execution ended in 00h 00m 45.47s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.30009 training's macroF1: 0.460667 valid_1's multi_logloss: 1.27615 valid_1's macroF1: 0.395346
Early stopping, best iteration is:
[7] training's multi_logloss: 1.3848 training's macroF1: 0.432936 valid_1's multi_logloss: 1.38439 valid_1's macroF1: 0.417813
******************** Execution ended in 00h 00m 36.17s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.29841 training's macroF1: 0.483632 valid_1's multi_logloss: 1.27353 valid_1's macroF1: 0.410683
Early stopping, best iteration is:
[7] training's multi_logloss: 1.38483 training's macroF1: 0.451192 valid_1's multi_logloss: 1.38425 valid_1's macroF1: 0.42068
******************** Execution ended in 00h 00m 39.64s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.29901 training's macroF1: 0.481841 valid_1's multi_logloss: 1.28631 valid_1's macroF1: 0.408987
[1000] training's multi_logloss: 1.23559 training's macroF1: 0.497927 valid_1's multi_logloss: 1.22382 valid_1's macroF1: 0.418935
[1500] training's multi_logloss: 1.18594 training's macroF1: 0.5059 valid_1's multi_logloss: 1.18069 valid_1's macroF1: 0.416248
Early stopping, best iteration is:
[1151] training's multi_logloss: 1.2195 training's macroF1: 0.499133 valid_1's multi_logloss: 1.20936 valid_1's macroF1: 0.424885
******************** Execution ended in 00h 01m 58.60s ********************
######################################## 13 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.796053 training's macroF1: 0.6769 valid_1's multi_logloss: 1.06737 valid_1's macroF1: 0.366604
[1000] training's multi_logloss: 0.639473 training's macroF1: 0.749079 valid_1's multi_logloss: 1.06012 valid_1's macroF1: 0.374075
Early stopping, best iteration is:
[621] training's multi_logloss: 0.748547 training's macroF1: 0.697439 valid_1's multi_logloss: 1.06386 valid_1's macroF1: 0.384188
******************** Execution ended in 00h 01m 02.29s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.810608 training's macroF1: 0.688407 valid_1's multi_logloss: 1.00303 valid_1's macroF1: 0.410421
Early stopping, best iteration is:
[36] training's multi_logloss: 1.26059 training's macroF1: 0.529108 valid_1's multi_logloss: 1.23036 valid_1's macroF1: 0.444541
******************** Execution ended in 00h 00m 29.62s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.81863 training's macroF1: 0.668042 valid_1's multi_logloss: 0.984283 valid_1's macroF1: 0.426227
[1000] training's multi_logloss: 0.659517 training's macroF1: 0.741733 valid_1's multi_logloss: 0.961818 valid_1's macroF1: 0.431293
[1500] training's multi_logloss: 0.56535 training's macroF1: 0.785813 valid_1's multi_logloss: 0.951673 valid_1's macroF1: 0.42772
[2000] training's multi_logloss: 0.501424 training's macroF1: 0.805714 valid_1's multi_logloss: 0.948709 valid_1's macroF1: 0.421404
Early stopping, best iteration is:
[1522] training's multi_logloss: 0.56183 training's macroF1: 0.787145 valid_1's multi_logloss: 0.952727 valid_1's macroF1: 0.438539
******************** Execution ended in 00h 01m 46.77s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.814833 training's macroF1: 0.68109 valid_1's multi_logloss: 0.989224 valid_1's macroF1: 0.423091
Early stopping, best iteration is:
[67] training's multi_logloss: 1.18275 training's macroF1: 0.543441 valid_1's multi_logloss: 1.16114 valid_1's macroF1: 0.448106
******************** Execution ended in 00h 00m 32.09s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.795034 training's macroF1: 0.683981 valid_1's multi_logloss: 1.04802 valid_1's macroF1: 0.415655
Early stopping, best iteration is:
[285] training's multi_logloss: 0.913446 training's macroF1: 0.634247 valid_1's multi_logloss: 1.05315 valid_1's macroF1: 0.427517
******************** Execution ended in 00h 00m 45.17s ********************
######################################## 14 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.11515 training's macroF1: 0.502067 valid_1's multi_logloss: 1.13088 valid_1's macroF1: 0.382403
[1000] training's multi_logloss: 1.02928 training's macroF1: 0.533553 valid_1's multi_logloss: 1.10583 valid_1's macroF1: 0.386869
[1500] training's multi_logloss: 0.976349 training's macroF1: 0.554248 valid_1's multi_logloss: 1.1039 valid_1's macroF1: 0.390098
Early stopping, best iteration is:
[1439] training's multi_logloss: 0.981934 training's macroF1: 0.553774 valid_1's multi_logloss: 1.10414 valid_1's macroF1: 0.392735
******************** Execution ended in 00h 01m 37.46s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.12244 training's macroF1: 0.502836 valid_1's multi_logloss: 1.09195 valid_1's macroF1: 0.401618
Early stopping, best iteration is:
[173] training's multi_logloss: 1.24366 training's macroF1: 0.47021 valid_1's multi_logloss: 1.19787 valid_1's macroF1: 0.416747
******************** Execution ended in 00h 00m 33.20s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.1167 training's macroF1: 0.494135 valid_1's multi_logloss: 1.05546 valid_1's macroF1: 0.404235
[1000] training's multi_logloss: 1.03462 training's macroF1: 0.538682 valid_1's multi_logloss: 1.01899 valid_1's macroF1: 0.410165
Early stopping, best iteration is:
[623] training's multi_logloss: 1.09099 training's macroF1: 0.507756 valid_1's multi_logloss: 1.04091 valid_1's macroF1: 0.415521
******************** Execution ended in 00h 00m 58.58s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.1164 training's macroF1: 0.511266 valid_1's multi_logloss: 1.13824 valid_1's macroF1: 0.382079
Early stopping, best iteration is:
[53] training's multi_logloss: 1.32614 training's macroF1: 0.456518 valid_1's multi_logloss: 1.3151 valid_1's macroF1: 0.385225
******************** Execution ended in 00h 00m 27.48s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.11219 training's macroF1: 0.491069 valid_1's multi_logloss: 1.07265 valid_1's macroF1: 0.385793
[1000] training's multi_logloss: 1.02932 training's macroF1: 0.527314 valid_1's multi_logloss: 1.03614 valid_1's macroF1: 0.390766
[1500] training's multi_logloss: 0.978791 training's macroF1: 0.549414 valid_1's multi_logloss: 1.02757 valid_1's macroF1: 0.399675
[2000] training's multi_logloss: 0.939191 training's macroF1: 0.576691 valid_1's multi_logloss: 1.02427 valid_1's macroF1: 0.390172
Early stopping, best iteration is:
[1765] training's multi_logloss: 0.956839 training's macroF1: 0.563314 valid_1's multi_logloss: 1.02566 valid_1's macroF1: 0.404754
******************** Execution ended in 00h 01m 48.52s ********************
######################################## 15 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.719719 training's macroF1: 0.718996 valid_1's multi_logloss: 1.04115 valid_1's macroF1: 0.38858
Early stopping, best iteration is:
[103] training's multi_logloss: 1.07686 training's macroF1: 0.580143 valid_1's multi_logloss: 1.13666 valid_1's macroF1: 0.403155
******************** Execution ended in 00h 00m 51.63s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.723133 training's macroF1: 0.729887 valid_1's multi_logloss: 0.987832 valid_1's macroF1: 0.415537
Early stopping, best iteration is:
[326] training's multi_logloss: 0.831968 training's macroF1: 0.684592 valid_1's multi_logloss: 0.998096 valid_1's macroF1: 0.432349
******************** Execution ended in 00h 01m 10.56s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.718914 training's macroF1: 0.723201 valid_1's multi_logloss: 0.970824 valid_1's macroF1: 0.399813
Early stopping, best iteration is:
[157] training's multi_logloss: 0.985924 training's macroF1: 0.629349 valid_1's multi_logloss: 1.03996 valid_1's macroF1: 0.41978
******************** Execution ended in 00h 00m 55.50s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.710193 training's macroF1: 0.733349 valid_1's multi_logloss: 1.05624 valid_1's macroF1: 0.416228
[1000] training's multi_logloss: 0.52302 training's macroF1: 0.814478 valid_1's multi_logloss: 1.04124 valid_1's macroF1: 0.397145
Early stopping, best iteration is:
[504] training's multi_logloss: 0.708226 training's macroF1: 0.735863 valid_1's multi_logloss: 1.05606 valid_1's macroF1: 0.417131
******************** Execution ended in 00h 01m 24.97s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.720316 training's macroF1: 0.720217 valid_1's multi_logloss: 0.998861 valid_1's macroF1: 0.437373
[1000] training's multi_logloss: 0.53047 training's macroF1: 0.800926 valid_1's multi_logloss: 0.97258 valid_1's macroF1: 0.45171
[1500] training's multi_logloss: 0.423284 training's macroF1: 0.852537 valid_1's multi_logloss: 0.96553 valid_1's macroF1: 0.454637
Early stopping, best iteration is:
[1278] training's multi_logloss: 0.464182 training's macroF1: 0.835227 valid_1's multi_logloss: 0.968516 valid_1's macroF1: 0.461752
******************** Execution ended in 00h 02m 27.49s ********************
######################################## 16 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.991341 training's macroF1: 0.548778 valid_1's multi_logloss: 1.07581 valid_1's macroF1: 0.366344
Early stopping, best iteration is:
[361] training's multi_logloss: 1.03856 training's macroF1: 0.533282 valid_1's multi_logloss: 1.08662 valid_1's macroF1: 0.379251
******************** Execution ended in 00h 00m 41.76s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.01557 training's macroF1: 0.546137 valid_1's multi_logloss: 1.04278 valid_1's macroF1: 0.393345
Early stopping, best iteration is:
[213] training's multi_logloss: 1.13356 training's macroF1: 0.510934 valid_1's multi_logloss: 1.1085 valid_1's macroF1: 0.420432
******************** Execution ended in 00h 00m 37.91s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.00182 training's macroF1: 0.570637 valid_1's multi_logloss: 1.05628 valid_1's macroF1: 0.410853
Early stopping, best iteration is:
[395] training's multi_logloss: 1.03678 training's macroF1: 0.552593 valid_1's multi_logloss: 1.06752 valid_1's macroF1: 0.415816
******************** Execution ended in 00h 00m 44.91s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.01552 training's macroF1: 0.556334 valid_1's multi_logloss: 1.05619 valid_1's macroF1: 0.405569
Early stopping, best iteration is:
[153] training's multi_logloss: 1.18173 training's macroF1: 0.506857 valid_1's multi_logloss: 1.15633 valid_1's macroF1: 0.410264
******************** Execution ended in 00h 00m 32.93s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.01017 training's macroF1: 0.547826 valid_1's multi_logloss: 1.05511 valid_1's macroF1: 0.406426
[1000] training's multi_logloss: 0.899636 training's macroF1: 0.599848 valid_1's multi_logloss: 1.04141 valid_1's macroF1: 0.431214
[1500] training's multi_logloss: 0.826699 training's macroF1: 0.639014 valid_1's multi_logloss: 1.03901 valid_1's macroF1: 0.431922
[2000] training's multi_logloss: 0.77062 training's macroF1: 0.672196 valid_1's multi_logloss: 1.03768 valid_1's macroF1: 0.434137
[2500] training's multi_logloss: 0.725227 training's macroF1: 0.699865 valid_1's multi_logloss: 1.038 valid_1's macroF1: 0.441531
Early stopping, best iteration is:
[2107] training's multi_logloss: 0.760165 training's macroF1: 0.680114 valid_1's multi_logloss: 1.03825 valid_1's macroF1: 0.44639
******************** Execution ended in 00h 02m 06.20s ********************
######################################## 17 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.16144 training's macroF1: 0.485398 valid_1's multi_logloss: 1.10625 valid_1's macroF1: 0.412715
Early stopping, best iteration is:
[430] training's multi_logloss: 1.17939 training's macroF1: 0.477711 valid_1's multi_logloss: 1.12221 valid_1's macroF1: 0.428811
******************** Execution ended in 00h 00m 45.84s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15249 training's macroF1: 0.485752 valid_1's multi_logloss: 1.12388 valid_1's macroF1: 0.417605
[1000] training's multi_logloss: 1.06534 training's macroF1: 0.513267 valid_1's multi_logloss: 1.07632 valid_1's macroF1: 0.414977
Early stopping, best iteration is:
[630] training's multi_logloss: 1.1237 training's macroF1: 0.495489 valid_1's multi_logloss: 1.10269 valid_1's macroF1: 0.42889
******************** Execution ended in 00h 00m 54.39s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15398 training's macroF1: 0.483677 valid_1's multi_logloss: 1.09515 valid_1's macroF1: 0.405397
[1000] training's multi_logloss: 1.06827 training's macroF1: 0.519403 valid_1's multi_logloss: 1.03886 valid_1's macroF1: 0.425609
[1500] training's multi_logloss: 1.01702 training's macroF1: 0.544883 valid_1's multi_logloss: 1.02252 valid_1's macroF1: 0.413679
Early stopping, best iteration is:
[1199] training's multi_logloss: 1.04571 training's macroF1: 0.53007 valid_1's multi_logloss: 1.03036 valid_1's macroF1: 0.428099
******************** Execution ended in 00h 01m 21.86s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.14959 training's macroF1: 0.487923 valid_1's multi_logloss: 1.12724 valid_1's macroF1: 0.371036
Early stopping, best iteration is:
[472] training's multi_logloss: 1.15639 training's macroF1: 0.486354 valid_1's multi_logloss: 1.13165 valid_1's macroF1: 0.375175
******************** Execution ended in 00h 00m 48.98s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15383 training's macroF1: 0.496044 valid_1's multi_logloss: 1.15408 valid_1's macroF1: 0.381772
[1000] training's multi_logloss: 1.07192 training's macroF1: 0.519119 valid_1's multi_logloss: 1.11394 valid_1's macroF1: 0.402939
[1500] training's multi_logloss: 1.02229 training's macroF1: 0.539649 valid_1's multi_logloss: 1.10419 valid_1's macroF1: 0.406172
Early stopping, best iteration is:
[1422] training's multi_logloss: 1.02903 training's macroF1: 0.533826 valid_1's multi_logloss: 1.10546 valid_1's macroF1: 0.410275
******************** Execution ended in 00h 01m 33.57s ********************
######################################## 18 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.10311 training's macroF1: 0.560559 valid_1's multi_logloss: 1.0851 valid_1's macroF1: 0.415552
Early stopping, best iteration is:
[168] training's multi_logloss: 1.25194 training's macroF1: 0.527789 valid_1's multi_logloss: 1.22333 valid_1's macroF1: 0.434388
******************** Execution ended in 00h 00m 46.63s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.09991 training's macroF1: 0.560437 valid_1's multi_logloss: 1.09192 valid_1's macroF1: 0.42913
Early stopping, best iteration is:
[21] training's multi_logloss: 1.36434 training's macroF1: 0.494973 valid_1's multi_logloss: 1.3571 valid_1's macroF1: 0.450147
******************** Execution ended in 00h 00m 40.47s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.10035 training's macroF1: 0.555457 valid_1's multi_logloss: 1.08409 valid_1's macroF1: 0.447443
Early stopping, best iteration is:
[325] training's multi_logloss: 1.16711 training's macroF1: 0.535076 valid_1's multi_logloss: 1.14041 valid_1's macroF1: 0.449101
******************** Execution ended in 00h 00m 56.82s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.07951 training's macroF1: 0.556213 valid_1's multi_logloss: 1.12567 valid_1's macroF1: 0.389544
Early stopping, best iteration is:
[16] training's multi_logloss: 1.36801 training's macroF1: 0.486192 valid_1's multi_logloss: 1.36709 valid_1's macroF1: 0.397304
******************** Execution ended in 00h 00m 35.74s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.08637 training's macroF1: 0.563228 valid_1's multi_logloss: 1.13928 valid_1's macroF1: 0.377075
Early stopping, best iteration is:
[18] training's multi_logloss: 1.36655 training's macroF1: 0.489961 valid_1's multi_logloss: 1.36515 valid_1's macroF1: 0.392596
******************** Execution ended in 00h 00m 35.37s ********************
######################################## 19 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.16096 training's macroF1: 0.528093 valid_1's multi_logloss: 1.12317 valid_1's macroF1: 0.422496
Early stopping, best iteration is:
[455] training's multi_logloss: 1.17441 training's macroF1: 0.525833 valid_1's multi_logloss: 1.13537 valid_1's macroF1: 0.430164
******************** Execution ended in 00h 01m 16.98s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.1651 training's macroF1: 0.520922 valid_1's multi_logloss: 1.14566 valid_1's macroF1: 0.395906
Early stopping, best iteration is:
[332] training's multi_logloss: 1.21984 training's macroF1: 0.498022 valid_1's multi_logloss: 1.19586 valid_1's macroF1: 0.409149
******************** Execution ended in 00h 01m 06.14s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15709 training's macroF1: 0.520787 valid_1's multi_logloss: 1.14558 valid_1's macroF1: 0.391539
Early stopping, best iteration is:
[2] training's multi_logloss: 1.38474 training's macroF1: 0.445344 valid_1's multi_logloss: 1.38438 valid_1's macroF1: 0.417154
******************** Execution ended in 00h 00m 39.71s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.1501 training's macroF1: 0.507617 valid_1's multi_logloss: 1.17402 valid_1's macroF1: 0.363787
Early stopping, best iteration is:
[3] training's multi_logloss: 1.38396 training's macroF1: 0.442248 valid_1's multi_logloss: 1.38362 valid_1's macroF1: 0.376224
******************** Execution ended in 00h 00m 41.43s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15215 training's macroF1: 0.525629 valid_1's multi_logloss: 1.17457 valid_1's macroF1: 0.361562
[1000] training's multi_logloss: 1.03978 training's macroF1: 0.558132 valid_1's multi_logloss: 1.11233 valid_1's macroF1: 0.379311
Early stopping, best iteration is:
[905] training's multi_logloss: 1.05657 training's macroF1: 0.54828 valid_1's multi_logloss: 1.11952 valid_1's macroF1: 0.383473
******************** Execution ended in 00h 01m 53.64s ********************
######################################## 20 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.04039 training's macroF1: 0.527414 valid_1's multi_logloss: 1.0534 valid_1's macroF1: 0.42005
[1000] training's multi_logloss: 0.945869 training's macroF1: 0.580036 valid_1's multi_logloss: 1.03864 valid_1's macroF1: 0.406612
Early stopping, best iteration is:
[628] training's multi_logloss: 1.01068 training's macroF1: 0.545473 valid_1's multi_logloss: 1.04708 valid_1's macroF1: 0.42716
******************** Execution ended in 00h 00m 54.36s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.03631 training's macroF1: 0.51629 valid_1's multi_logloss: 1.06409 valid_1's macroF1: 0.407714
Early stopping, best iteration is:
[259] training's multi_logloss: 1.11757 training's macroF1: 0.492983 valid_1's multi_logloss: 1.09678 valid_1's macroF1: 0.429086
******************** Execution ended in 00h 00m 37.09s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.0356 training's macroF1: 0.54323 valid_1's multi_logloss: 1.0763 valid_1's macroF1: 0.399144
Early stopping, best iteration is:
[309] training's multi_logloss: 1.09434 training's macroF1: 0.513538 valid_1's multi_logloss: 1.09515 valid_1's macroF1: 0.41528
******************** Execution ended in 00h 00m 43.94s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.02369 training's macroF1: 0.548811 valid_1's multi_logloss: 1.11332 valid_1's macroF1: 0.387036
[1000] training's multi_logloss: 0.92829 training's macroF1: 0.587162 valid_1's multi_logloss: 1.11578 valid_1's macroF1: 0.395637
Early stopping, best iteration is:
[544] training's multi_logloss: 1.01359 training's macroF1: 0.554965 valid_1's multi_logloss: 1.11091 valid_1's macroF1: 0.381769
******************** Execution ended in 00h 00m 50.26s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.03283 training's macroF1: 0.522996 valid_1's multi_logloss: 1.03461 valid_1's macroF1: 0.420961
[1000] training's multi_logloss: 0.940988 training's macroF1: 0.560025 valid_1's multi_logloss: 1.01778 valid_1's macroF1: 0.41203
Early stopping, best iteration is:
[605] training's multi_logloss: 1.00911 training's macroF1: 0.534159 valid_1's multi_logloss: 1.02658 valid_1's macroF1: 0.436975
******************** Execution ended in 00h 00m 52.41s ********************
######################################## 21 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.8295 training's macroF1: 0.716314 valid_1's multi_logloss: 1.01155 valid_1's macroF1: 0.422958
Early stopping, best iteration is:
[13] training's multi_logloss: 1.34781 training's macroF1: 0.525285 valid_1's multi_logloss: 1.3452 valid_1's macroF1: 0.435518
******************** Execution ended in 00h 00m 29.77s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.821289 training's macroF1: 0.696045 valid_1's multi_logloss: 0.983 valid_1's macroF1: 0.39235
Early stopping, best iteration is:
[12] training's multi_logloss: 1.35023 training's macroF1: 0.506874 valid_1's multi_logloss: 1.34009 valid_1's macroF1: 0.429818
******************** Execution ended in 00h 00m 30.32s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.80744 training's macroF1: 0.697536 valid_1's multi_logloss: 1.00423 valid_1's macroF1: 0.383837
Early stopping, best iteration is:
[9] training's multi_logloss: 1.35814 training's macroF1: 0.523172 valid_1's multi_logloss: 1.35401 valid_1's macroF1: 0.402235
******************** Execution ended in 00h 00m 29.81s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.823709 training's macroF1: 0.68925 valid_1's multi_logloss: 1.02915 valid_1's macroF1: 0.400118
Early stopping, best iteration is:
[386] training's multi_logloss: 0.885753 training's macroF1: 0.66877 valid_1's multi_logloss: 1.04237 valid_1's macroF1: 0.417867
******************** Execution ended in 00h 00m 55.31s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.828871 training's macroF1: 0.697904 valid_1's multi_logloss: 1.01026 valid_1's macroF1: 0.448786
[1000] training's multi_logloss: 0.647254 training's macroF1: 0.765756 valid_1's multi_logloss: 0.985841 valid_1's macroF1: 0.442639
Early stopping, best iteration is:
[503] training's multi_logloss: 0.827499 training's macroF1: 0.699775 valid_1's multi_logloss: 1.0104 valid_1's macroF1: 0.452063
******************** Execution ended in 00h 01m 02.04s ********************
######################################## 22 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.853062 training's macroF1: 0.67499 valid_1's multi_logloss: 1.03501 valid_1's macroF1: 0.422689
[1000] training's multi_logloss: 0.673052 training's macroF1: 0.752262 valid_1's multi_logloss: 1.00532 valid_1's macroF1: 0.418821
Early stopping, best iteration is:
[501] training's multi_logloss: 0.852552 training's macroF1: 0.674994 valid_1's multi_logloss: 1.03481 valid_1's macroF1: 0.42654
******************** Execution ended in 00h 01m 27.92s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.854253 training's macroF1: 0.697153 valid_1's multi_logloss: 1.02194 valid_1's macroF1: 0.414942
[1000] training's multi_logloss: 0.672828 training's macroF1: 0.766584 valid_1's multi_logloss: 0.999475 valid_1's macroF1: 0.432806
Early stopping, best iteration is:
[886] training's multi_logloss: 0.704918 training's macroF1: 0.758429 valid_1's multi_logloss: 1.0024 valid_1's macroF1: 0.436967
******************** Execution ended in 00h 02m 01.38s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.853408 training's macroF1: 0.686406 valid_1's multi_logloss: 0.997469 valid_1's macroF1: 0.40929
[1000] training's multi_logloss: 0.674496 training's macroF1: 0.759269 valid_1's multi_logloss: 0.965511 valid_1's macroF1: 0.424854
Early stopping, best iteration is:
[860] training's multi_logloss: 0.713759 training's macroF1: 0.745124 valid_1's multi_logloss: 0.970084 valid_1's macroF1: 0.434274
******************** Execution ended in 00h 02m 02.50s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.855503 training's macroF1: 0.682209 valid_1's multi_logloss: 1.00245 valid_1's macroF1: 0.408855
Early stopping, best iteration is:
[221] training's multi_logloss: 1.04339 training's macroF1: 0.618299 valid_1's multi_logloss: 1.07951 valid_1's macroF1: 0.420676
******************** Execution ended in 00h 01m 01.94s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.851665 training's macroF1: 0.679089 valid_1's multi_logloss: 1.0501 valid_1's macroF1: 0.390139
Early stopping, best iteration is:
[8] training's multi_logloss: 1.36429 training's macroF1: 0.499563 valid_1's multi_logloss: 1.36444 valid_1's macroF1: 0.39964
******************** Execution ended in 00h 00m 43.59s ********************
######################################## 23 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.07651 training's macroF1: 0.576284 valid_1's multi_logloss: 1.09385 valid_1's macroF1: 0.406565
Early stopping, best iteration is:
[452] training's multi_logloss: 1.09473 training's macroF1: 0.57358 valid_1's multi_logloss: 1.10612 valid_1's macroF1: 0.414289
******************** Execution ended in 00h 01m 11.73s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.07664 training's macroF1: 0.562097 valid_1's multi_logloss: 1.10442 valid_1's macroF1: 0.380507
Early stopping, best iteration is:
[5] training's multi_logloss: 1.38067 training's macroF1: 0.468954 valid_1's multi_logloss: 1.3799 valid_1's macroF1: 0.392206
******************** Execution ended in 00h 00m 37.61s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.07919 training's macroF1: 0.56974 valid_1's multi_logloss: 1.10431 valid_1's macroF1: 0.403978
Early stopping, best iteration is:
[314] training's multi_logloss: 1.15861 training's macroF1: 0.556978 valid_1's multi_logloss: 1.15816 valid_1's macroF1: 0.417018
******************** Execution ended in 00h 01m 03.50s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.08752 training's macroF1: 0.570788 valid_1's multi_logloss: 1.09413 valid_1's macroF1: 0.419409
[1000] training's multi_logloss: 0.955432 training's macroF1: 0.613257 valid_1's multi_logloss: 1.02056 valid_1's macroF1: 0.435846
[1500] training's multi_logloss: 0.869719 training's macroF1: 0.654768 valid_1's multi_logloss: 0.995543 valid_1's macroF1: 0.438266
[2000] training's multi_logloss: 0.803634 training's macroF1: 0.691102 valid_1's multi_logloss: 0.983469 valid_1's macroF1: 0.453552
[2500] training's multi_logloss: 0.750031 training's macroF1: 0.716025 valid_1's multi_logloss: 0.975284 valid_1's macroF1: 0.4533
Early stopping, best iteration is:
[2137] training's multi_logloss: 0.787831 training's macroF1: 0.698501 valid_1's multi_logloss: 0.980244 valid_1's macroF1: 0.458939
******************** Execution ended in 00h 03m 11.28s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.07572 training's macroF1: 0.56772 valid_1's multi_logloss: 1.10394 valid_1's macroF1: 0.401905
Early stopping, best iteration is:
[496] training's multi_logloss: 1.07714 training's macroF1: 0.567087 valid_1's multi_logloss: 1.10453 valid_1's macroF1: 0.409293
******************** Execution ended in 00h 01m 14.87s ********************
######################################## 24 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.36872 training's macroF1: 0.476035 valid_1's multi_logloss: 1.36434 valid_1's macroF1: 0.371415
Early stopping, best iteration is:
[4] training's multi_logloss: 1.38615 training's macroF1: 0.45417 valid_1's multi_logloss: 1.38612 valid_1's macroF1: 0.377204
******************** Execution ended in 00h 00m 35.62s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.36778 training's macroF1: 0.475893 valid_1's multi_logloss: 1.36227 valid_1's macroF1: 0.380175
Early stopping, best iteration is:
[44] training's multi_logloss: 1.38461 training's macroF1: 0.476616 valid_1's multi_logloss: 1.3841 valid_1's macroF1: 0.410753
******************** Execution ended in 00h 00m 35.43s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.36791 training's macroF1: 0.493256 valid_1's multi_logloss: 1.36258 valid_1's macroF1: 0.371391
Early stopping, best iteration is:
[2] training's multi_logloss: 1.38622 training's macroF1: 0.446181 valid_1's multi_logloss: 1.38619 valid_1's macroF1: 0.394467
******************** Execution ended in 00h 00m 32.39s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.36786 training's macroF1: 0.470686 valid_1's multi_logloss: 1.36366 valid_1's macroF1: 0.394163
Early stopping, best iteration is:
[14] training's multi_logloss: 1.38577 training's macroF1: 0.482172 valid_1's multi_logloss: 1.38564 valid_1's macroF1: 0.377765
******************** Execution ended in 00h 00m 31.95s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.3677 training's macroF1: 0.462665 valid_1's multi_logloss: 1.36474 valid_1's macroF1: 0.396157
Early stopping, best iteration is:
[58] training's multi_logloss: 1.38406 training's macroF1: 0.472283 valid_1's multi_logloss: 1.38367 valid_1's macroF1: 0.397819
******************** Execution ended in 00h 00m 34.09s ********************
######################################## 25 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.11832 training's macroF1: 0.488679 valid_1's multi_logloss: 1.06334 valid_1's macroF1: 0.406784
[1000] training's multi_logloss: 1.05304 training's macroF1: 0.524668 valid_1's multi_logloss: 1.043 valid_1's macroF1: 0.415379
[1500] training's multi_logloss: 1.01016 training's macroF1: 0.536585 valid_1's multi_logloss: 1.0435 valid_1's macroF1: 0.406986
Early stopping, best iteration is:
[1120] training's multi_logloss: 1.04213 training's macroF1: 0.526782 valid_1's multi_logloss: 1.03906 valid_1's macroF1: 0.413274
******************** Execution ended in 00h 01m 01.05s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.11296 training's macroF1: 0.492108 valid_1's multi_logloss: 1.05919 valid_1's macroF1: 0.393494
Early stopping, best iteration is:
[30] training's multi_logloss: 1.34225 training's macroF1: 0.442494 valid_1's multi_logloss: 1.32048 valid_1's macroF1: 0.418895
******************** Execution ended in 00h 00m 20.37s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.10973 training's macroF1: 0.492034 valid_1's multi_logloss: 1.10373 valid_1's macroF1: 0.391655
Early stopping, best iteration is:
[156] training's multi_logloss: 1.22373 training's macroF1: 0.448961 valid_1's multi_logloss: 1.1895 valid_1's macroF1: 0.404394
******************** Execution ended in 00h 00m 25.40s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.1092 training's macroF1: 0.488976 valid_1's multi_logloss: 1.09781 valid_1's macroF1: 0.425358
[1000] training's multi_logloss: 1.04403 training's macroF1: 0.516806 valid_1's multi_logloss: 1.08192 valid_1's macroF1: 0.413425
Early stopping, best iteration is:
[517] training's multi_logloss: 1.10579 training's macroF1: 0.488927 valid_1's multi_logloss: 1.09574 valid_1's macroF1: 0.429606
******************** Execution ended in 00h 00m 38.88s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.1075 training's macroF1: 0.499065 valid_1's multi_logloss: 1.09136 valid_1's macroF1: 0.383037
Early stopping, best iteration is:
[148] training's multi_logloss: 1.22991 training's macroF1: 0.468456 valid_1's multi_logloss: 1.18295 valid_1's macroF1: 0.398795
******************** Execution ended in 00h 00m 25.06s ********************
######################################## 26 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.898636 training's macroF1: 0.602686 valid_1's multi_logloss: 1.01344 valid_1's macroF1: 0.416381
[1000] training's multi_logloss: 0.754101 training's macroF1: 0.678634 valid_1's multi_logloss: 1.00585 valid_1's macroF1: 0.42001
Early stopping, best iteration is:
[964] training's multi_logloss: 0.761773 training's macroF1: 0.673882 valid_1's multi_logloss: 1.00345 valid_1's macroF1: 0.421104
******************** Execution ended in 00h 01m 49.52s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.907838 training's macroF1: 0.619255 valid_1's multi_logloss: 1.04572 valid_1's macroF1: 0.403503
[1000] training's multi_logloss: 0.765939 training's macroF1: 0.680588 valid_1's multi_logloss: 1.03142 valid_1's macroF1: 0.415198
[1500] training's multi_logloss: 0.674712 training's macroF1: 0.733827 valid_1's multi_logloss: 1.01706 valid_1's macroF1: 0.427487
[2000] training's multi_logloss: 0.610025 training's macroF1: 0.766093 valid_1's multi_logloss: 1.01442 valid_1's macroF1: 0.425556
[2500] training's multi_logloss: 0.559445 training's macroF1: 0.791199 valid_1's multi_logloss: 1.01278 valid_1's macroF1: 0.433576
Early stopping, best iteration is:
[2064] training's multi_logloss: 0.602841 training's macroF1: 0.770507 valid_1's multi_logloss: 1.01589 valid_1's macroF1: 0.438177
******************** Execution ended in 00h 03m 09.63s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.895739 training's macroF1: 0.626521 valid_1's multi_logloss: 1.08062 valid_1's macroF1: 0.385384
Early stopping, best iteration is:
[26] training's multi_logloss: 1.31069 training's macroF1: 0.480752 valid_1's multi_logloss: 1.3042 valid_1's macroF1: 0.420754
******************** Execution ended in 00h 00m 41.80s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.922218 training's macroF1: 0.615951 valid_1's multi_logloss: 0.956162 valid_1's macroF1: 0.486686
Early stopping, best iteration is:
[431] training's multi_logloss: 0.950119 training's macroF1: 0.596526 valid_1's multi_logloss: 0.963887 valid_1's macroF1: 0.492032
******************** Execution ended in 00h 01m 11.48s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.90137 training's macroF1: 0.618391 valid_1's multi_logloss: 1.07553 valid_1's macroF1: 0.389735
Early stopping, best iteration is:
[407] training's multi_logloss: 0.940399 training's macroF1: 0.602883 valid_1's multi_logloss: 1.0817 valid_1's macroF1: 0.405909
******************** Execution ended in 00h 01m 09.97s ********************
######################################## 27 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.964246 training's macroF1: 0.584653 valid_1's multi_logloss: 1.02534 valid_1's macroF1: 0.452632
[1000] training's multi_logloss: 0.83076 training's macroF1: 0.64756 valid_1's multi_logloss: 1.00605 valid_1's macroF1: 0.463799
Early stopping, best iteration is:
[848] training's multi_logloss: 0.864505 training's macroF1: 0.638089 valid_1's multi_logloss: 1.01154 valid_1's macroF1: 0.467193
******************** Execution ended in 00h 01m 15.27s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.953429 training's macroF1: 0.587224 valid_1's multi_logloss: 1.03686 valid_1's macroF1: 0.400687
Early stopping, best iteration is:
[45] training's multi_logloss: 1.29274 training's macroF1: 0.495714 valid_1's multi_logloss: 1.27677 valid_1's macroF1: 0.408356
******************** Execution ended in 00h 00m 31.04s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.949186 training's macroF1: 0.595898 valid_1's multi_logloss: 1.0566 valid_1's macroF1: 0.368725
Early stopping, best iteration is:
[14] training's multi_logloss: 1.35418 training's macroF1: 0.464546 valid_1's multi_logloss: 1.34839 valid_1's macroF1: 0.380296
******************** Execution ended in 00h 00m 28.75s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.959856 training's macroF1: 0.591938 valid_1's multi_logloss: 1.02286 valid_1's macroF1: 0.436531
[1000] training's multi_logloss: 0.825508 training's macroF1: 0.665823 valid_1's multi_logloss: 1.01105 valid_1's macroF1: 0.448329
[1500] training's multi_logloss: 0.737028 training's macroF1: 0.710174 valid_1's multi_logloss: 1.00841 valid_1's macroF1: 0.448708
[2000] training's multi_logloss: 0.671124 training's macroF1: 0.741354 valid_1's multi_logloss: 1.00777 valid_1's macroF1: 0.464259
Early stopping, best iteration is:
[1839] training's multi_logloss: 0.690865 training's macroF1: 0.730756 valid_1's multi_logloss: 1.00276 valid_1's macroF1: 0.456267
******************** Execution ended in 00h 02m 09.13s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.962607 training's macroF1: 0.596246 valid_1's multi_logloss: 1.02851 valid_1's macroF1: 0.429108
[1000] training's multi_logloss: 0.82895 training's macroF1: 0.66177 valid_1's multi_logloss: 1.00816 valid_1's macroF1: 0.422174
Early stopping, best iteration is:
[613] training's multi_logloss: 0.924642 training's macroF1: 0.610231 valid_1's multi_logloss: 1.01979 valid_1's macroF1: 0.437229
******************** Execution ended in 00h 01m 02.93s ********************
######################################## 28 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.1064 training's macroF1: 0.621124 valid_1's multi_logloss: 1.1217 valid_1's macroF1: 0.386253
Early stopping, best iteration is:
[20] training's multi_logloss: 1.36984 training's macroF1: 0.525691 valid_1's multi_logloss: 1.36778 valid_1's macroF1: 0.412845
******************** Execution ended in 00h 00m 48.20s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.10942 training's macroF1: 0.61295 valid_1's multi_logloss: 1.1186 valid_1's macroF1: 0.404855
[1000] training's multi_logloss: 0.958931 training's macroF1: 0.659805 valid_1's multi_logloss: 1.0312 valid_1's macroF1: 0.409979
Early stopping, best iteration is:
[576] training's multi_logloss: 1.0818 training's macroF1: 0.620317 valid_1's multi_logloss: 1.10014 valid_1's macroF1: 0.417177
******************** Execution ended in 00h 01m 50.96s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.1085 training's macroF1: 0.611803 valid_1's multi_logloss: 1.13382 valid_1's macroF1: 0.426722
Early stopping, best iteration is:
[58] training's multi_logloss: 1.34197 training's macroF1: 0.559251 valid_1's multi_logloss: 1.33674 valid_1's macroF1: 0.434148
******************** Execution ended in 00h 00m 52.16s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.10688 training's macroF1: 0.623805 valid_1's multi_logloss: 1.15181 valid_1's macroF1: 0.419124
[1000] training's multi_logloss: 0.955646 training's macroF1: 0.678811 valid_1's multi_logloss: 1.07043 valid_1's macroF1: 0.421485
Early stopping, best iteration is:
[691] training's multi_logloss: 1.04024 training's macroF1: 0.641187 valid_1's multi_logloss: 1.11263 valid_1's macroF1: 0.428778
******************** Execution ended in 00h 02m 01.56s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.10045 training's macroF1: 0.612875 valid_1's multi_logloss: 1.15994 valid_1's macroF1: 0.368683
[1000] training's multi_logloss: 0.945354 training's macroF1: 0.655267 valid_1's multi_logloss: 1.09103 valid_1's macroF1: 0.366159
Early stopping, best iteration is:
[526] training's multi_logloss: 1.0904 training's macroF1: 0.618306 valid_1's multi_logloss: 1.15403 valid_1's macroF1: 0.377613
******************** Execution ended in 00h 01m 42.45s ********************
######################################## 29 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.03807 training's macroF1: 0.525745 valid_1's multi_logloss: 1.09533 valid_1's macroF1: 0.36886
Early stopping, best iteration is:
[21] training's multi_logloss: 1.34868 training's macroF1: 0.464098 valid_1's multi_logloss: 1.34108 valid_1's macroF1: 0.380429
******************** Execution ended in 00h 00m 26.17s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.05208 training's macroF1: 0.54688 valid_1's multi_logloss: 1.08021 valid_1's macroF1: 0.404179
Early stopping, best iteration is:
[38] training's multi_logloss: 1.32729 training's macroF1: 0.466128 valid_1's multi_logloss: 1.30951 valid_1's macroF1: 0.428354
******************** Execution ended in 00h 00m 26.05s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.06078 training's macroF1: 0.536397 valid_1's multi_logloss: 1.06636 valid_1's macroF1: 0.422467
[1000] training's multi_logloss: 0.96103 training's macroF1: 0.575414 valid_1's multi_logloss: 1.03914 valid_1's macroF1: 0.430793
Early stopping, best iteration is:
[735] training's multi_logloss: 1.00676 training's macroF1: 0.555576 valid_1's multi_logloss: 1.0481 valid_1's macroF1: 0.44291
******************** Execution ended in 00h 00m 58.41s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.05332 training's macroF1: 0.5335 valid_1's multi_logloss: 1.02225 valid_1's macroF1: 0.413933
Early stopping, best iteration is:
[400] training's multi_logloss: 1.08394 training's macroF1: 0.526896 valid_1's multi_logloss: 1.03935 valid_1's macroF1: 0.423188
******************** Execution ended in 00h 00m 43.51s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.04599 training's macroF1: 0.543112 valid_1's multi_logloss: 1.07255 valid_1's macroF1: 0.406737
[1000] training's multi_logloss: 0.942729 training's macroF1: 0.582952 valid_1's multi_logloss: 1.05751 valid_1's macroF1: 0.394757
Early stopping, best iteration is:
[712] training's multi_logloss: 0.994178 training's macroF1: 0.561105 valid_1's multi_logloss: 1.06215 valid_1's macroF1: 0.414359
******************** Execution ended in 00h 00m 58.65s ********************
######################################## 30 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.645191 training's macroF1: 0.782338 valid_1's multi_logloss: 0.971783 valid_1's macroF1: 0.401896
Early stopping, best iteration is:
[231] training's multi_logloss: 0.866802 training's macroF1: 0.72021 valid_1's multi_logloss: 1.02972 valid_1's macroF1: 0.424846
******************** Execution ended in 00h 01m 13.83s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.63992 training's macroF1: 0.779209 valid_1's multi_logloss: 0.972584 valid_1's macroF1: 0.416659
Early stopping, best iteration is:
[30] training's multi_logloss: 1.26468 training's macroF1: 0.592983 valid_1's multi_logloss: 1.26968 valid_1's macroF1: 0.448569
******************** Execution ended in 00h 00m 49.72s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.644894 training's macroF1: 0.774475 valid_1's multi_logloss: 1.00835 valid_1's macroF1: 0.41355
Early stopping, best iteration is:
[57] training's multi_logloss: 1.17835 training's macroF1: 0.628495 valid_1's multi_logloss: 1.21817 valid_1's macroF1: 0.435933
******************** Execution ended in 00h 00m 51.30s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.639858 training's macroF1: 0.7926 valid_1's multi_logloss: 0.997143 valid_1's macroF1: 0.420676
[1000] training's multi_logloss: 0.447524 training's macroF1: 0.85148 valid_1's multi_logloss: 0.982106 valid_1's macroF1: 0.42299
[1500] training's multi_logloss: 0.357004 training's macroF1: 0.886419 valid_1's multi_logloss: 0.978627 valid_1's macroF1: 0.429918
Early stopping, best iteration is:
[1392] training's multi_logloss: 0.372025 training's macroF1: 0.880198 valid_1's multi_logloss: 0.979598 valid_1's macroF1: 0.432092
******************** Execution ended in 00h 02m 54.87s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.64265 training's macroF1: 0.774327 valid_1's multi_logloss: 1.00109 valid_1's macroF1: 0.384584
Early stopping, best iteration is:
[20] training's multi_logloss: 1.3018 training's macroF1: 0.595838 valid_1's multi_logloss: 1.30393 valid_1's macroF1: 0.402713
******************** Execution ended in 00h 00m 49.37s ********************
######################################## 31 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.904224 training's macroF1: 0.611103 valid_1's multi_logloss: 1.06779 valid_1's macroF1: 0.385332
Early stopping, best iteration is:
[192] training's multi_logloss: 1.05865 training's macroF1: 0.545753 valid_1's multi_logloss: 1.08913 valid_1's macroF1: 0.401227
******************** Execution ended in 00h 00m 42.10s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.909831 training's macroF1: 0.609324 valid_1's multi_logloss: 1.05674 valid_1's macroF1: 0.402736
Early stopping, best iteration is:
[353] training's multi_logloss: 0.966867 training's macroF1: 0.578357 valid_1's multi_logloss: 1.06388 valid_1's macroF1: 0.413389
******************** Execution ended in 00h 00m 52.05s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.908514 training's macroF1: 0.605167 valid_1's multi_logloss: 1.05894 valid_1's macroF1: 0.397606
Early stopping, best iteration is:
[334] training's multi_logloss: 0.97626 training's macroF1: 0.578249 valid_1's multi_logloss: 1.06431 valid_1's macroF1: 0.413256
******************** Execution ended in 00h 00m 51.43s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.916911 training's macroF1: 0.606472 valid_1's multi_logloss: 1.00978 valid_1's macroF1: 0.444424
[1000] training's multi_logloss: 0.788544 training's macroF1: 0.670914 valid_1's multi_logloss: 0.992011 valid_1's macroF1: 0.439945
Early stopping, best iteration is:
[579] training's multi_logloss: 0.891163 training's macroF1: 0.620844 valid_1's multi_logloss: 1.00626 valid_1's macroF1: 0.446838
******************** Execution ended in 00h 01m 05.15s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.909625 training's macroF1: 0.602566 valid_1's multi_logloss: 1.0206 valid_1's macroF1: 0.42355
Early stopping, best iteration is:
[192] training's multi_logloss: 1.0617 training's macroF1: 0.541712 valid_1's multi_logloss: 1.06113 valid_1's macroF1: 0.446153
******************** Execution ended in 00h 00m 41.95s ********************
######################################## 32 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.05833 training's macroF1: 0.550869 valid_1's multi_logloss: 1.06454 valid_1's macroF1: 0.370216
Early stopping, best iteration is:
[343] training's multi_logloss: 1.11429 training's macroF1: 0.522876 valid_1's multi_logloss: 1.09302 valid_1's macroF1: 0.383378
******************** Execution ended in 00h 00m 39.39s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.06549 training's macroF1: 0.534841 valid_1's multi_logloss: 1.0534 valid_1's macroF1: 0.384535
[1000] training's multi_logloss: 0.962423 training's macroF1: 0.580359 valid_1's multi_logloss: 1.02345 valid_1's macroF1: 0.399813
[1500] training's multi_logloss: 0.895626 training's macroF1: 0.615252 valid_1's multi_logloss: 1.01704 valid_1's macroF1: 0.406948
[2000] training's multi_logloss: 0.84534 training's macroF1: 0.639609 valid_1's multi_logloss: 1.01367 valid_1's macroF1: 0.412471
Early stopping, best iteration is:
[1966] training's multi_logloss: 0.848541 training's macroF1: 0.636192 valid_1's multi_logloss: 1.01381 valid_1's macroF1: 0.420376
******************** Execution ended in 00h 01m 37.62s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.06965 training's macroF1: 0.540439 valid_1's multi_logloss: 1.06914 valid_1's macroF1: 0.41693
Early stopping, best iteration is:
[417] training's multi_logloss: 1.09667 training's macroF1: 0.531259 valid_1's multi_logloss: 1.08425 valid_1's macroF1: 0.431676
******************** Execution ended in 00h 00m 38.11s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.06988 training's macroF1: 0.538012 valid_1's multi_logloss: 1.07106 valid_1's macroF1: 0.372955
Early stopping, best iteration is:
[121] training's multi_logloss: 1.25043 training's macroF1: 0.491658 valid_1's multi_logloss: 1.2189 valid_1's macroF1: 0.398413
******************** Execution ended in 00h 00m 26.03s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.06315 training's macroF1: 0.548751 valid_1's multi_logloss: 1.08975 valid_1's macroF1: 0.395899
[1000] training's multi_logloss: 0.961405 training's macroF1: 0.576417 valid_1's multi_logloss: 1.06971 valid_1's macroF1: 0.400837
Early stopping, best iteration is:
[846] training's multi_logloss: 0.986443 training's macroF1: 0.563439 valid_1's multi_logloss: 1.07041 valid_1's macroF1: 0.408452
******************** Execution ended in 00h 00m 54.59s ********************
######################################## 33 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.869955 training's macroF1: 0.660329 valid_1's multi_logloss: 0.994284 valid_1's macroF1: 0.444177
Early stopping, best iteration is:
[271] training's multi_logloss: 1.00289 training's macroF1: 0.612345 valid_1's multi_logloss: 1.0425 valid_1's macroF1: 0.447601
******************** Execution ended in 00h 01m 08.85s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.864985 training's macroF1: 0.659996 valid_1's multi_logloss: 1.01526 valid_1's macroF1: 0.430609
[1000] training's multi_logloss: 0.697442 training's macroF1: 0.739491 valid_1's multi_logloss: 0.997228 valid_1's macroF1: 0.434441
Early stopping, best iteration is:
[613] training's multi_logloss: 0.817406 training's macroF1: 0.678198 valid_1's multi_logloss: 1.00738 valid_1's macroF1: 0.441088
******************** Execution ended in 00h 01m 38.79s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.862471 training's macroF1: 0.658302 valid_1's multi_logloss: 1.00202 valid_1's macroF1: 0.411821
Early stopping, best iteration is:
[272] training's multi_logloss: 0.995109 training's macroF1: 0.595925 valid_1's multi_logloss: 1.04182 valid_1's macroF1: 0.424301
******************** Execution ended in 00h 01m 08.35s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.864335 training's macroF1: 0.675712 valid_1's multi_logloss: 1.05802 valid_1's macroF1: 0.405308
Early stopping, best iteration is:
[211] training's multi_logloss: 1.04696 training's macroF1: 0.616735 valid_1's multi_logloss: 1.11385 valid_1's macroF1: 0.424722
******************** Execution ended in 00h 01m 02.63s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.853958 training's macroF1: 0.652339 valid_1's multi_logloss: 1.05299 valid_1's macroF1: 0.383835
[1000] training's multi_logloss: 0.690527 training's macroF1: 0.72592 valid_1's multi_logloss: 1.03531 valid_1's macroF1: 0.392891
Early stopping, best iteration is:
[755] training's multi_logloss: 0.758989 training's macroF1: 0.693673 valid_1's multi_logloss: 1.04183 valid_1's macroF1: 0.405534
******************** Execution ended in 00h 01m 55.76s ********************
######################################## 34 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.08916 training's macroF1: 0.50764 valid_1's multi_logloss: 1.10321 valid_1's macroF1: 0.415162
[1000] training's multi_logloss: 1.01135 training's macroF1: 0.545349 valid_1's multi_logloss: 1.08305 valid_1's macroF1: 0.421022
Early stopping, best iteration is:
[540] training's multi_logloss: 1.08063 training's macroF1: 0.505099 valid_1's multi_logloss: 1.10216 valid_1's macroF1: 0.4334
******************** Execution ended in 00h 00m 33.70s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.08482 training's macroF1: 0.521816 valid_1's multi_logloss: 1.05027 valid_1's macroF1: 0.396627
[1000] training's multi_logloss: 1.00221 training's macroF1: 0.545941 valid_1's multi_logloss: 1.03549 valid_1's macroF1: 0.401351
[1500] training's multi_logloss: 0.947808 training's macroF1: 0.578846 valid_1's multi_logloss: 1.03253 valid_1's macroF1: 0.399981
Early stopping, best iteration is:
[1377] training's multi_logloss: 0.959963 training's macroF1: 0.569978 valid_1's multi_logloss: 1.03275 valid_1's macroF1: 0.413781
******************** Execution ended in 00h 00m 59.51s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.07866 training's macroF1: 0.522615 valid_1's multi_logloss: 1.11704 valid_1's macroF1: 0.410071
Early stopping, best iteration is:
[378] training's multi_logloss: 1.11008 training's macroF1: 0.518178 valid_1's multi_logloss: 1.12543 valid_1's macroF1: 0.414591
******************** Execution ended in 00h 00m 28.61s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.08065 training's macroF1: 0.509672 valid_1's multi_logloss: 1.06682 valid_1's macroF1: 0.415609
[1000] training's multi_logloss: 0.99958 training's macroF1: 0.545949 valid_1's multi_logloss: 1.04398 valid_1's macroF1: 0.419905
[1500] training's multi_logloss: 0.94819 training's macroF1: 0.574207 valid_1's multi_logloss: 1.04215 valid_1's macroF1: 0.412599
Early stopping, best iteration is:
[1238] training's multi_logloss: 0.972718 training's macroF1: 0.556604 valid_1's multi_logloss: 1.04332 valid_1's macroF1: 0.427397
******************** Execution ended in 00h 00m 55.03s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.0755 training's macroF1: 0.505342 valid_1's multi_logloss: 1.02663 valid_1's macroF1: 0.390627
Early stopping, best iteration is:
[10] training's multi_logloss: 1.36838 training's macroF1: 0.424073 valid_1's multi_logloss: 1.35602 valid_1's macroF1: 0.415529
******************** Execution ended in 00h 00m 16.59s ********************
######################################## 35 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.12479 training's macroF1: 0.566932 valid_1's multi_logloss: 1.10265 valid_1's macroF1: 0.451481
[1000] training's multi_logloss: 1.00988 training's macroF1: 0.592613 valid_1's multi_logloss: 1.03074 valid_1's macroF1: 0.431464
Early stopping, best iteration is:
[530] training's multi_logloss: 1.1158 training's macroF1: 0.568653 valid_1's multi_logloss: 1.09559 valid_1's macroF1: 0.454459
******************** Execution ended in 00h 00m 52.75s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.11502 training's macroF1: 0.550964 valid_1's multi_logloss: 1.1411 valid_1's macroF1: 0.37313
Early stopping, best iteration is:
[111] training's multi_logloss: 1.29554 training's macroF1: 0.504182 valid_1's multi_logloss: 1.29028 valid_1's macroF1: 0.390227
******************** Execution ended in 00h 00m 30.84s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.1116 training's macroF1: 0.558426 valid_1's multi_logloss: 1.10985 valid_1's macroF1: 0.412323
[1000] training's multi_logloss: 0.99325 training's macroF1: 0.60369 valid_1's multi_logloss: 1.05112 valid_1's macroF1: 0.420128
Early stopping, best iteration is:
[938] training's multi_logloss: 1.00475 training's macroF1: 0.597078 valid_1's multi_logloss: 1.05506 valid_1's macroF1: 0.423926
******************** Execution ended in 00h 01m 13.84s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.12692 training's macroF1: 0.555206 valid_1's multi_logloss: 1.11367 valid_1's macroF1: 0.397537
Early stopping, best iteration is:
[285] training's multi_logloss: 1.20772 training's macroF1: 0.521734 valid_1's multi_logloss: 1.18411 valid_1's macroF1: 0.408667
******************** Execution ended in 00h 00m 39.81s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.11754 training's macroF1: 0.551218 valid_1's multi_logloss: 1.10194 valid_1's macroF1: 0.411145
Early stopping, best iteration is:
[44] training's multi_logloss: 1.34662 training's macroF1: 0.492833 valid_1's multi_logloss: 1.33712 valid_1's macroF1: 0.415093
******************** Execution ended in 00h 00m 26.95s ********************
######################################## 36 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.594738 training's macroF1: 0.802879 valid_1's multi_logloss: 0.955162 valid_1's macroF1: 0.411936
Early stopping, best iteration is:
[178] training's multi_logloss: 0.907338 training's macroF1: 0.713464 valid_1's multi_logloss: 1.03527 valid_1's macroF1: 0.433103
******************** Execution ended in 00h 01m 08.64s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.591779 training's macroF1: 0.80527 valid_1's multi_logloss: 1.03648 valid_1's macroF1: 0.387512
[1000] training's multi_logloss: 0.413517 training's macroF1: 0.865685 valid_1's multi_logloss: 1.0168 valid_1's macroF1: 0.377673
Early stopping, best iteration is:
[943] training's multi_logloss: 0.426294 training's macroF1: 0.863753 valid_1's multi_logloss: 1.01865 valid_1's macroF1: 0.398903
******************** Execution ended in 00h 02m 29.20s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.582462 training's macroF1: 0.817363 valid_1's multi_logloss: 0.975324 valid_1's macroF1: 0.382348
[1000] training's multi_logloss: 0.406282 training's macroF1: 0.866316 valid_1's multi_logloss: 0.972768 valid_1's macroF1: 0.386858
Early stopping, best iteration is:
[639] training's multi_logloss: 0.513666 training's macroF1: 0.837524 valid_1's multi_logloss: 0.971406 valid_1's macroF1: 0.395855
******************** Execution ended in 00h 01m 57.08s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.604397 training's macroF1: 0.800887 valid_1's multi_logloss: 0.953704 valid_1's macroF1: 0.408685
Early stopping, best iteration is:
[449] training's multi_logloss: 0.63664 training's macroF1: 0.784531 valid_1's multi_logloss: 0.961957 valid_1's macroF1: 0.424926
******************** Execution ended in 00h 01m 38.29s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.59178 training's macroF1: 0.803338 valid_1's multi_logloss: 1.01139 valid_1's macroF1: 0.408832
[1000] training's multi_logloss: 0.414392 training's macroF1: 0.86242 valid_1's multi_logloss: 0.986926 valid_1's macroF1: 0.438745
[1500] training's multi_logloss: 0.339732 training's macroF1: 0.886646 valid_1's multi_logloss: 0.979916 valid_1's macroF1: 0.426275
Early stopping, best iteration is:
[1145] training's multi_logloss: 0.387347 training's macroF1: 0.869148 valid_1's multi_logloss: 0.984027 valid_1's macroF1: 0.443808
******************** Execution ended in 00h 02m 40.09s ********************
######################################## 37 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.2577 training's macroF1: 0.488215 valid_1's multi_logloss: 1.2345 valid_1's macroF1: 0.374227
Early stopping, best iteration is:
[7] training's multi_logloss: 1.38388 training's macroF1: 0.447279 valid_1's multi_logloss: 1.38309 valid_1's macroF1: 0.388279
******************** Execution ended in 00h 00m 35.23s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.25996 training's macroF1: 0.480892 valid_1's multi_logloss: 1.24163 valid_1's macroF1: 0.390542
Early stopping, best iteration is:
[51] training's multi_logloss: 1.36978 training's macroF1: 0.462242 valid_1's multi_logloss: 1.36549 valid_1's macroF1: 0.410506
******************** Execution ended in 00h 00m 40.25s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.25776 training's macroF1: 0.491685 valid_1's multi_logloss: 1.23015 valid_1's macroF1: 0.391796
[1000] training's multi_logloss: 1.1749 training's macroF1: 0.514329 valid_1's multi_logloss: 1.15178 valid_1's macroF1: 0.390986
Early stopping, best iteration is:
[751] training's multi_logloss: 1.21168 training's macroF1: 0.50341 valid_1's multi_logloss: 1.18475 valid_1's macroF1: 0.402807
******************** Execution ended in 00h 01m 35.28s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.26241 training's macroF1: 0.494631 valid_1's multi_logloss: 1.23409 valid_1's macroF1: 0.412751
Early stopping, best iteration is:
[402] training's multi_logloss: 1.28258 training's macroF1: 0.484972 valid_1's multi_logloss: 1.25643 valid_1's macroF1: 0.428322
******************** Execution ended in 00h 01m 05.95s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.2558 training's macroF1: 0.517966 valid_1's multi_logloss: 1.24026 valid_1's macroF1: 0.379645
Early stopping, best iteration is:
[16] training's multi_logloss: 1.38082 training's macroF1: 0.470095 valid_1's multi_logloss: 1.37941 valid_1's macroF1: 0.3966
******************** Execution ended in 00h 00m 37.38s ********************
######################################## 38 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.14074 training's macroF1: 0.516031 valid_1's multi_logloss: 1.12059 valid_1's macroF1: 0.392192
[1000] training's multi_logloss: 1.03391 training's macroF1: 0.548701 valid_1's multi_logloss: 1.05962 valid_1's macroF1: 0.402237
Early stopping, best iteration is:
[978] training's multi_logloss: 1.0376 training's macroF1: 0.544943 valid_1's multi_logloss: 1.06128 valid_1's macroF1: 0.409218
******************** Execution ended in 00h 01m 27.50s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.13431 training's macroF1: 0.533157 valid_1's multi_logloss: 1.13219 valid_1's macroF1: 0.383539
[1000] training's multi_logloss: 1.02554 training's macroF1: 0.564812 valid_1's multi_logloss: 1.07394 valid_1's macroF1: 0.395119
[1500] training's multi_logloss: 0.958499 training's macroF1: 0.587929 valid_1's multi_logloss: 1.06045 valid_1's macroF1: 0.398654
Early stopping, best iteration is:
[1293] training's multi_logloss: 0.983309 training's macroF1: 0.579312 valid_1's multi_logloss: 1.06314 valid_1's macroF1: 0.404405
******************** Execution ended in 00h 01m 43.33s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.14383 training's macroF1: 0.52709 valid_1's multi_logloss: 1.12558 valid_1's macroF1: 0.404605
Early stopping, best iteration is:
[390] training's multi_logloss: 1.17823 training's macroF1: 0.518382 valid_1's multi_logloss: 1.15295 valid_1's macroF1: 0.414989
******************** Execution ended in 00h 00m 49.31s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.14268 training's macroF1: 0.531082 valid_1's multi_logloss: 1.10133 valid_1's macroF1: 0.398754
Early stopping, best iteration is:
[27] training's multi_logloss: 1.36433 training's macroF1: 0.45498 valid_1's multi_logloss: 1.35415 valid_1's macroF1: 0.426706
******************** Execution ended in 00h 00m 29.24s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.14384 training's macroF1: 0.533249 valid_1's multi_logloss: 1.12562 valid_1's macroF1: 0.414502
[1000] training's multi_logloss: 1.03884 training's macroF1: 0.564221 valid_1's multi_logloss: 1.06224 valid_1's macroF1: 0.407688
Early stopping, best iteration is:
[527] training's multi_logloss: 1.13625 training's macroF1: 0.535358 valid_1's multi_logloss: 1.12004 valid_1's macroF1: 0.416617
******************** Execution ended in 00h 00m 57.38s ********************
######################################## 39 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.11294 training's macroF1: 0.484685 valid_1's multi_logloss: 1.08443 valid_1's macroF1: 0.412734
Early stopping, best iteration is:
[260] training's multi_logloss: 1.17315 training's macroF1: 0.475675 valid_1's multi_logloss: 1.12432 valid_1's macroF1: 0.427565
******************** Execution ended in 00h 00m 21.88s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.10516 training's macroF1: 0.484149 valid_1's multi_logloss: 1.06738 valid_1's macroF1: 0.389824
Early stopping, best iteration is:
[441] training's multi_logloss: 1.11608 training's macroF1: 0.485254 valid_1's multi_logloss: 1.07286 valid_1's macroF1: 0.39981
******************** Execution ended in 00h 00m 26.72s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.1143 training's macroF1: 0.481821 valid_1's multi_logloss: 1.07888 valid_1's macroF1: 0.401004
Early stopping, best iteration is:
[335] training's multi_logloss: 1.15022 training's macroF1: 0.475125 valid_1's multi_logloss: 1.09236 valid_1's macroF1: 0.419557
******************** Execution ended in 00h 00m 24.04s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.10833 training's macroF1: 0.483019 valid_1's multi_logloss: 1.09362 valid_1's macroF1: 0.394489
[1000] training's multi_logloss: 1.04626 training's macroF1: 0.517864 valid_1's multi_logloss: 1.08117 valid_1's macroF1: 0.402033
Early stopping, best iteration is:
[992] training's multi_logloss: 1.04705 training's macroF1: 0.514815 valid_1's multi_logloss: 1.08122 valid_1's macroF1: 0.405579
******************** Execution ended in 00h 00m 42.15s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.113 training's macroF1: 0.482034 valid_1's multi_logloss: 1.1112 valid_1's macroF1: 0.393716
Early stopping, best iteration is:
[10] training's multi_logloss: 1.37023 training's macroF1: 0.412393 valid_1's multi_logloss: 1.36237 valid_1's macroF1: 0.402828
******************** Execution ended in 00h 00m 15.02s ********************
######################################## 40 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.04014 training's macroF1: 0.520467 valid_1's multi_logloss: 1.02889 valid_1's macroF1: 0.426815
Early stopping, best iteration is:
[373] training's multi_logloss: 1.07351 training's macroF1: 0.506163 valid_1's multi_logloss: 1.0348 valid_1's macroF1: 0.43991
******************** Execution ended in 00h 00m 39.45s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.03129 training's macroF1: 0.535138 valid_1's multi_logloss: 1.05016 valid_1's macroF1: 0.416594
[1000] training's multi_logloss: 0.946173 training's macroF1: 0.576429 valid_1's multi_logloss: 1.04836 valid_1's macroF1: 0.425101
Early stopping, best iteration is:
[890] training's multi_logloss: 0.961534 training's macroF1: 0.56605 valid_1's multi_logloss: 1.05001 valid_1's macroF1: 0.433966
******************** Execution ended in 00h 00m 57.00s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.02408 training's macroF1: 0.532041 valid_1's multi_logloss: 1.1079 valid_1's macroF1: 0.38635
Early stopping, best iteration is:
[259] training's multi_logloss: 1.09935 training's macroF1: 0.4964 valid_1's multi_logloss: 1.12086 valid_1's macroF1: 0.397631
******************** Execution ended in 00h 00m 31.36s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.02369 training's macroF1: 0.532613 valid_1's multi_logloss: 1.05888 valid_1's macroF1: 0.39882
Early stopping, best iteration is:
[5] training's multi_logloss: 1.37137 training's macroF1: 0.416943 valid_1's multi_logloss: 1.36531 valid_1's macroF1: 0.410584
******************** Execution ended in 00h 00m 21.11s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.02547 training's macroF1: 0.533093 valid_1's multi_logloss: 1.08358 valid_1's macroF1: 0.388974
Early stopping, best iteration is:
[104] training's multi_logloss: 1.19854 training's macroF1: 0.473315 valid_1's multi_logloss: 1.16579 valid_1's macroF1: 0.405119
******************** Execution ended in 00h 00m 25.32s ********************
######################################## 41 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.23509 training's macroF1: 0.503037 valid_1's multi_logloss: 1.20367 valid_1's macroF1: 0.392491
Early stopping, best iteration is:
[54] training's multi_logloss: 1.36366 training's macroF1: 0.470994 valid_1's multi_logloss: 1.3557 valid_1's macroF1: 0.414576
******************** Execution ended in 00h 00m 25.78s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.23808 training's macroF1: 0.47522 valid_1's multi_logloss: 1.19177 valid_1's macroF1: 0.390301
Early stopping, best iteration is:
[23] training's multi_logloss: 1.37648 training's macroF1: 0.448003 valid_1's multi_logloss: 1.37203 valid_1's macroF1: 0.396884
******************** Execution ended in 00h 00m 23.78s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.2318 training's macroF1: 0.499995 valid_1's multi_logloss: 1.20854 valid_1's macroF1: 0.363396
Early stopping, best iteration is:
[1] training's multi_logloss: 1.38584 training's macroF1: 0.417351 valid_1's multi_logloss: 1.38565 valid_1's macroF1: 0.382147
******************** Execution ended in 00h 00m 22.49s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.23005 training's macroF1: 0.482272 valid_1's multi_logloss: 1.2131 valid_1's macroF1: 0.368922
Early stopping, best iteration is:
[165] training's multi_logloss: 1.32168 training's macroF1: 0.465481 valid_1's multi_logloss: 1.3094 valid_1's macroF1: 0.389235
******************** Execution ended in 00h 00m 29.59s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.23676 training's macroF1: 0.497106 valid_1's multi_logloss: 1.18902 valid_1's macroF1: 0.401446
Early stopping, best iteration is:
[165] training's multi_logloss: 1.32543 training's macroF1: 0.473608 valid_1's multi_logloss: 1.29781 valid_1's macroF1: 0.428447
******************** Execution ended in 00h 00m 29.77s ********************
######################################## 42 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.889546 training's macroF1: 0.652826 valid_1's multi_logloss: 1.0798 valid_1's macroF1: 0.383314
Early stopping, best iteration is:
[32] training's multi_logloss: 1.3063 training's macroF1: 0.531342 valid_1's multi_logloss: 1.309 valid_1's macroF1: 0.385634
******************** Execution ended in 00h 00m 42.10s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.896787 training's macroF1: 0.647364 valid_1's multi_logloss: 1.01983 valid_1's macroF1: 0.438611
[1000] training's multi_logloss: 0.741668 training's macroF1: 0.71529 valid_1's multi_logloss: 0.992336 valid_1's macroF1: 0.444097
Early stopping, best iteration is:
[780] training's multi_logloss: 0.801303 training's macroF1: 0.681928 valid_1's multi_logloss: 1.00041 valid_1's macroF1: 0.457308
******************** Execution ended in 00h 01m 40.72s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.907669 training's macroF1: 0.629332 valid_1's multi_logloss: 0.974822 valid_1's macroF1: 0.448167
[1000] training's multi_logloss: 0.746819 training's macroF1: 0.697101 valid_1's multi_logloss: 0.942006 valid_1's macroF1: 0.450647
[1500] training's multi_logloss: 0.638433 training's macroF1: 0.752848 valid_1's multi_logloss: 0.930953 valid_1's macroF1: 0.467112
[2000] training's multi_logloss: 0.560742 training's macroF1: 0.789906 valid_1's multi_logloss: 0.923515 valid_1's macroF1: 0.472055
[2500] training's multi_logloss: 0.501909 training's macroF1: 0.815554 valid_1's multi_logloss: 0.920454 valid_1's macroF1: 0.483169
Early stopping, best iteration is:
[2249] training's multi_logloss: 0.529697 training's macroF1: 0.801963 valid_1's multi_logloss: 0.922251 valid_1's macroF1: 0.486116
******************** Execution ended in 00h 03m 27.72s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.898249 training's macroF1: 0.649344 valid_1's multi_logloss: 1.04874 valid_1's macroF1: 0.422564
Early stopping, best iteration is:
[397] training's multi_logloss: 0.944895 training's macroF1: 0.631595 valid_1's multi_logloss: 1.06135 valid_1's macroF1: 0.437156
******************** Execution ended in 00h 01m 13.37s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.898129 training's macroF1: 0.639093 valid_1's multi_logloss: 1.03037 valid_1's macroF1: 0.413664
Early stopping, best iteration is:
[121] training's multi_logloss: 1.16143 training's macroF1: 0.554013 valid_1's multi_logloss: 1.16101 valid_1's macroF1: 0.427538
******************** Execution ended in 00h 00m 51.84s ********************
######################################## 43 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.11196 training's macroF1: 0.504259 valid_1's multi_logloss: 1.0803 valid_1's macroF1: 0.389258
[1000] training's multi_logloss: 1.03571 training's macroF1: 0.524563 valid_1's multi_logloss: 1.05815 valid_1's macroF1: 0.382977
Early stopping, best iteration is:
[877] training's multi_logloss: 1.04996 training's macroF1: 0.519737 valid_1's multi_logloss: 1.0593 valid_1's macroF1: 0.408931
******************** Execution ended in 00h 00m 48.36s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.12878 training's macroF1: 0.493121 valid_1's multi_logloss: 1.09028 valid_1's macroF1: 0.409471
[1000] training's multi_logloss: 1.05606 training's macroF1: 0.519611 valid_1's multi_logloss: 1.05386 valid_1's macroF1: 0.419115
Early stopping, best iteration is:
[979] training's multi_logloss: 1.0583 training's macroF1: 0.520011 valid_1's multi_logloss: 1.05508 valid_1's macroF1: 0.420111
******************** Execution ended in 00h 00m 51.35s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.12305 training's macroF1: 0.490997 valid_1's multi_logloss: 1.06293 valid_1's macroF1: 0.465243
Early stopping, best iteration is:
[384] training's multi_logloss: 1.15304 training's macroF1: 0.475243 valid_1's multi_logloss: 1.08284 valid_1's macroF1: 0.467708
******************** Execution ended in 00h 00m 31.41s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.12727 training's macroF1: 0.499517 valid_1's multi_logloss: 1.11572 valid_1's macroF1: 0.361759
Early stopping, best iteration is:
[60] training's multi_logloss: 1.31666 training's macroF1: 0.453537 valid_1's multi_logloss: 1.29174 valid_1's macroF1: 0.380608
******************** Execution ended in 00h 00m 20.51s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.11776 training's macroF1: 0.486488 valid_1's multi_logloss: 1.12242 valid_1's macroF1: 0.391126
[1000] training's multi_logloss: 1.04245 training's macroF1: 0.524056 valid_1's multi_logloss: 1.10541 valid_1's macroF1: 0.408292
[1500] training's multi_logloss: 0.995414 training's macroF1: 0.551524 valid_1's multi_logloss: 1.10192 valid_1's macroF1: 0.416645
[2000] training's multi_logloss: 0.959362 training's macroF1: 0.563075 valid_1's multi_logloss: 1.10111 valid_1's macroF1: 0.411329
Early stopping, best iteration is:
[1572] training's multi_logloss: 0.989864 training's macroF1: 0.552538 valid_1's multi_logloss: 1.10227 valid_1's macroF1: 0.42512
******************** Execution ended in 00h 01m 13.23s ********************
######################################## 44 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.820078 training's macroF1: 0.665156 valid_1's multi_logloss: 1.02735 valid_1's macroF1: 0.414773
Early stopping, best iteration is:
[187] training's multi_logloss: 1.01082 training's macroF1: 0.572393 valid_1's multi_logloss: 1.06414 valid_1's macroF1: 0.441662
******************** Execution ended in 00h 00m 44.45s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.80254 training's macroF1: 0.666306 valid_1's multi_logloss: 1.04256 valid_1's macroF1: 0.388427
[1000] training's multi_logloss: 0.647958 training's macroF1: 0.742893 valid_1's multi_logloss: 1.03417 valid_1's macroF1: 0.388625
Early stopping, best iteration is:
[787] training's multi_logloss: 0.702192 training's macroF1: 0.71535 valid_1's multi_logloss: 1.03752 valid_1's macroF1: 0.405027
******************** Execution ended in 00h 01m 22.27s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.808003 training's macroF1: 0.678017 valid_1's multi_logloss: 1.01835 valid_1's macroF1: 0.432496
Early stopping, best iteration is:
[189] training's multi_logloss: 1.00249 training's macroF1: 0.584859 valid_1's multi_logloss: 1.0591 valid_1's macroF1: 0.443377
******************** Execution ended in 00h 00m 45.26s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.822026 training's macroF1: 0.664212 valid_1's multi_logloss: 0.983132 valid_1's macroF1: 0.422824
Early stopping, best iteration is:
[131] training's multi_logloss: 1.07427 training's macroF1: 0.559816 valid_1's multi_logloss: 1.07254 valid_1's macroF1: 0.442215
******************** Execution ended in 00h 00m 40.11s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.797009 training's macroF1: 0.667564 valid_1's multi_logloss: 1.03259 valid_1's macroF1: 0.402937
Early stopping, best iteration is:
[111] training's multi_logloss: 1.09313 training's macroF1: 0.552966 valid_1's multi_logloss: 1.09779 valid_1's macroF1: 0.421831
******************** Execution ended in 00h 00m 43.89s ********************
######################################## 45 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.13046 training's macroF1: 0.542273 valid_1's multi_logloss: 1.14415 valid_1's macroF1: 0.407926
Early stopping, best iteration is:
[2] training's multi_logloss: 1.38451 training's macroF1: 0.444451 valid_1's multi_logloss: 1.38406 valid_1's macroF1: 0.418685
******************** Execution ended in 00h 00m 32.22s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.1383 training's macroF1: 0.530167 valid_1's multi_logloss: 1.11988 valid_1's macroF1: 0.41558
[1000] training's multi_logloss: 1.0312 training's macroF1: 0.561026 valid_1's multi_logloss: 1.04854 valid_1's macroF1: 0.416117
Early stopping, best iteration is:
[743] training's multi_logloss: 1.07809 training's macroF1: 0.542512 valid_1's multi_logloss: 1.07468 valid_1's macroF1: 0.423817
******************** Execution ended in 00h 01m 19.51s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.13462 training's macroF1: 0.527095 valid_1's multi_logloss: 1.11482 valid_1's macroF1: 0.389953
Early stopping, best iteration is:
[323] training's multi_logloss: 1.19679 training's macroF1: 0.508338 valid_1's multi_logloss: 1.16876 valid_1's macroF1: 0.406442
******************** Execution ended in 00h 00m 52.42s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.13123 training's macroF1: 0.539203 valid_1's multi_logloss: 1.12962 valid_1's macroF1: 0.377098
[1000] training's multi_logloss: 1.02152 training's macroF1: 0.572134 valid_1's multi_logloss: 1.0657 valid_1's macroF1: 0.386884
[1500] training's multi_logloss: 0.953087 training's macroF1: 0.594816 valid_1's multi_logloss: 1.04669 valid_1's macroF1: 0.392709
Early stopping, best iteration is:
[1469] training's multi_logloss: 0.956754 training's macroF1: 0.593971 valid_1's multi_logloss: 1.0474 valid_1's macroF1: 0.399109
******************** Execution ended in 00h 02m 02.05s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.14035 training's macroF1: 0.530468 valid_1's multi_logloss: 1.13482 valid_1's macroF1: 0.391258
[1000] training's multi_logloss: 1.02942 training's macroF1: 0.558691 valid_1's multi_logloss: 1.06955 valid_1's macroF1: 0.397508
[1500] training's multi_logloss: 0.960629 training's macroF1: 0.590336 valid_1's multi_logloss: 1.04884 valid_1's macroF1: 0.423985
[2000] training's multi_logloss: 0.909239 training's macroF1: 0.610841 valid_1's multi_logloss: 1.04082 valid_1's macroF1: 0.424011
Early stopping, best iteration is:
[1665] training's multi_logloss: 0.942381 training's macroF1: 0.597739 valid_1's multi_logloss: 1.04549 valid_1's macroF1: 0.42744
******************** Execution ended in 00h 02m 12.88s ********************
######################################## 46 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.23944 training's macroF1: 0.466684 valid_1's multi_logloss: 1.1629 valid_1's macroF1: 0.438267
Early stopping, best iteration is:
[39] training's multi_logloss: 1.37005 training's macroF1: 0.437431 valid_1's multi_logloss: 1.35939 valid_1's macroF1: 0.442367
******************** Execution ended in 00h 00m 26.36s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.2395 training's macroF1: 0.473864 valid_1's multi_logloss: 1.18588 valid_1's macroF1: 0.419919
Early stopping, best iteration is:
[58] training's multi_logloss: 1.3622 training's macroF1: 0.440625 valid_1's multi_logloss: 1.3493 valid_1's macroF1: 0.42835
******************** Execution ended in 00h 00m 26.33s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.23113 training's macroF1: 0.472421 valid_1's multi_logloss: 1.19936 valid_1's macroF1: 0.379946
Early stopping, best iteration is:
[458] training's multi_logloss: 1.2402 training's macroF1: 0.471238 valid_1's multi_logloss: 1.20833 valid_1's macroF1: 0.389444
******************** Execution ended in 00h 00m 45.80s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.23263 training's macroF1: 0.460599 valid_1's multi_logloss: 1.20267 valid_1's macroF1: 0.369858
Early stopping, best iteration is:
[362] training's multi_logloss: 1.26398 training's macroF1: 0.450829 valid_1's multi_logloss: 1.23575 valid_1's macroF1: 0.380211
******************** Execution ended in 00h 00m 42.09s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.22938 training's macroF1: 0.478254 valid_1's multi_logloss: 1.21713 valid_1's macroF1: 0.387789
Early stopping, best iteration is:
[404] training's multi_logloss: 1.25086 training's macroF1: 0.472834 valid_1's multi_logloss: 1.23782 valid_1's macroF1: 0.393494
******************** Execution ended in 00h 00m 47.69s ********************
######################################## 47 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.00031 training's macroF1: 0.626342 valid_1's multi_logloss: 1.06127 valid_1's macroF1: 0.409881
Early stopping, best iteration is:
[428] training's multi_logloss: 1.03115 training's macroF1: 0.616777 valid_1's multi_logloss: 1.07478 valid_1's macroF1: 0.422387
******************** Execution ended in 00h 00m 49.58s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.00173 training's macroF1: 0.614745 valid_1's multi_logloss: 1.0657 valid_1's macroF1: 0.419639
[1000] training's multi_logloss: 0.855671 training's macroF1: 0.659258 valid_1's multi_logloss: 1.0181 valid_1's macroF1: 0.417516
[1500] training's multi_logloss: 0.760623 training's macroF1: 0.703882 valid_1's multi_logloss: 1.00042 valid_1's macroF1: 0.428468
[2000] training's multi_logloss: 0.689996 training's macroF1: 0.737662 valid_1's multi_logloss: 0.991489 valid_1's macroF1: 0.423894
Early stopping, best iteration is:
[1729] training's multi_logloss: 0.725886 training's macroF1: 0.721141 valid_1's multi_logloss: 0.995122 valid_1's macroF1: 0.43257
******************** Execution ended in 00h 01m 59.01s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.00725 training's macroF1: 0.618185 valid_1's multi_logloss: 1.04318 valid_1's macroF1: 0.421554
Early stopping, best iteration is:
[5] training's multi_logloss: 1.37856 training's macroF1: 0.484139 valid_1's multi_logloss: 1.37578 valid_1's macroF1: 0.433348
******************** Execution ended in 00h 00m 24.84s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.0089 training's macroF1: 0.607344 valid_1's multi_logloss: 1.04721 valid_1's macroF1: 0.405481
Early stopping, best iteration is:
[71] training's multi_logloss: 1.2926 training's macroF1: 0.53826 valid_1's multi_logloss: 1.27648 valid_1's macroF1: 0.422041
******************** Execution ended in 00h 00m 28.76s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.998805 training's macroF1: 0.609789 valid_1's multi_logloss: 1.04472 valid_1's macroF1: 0.374976
Early stopping, best iteration is:
[178] training's multi_logloss: 1.18252 training's macroF1: 0.566823 valid_1's multi_logloss: 1.17038 valid_1's macroF1: 0.393856
******************** Execution ended in 00h 00m 35.03s ********************
######################################## 48 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.8071 training's macroF1: 0.662279 valid_1's multi_logloss: 1.01148 valid_1's macroF1: 0.404122
Early stopping, best iteration is:
[462] training's multi_logloss: 0.824204 training's macroF1: 0.656333 valid_1's multi_logloss: 1.01275 valid_1's macroF1: 0.412464
******************** Execution ended in 00h 00m 51.20s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.799244 training's macroF1: 0.674325 valid_1's multi_logloss: 0.971178 valid_1's macroF1: 0.448099
Early stopping, best iteration is:
[283] training's multi_logloss: 0.918767 training's macroF1: 0.622818 valid_1's multi_logloss: 0.989862 valid_1's macroF1: 0.474071
******************** Execution ended in 00h 00m 41.54s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.802105 training's macroF1: 0.681719 valid_1's multi_logloss: 1.02001 valid_1's macroF1: 0.410017
Early stopping, best iteration is:
[338] training's multi_logloss: 0.883788 training's macroF1: 0.643205 valid_1's multi_logloss: 1.02797 valid_1's macroF1: 0.429726
******************** Execution ended in 00h 00m 42.93s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.788106 training's macroF1: 0.676333 valid_1's multi_logloss: 1.07599 valid_1's macroF1: 0.367092
Early stopping, best iteration is:
[104] training's multi_logloss: 1.08921 training's macroF1: 0.559944 valid_1's multi_logloss: 1.13602 valid_1's macroF1: 0.398403
******************** Execution ended in 00h 00m 30.57s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.794707 training's macroF1: 0.668448 valid_1's multi_logloss: 1.04107 valid_1's macroF1: 0.392949
Early stopping, best iteration is:
[173] training's multi_logloss: 1.00595 training's macroF1: 0.572598 valid_1's multi_logloss: 1.07324 valid_1's macroF1: 0.409823
******************** Execution ended in 00h 00m 34.36s ********************
######################################## 49 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.26168 training's macroF1: 0.487832 valid_1's multi_logloss: 1.22688 valid_1's macroF1: 0.391319
Early stopping, best iteration is:
[13] training's multi_logloss: 1.38197 training's macroF1: 0.433916 valid_1's multi_logloss: 1.38039 valid_1's macroF1: 0.399763
******************** Execution ended in 00h 00m 26.78s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.25725 training's macroF1: 0.49344 valid_1's multi_logloss: 1.23395 valid_1's macroF1: 0.363941
Early stopping, best iteration is:
[28] training's multi_logloss: 1.37696 training's macroF1: 0.46196 valid_1's multi_logloss: 1.3748 valid_1's macroF1: 0.382518
******************** Execution ended in 00h 00m 28.31s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.25907 training's macroF1: 0.492664 valid_1's multi_logloss: 1.23247 valid_1's macroF1: 0.384621
Early stopping, best iteration is:
[16] training's multi_logloss: 1.38094 training's macroF1: 0.462733 valid_1's multi_logloss: 1.37894 valid_1's macroF1: 0.417659
******************** Execution ended in 00h 00m 27.27s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.25378 training's macroF1: 0.503521 valid_1's multi_logloss: 1.23431 valid_1's macroF1: 0.384534
Early stopping, best iteration is:
[33] training's multi_logloss: 1.37508 training's macroF1: 0.445598 valid_1's multi_logloss: 1.37155 valid_1's macroF1: 0.403611
******************** Execution ended in 00h 00m 28.30s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.25653 training's macroF1: 0.492158 valid_1's multi_logloss: 1.22414 valid_1's macroF1: 0.388131
Early stopping, best iteration is:
[62] training's multi_logloss: 1.36623 training's macroF1: 0.462166 valid_1's multi_logloss: 1.35954 valid_1's macroF1: 0.410976
******************** Execution ended in 00h 00m 30.75s ********************
######################################## 50 of 50 iterations ########################################
============================== 1 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.20308 training's macroF1: 0.492346 valid_1's multi_logloss: 1.13667 valid_1's macroF1: 0.439857
[1000] training's multi_logloss: 1.11532 training's macroF1: 0.509183 valid_1's multi_logloss: 1.05185 valid_1's macroF1: 0.436571
Early stopping, best iteration is:
[542] training's multi_logloss: 1.19321 training's macroF1: 0.489722 valid_1's multi_logloss: 1.1252 valid_1's macroF1: 0.452032
******************** Execution ended in 00h 00m 58.62s ********************
============================== 2 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.20211 training's macroF1: 0.502901 valid_1's multi_logloss: 1.16359 valid_1's macroF1: 0.394206
[1000] training's multi_logloss: 1.11377 training's macroF1: 0.53092 valid_1's multi_logloss: 1.09168 valid_1's macroF1: 0.40153
[1500] training's multi_logloss: 1.05951 training's macroF1: 0.5409 valid_1's multi_logloss: 1.06526 valid_1's macroF1: 0.404961
Early stopping, best iteration is:
[1432] training's multi_logloss: 1.06584 training's macroF1: 0.539975 valid_1's multi_logloss: 1.06719 valid_1's macroF1: 0.411713
******************** Execution ended in 00h 01m 39.81s ********************
============================== 3 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.19885 training's macroF1: 0.489688 valid_1's multi_logloss: 1.14254 valid_1's macroF1: 0.412708
Early stopping, best iteration is:
[136] training's multi_logloss: 1.31544 training's macroF1: 0.461815 valid_1's multi_logloss: 1.28364 valid_1's macroF1: 0.422691
******************** Execution ended in 00h 00m 32.58s ********************
============================== 4 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.19172 training's macroF1: 0.497777 valid_1's multi_logloss: 1.17207 valid_1's macroF1: 0.382614
Early stopping, best iteration is:
[38] training's multi_logloss: 1.36292 training's macroF1: 0.454936 valid_1's multi_logloss: 1.3571 valid_1's macroF1: 0.403982
******************** Execution ended in 00h 00m 27.01s ********************
============================== 5 of 5 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.18896 training's macroF1: 0.498736 valid_1's multi_logloss: 1.18409 valid_1's macroF1: 0.353528
Early stopping, best iteration is:
[147] training's multi_logloss: 1.30661 training's macroF1: 0.47536 valid_1's multi_logloss: 1.29646 valid_1's macroF1: 0.365157
******************** Execution ended in 00h 00m 32.80s ********************
In [101]:
shop_sorted_df = total_shap_df.groupby('feature').mean().sort_values('shap_values', ascending = False).reset_index()
feat_imp_sorted_df = total_shap_df.groupby('feature').mean().sort_values('feat_imp', ascending=False).reset_index()
features_top_shap = shop_sorted_df['feature'][:500] # 500개 컬럼만 추출
features_top_feat_imp = feat_imp_sorted_df['feature'][:500] # 500개 컬럼만 추출
top_features = pd.Series(features_top_shap.tolist() + features_top_feat_imp.tolist())
top_features = top_features.unique() # features_top_shap과 features_top_feat_imp 겹치는 것은 삭제하고 유일한 값만 둔다.
4. Model Development
In [104]:
new_train = train[top_features].copy()
new_test = test[top_features].copy()
In [106]:
print('new train shape:', new_train.shape, 'new test shape:', new_test.shape)
new train shape: (2973, 532) new test shape: (23856, 532)
LGB를 통한 예측 및 변수 중요도 생성
In [107]:
# 카테고리 변수 리스트 생성
new_categorical_feats = [col for col in top_features if col in categorical_feats]
In [110]:
# LGB : Light Gradient Boost 함수 생성
def LGB_OOF(params, categorical_feats, N_FOLDs, SEED=1989):
clf = lgb.LGBMClassifier(objective='multiclass',
random_state=1989,
max_depth=params['max_depth'],
learning_rate=params['learning_rate'],
silent=True,
metric='multi_logloss',
n_jobs=-1, n_estimators=10000,
class_weight='balanced',
colsample_bytree = params['colsample_bytree'],
min_split_gain= params['min_split_gain'],
bagging_freq = params['bagging_freq'],
min_child_weight=params['min_child_weight'],
num_leaves = params['num_leaves'],
subsample = params['subsample'],
reg_alpha= params['reg_alpha'],
reg_lambda= params['reg_lambda'],
num_class=len(np.unique(y)),
bagging_seed=SEED,
seed=SEED,
)
kfold = 10
kf = StratifiedKFold(n_splits=kfold, shuffle=True)
feat_importance_df = pd.DataFrame()
predicts_result = []
for i, (train_index, test_index) in enumerate(kf.split(new_train, y)):
print('='*30, '{} of {} folds'.format(i+1, kfold), '='*30)
start = time.time()
X_train, X_val = new_train.iloc[train_index], new_train.iloc[test_index]
y_train, y_val = y.iloc[train_index], y.iloc[test_index]
clf.fit(X_train, y_train, eval_set=[(X_train, y_train), (X_val, y_val)], eval_metric=evaluate_macroF1_lgb,categorical_feature=new_categorical_feats,
early_stopping_rounds=500, verbose=500)
shap_values = shap.TreeExplainer(clf.booster_).shap_values(X_train)
fold_importance_df = pd.DataFrame()
fold_importance_df['feature'] = X_train.columns
fold_importance_df['shap_values'] = abs(np.array(shap_values)[:, :].mean(1).mean(0))
fold_importance_df['feat_imp'] = clf.feature_importances_
feat_importance_df = pd.concat([feat_importance_df, fold_importance_df])
predicts_result.append(clf.predict(new_test))
print_execution_time(start)
return predicts_result, feat_importance_df
In [111]:
# 셋팅값 입력
params = {'max_depth': 6,
'learning_rate': 0.002,
'colsample_bytree': 0.8,
'subsample': 0.8,
'min_split_gain': 0.02,
'num_leaves': 48,
'reg_alpha': 0.04,
'reg_lambda': 0.073,
'bagging_freq': 2,
'min_child_weight': 40
}
N_Folds = 20
SEED = 1989
predicts_result, feat_importance_df = LGB_OOF(params, new_categorical_feats, N_Folds, SEED=1989)
============================== 1 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.16249 training's macroF1: 0.569094 valid_1's multi_logloss: 1.18146 valid_1's macroF1: 0.42358
Early stopping, best iteration is:
[164] training's multi_logloss: 1.29343 training's macroF1: 0.537105 valid_1's multi_logloss: 1.29374 valid_1's macroF1: 0.439404
******************** Execution ended in 00h 00m 19.45s ********************
============================== 2 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15839 training's macroF1: 0.575707 valid_1's multi_logloss: 1.1554 valid_1's macroF1: 0.394559
Early stopping, best iteration is:
[148] training's multi_logloss: 1.29955 training's macroF1: 0.544585 valid_1's multi_logloss: 1.28984 valid_1's macroF1: 0.399304
******************** Execution ended in 00h 00m 18.94s ********************
============================== 3 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.1647 training's macroF1: 0.576162 valid_1's multi_logloss: 1.15358 valid_1's macroF1: 0.426385
Early stopping, best iteration is:
[29] training's multi_logloss: 1.36818 training's macroF1: 0.524073 valid_1's multi_logloss: 1.36502 valid_1's macroF1: 0.448364
******************** Execution ended in 00h 00m 13.28s ********************
============================== 4 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15892 training's macroF1: 0.567091 valid_1's multi_logloss: 1.15141 valid_1's macroF1: 0.407524
Early stopping, best iteration is:
[32] training's multi_logloss: 1.36582 training's macroF1: 0.538317 valid_1's multi_logloss: 1.36181 valid_1's macroF1: 0.416631
******************** Execution ended in 00h 00m 13.53s ********************
============================== 5 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.16022 training's macroF1: 0.569641 valid_1's multi_logloss: 1.17066 valid_1's macroF1: 0.38358
Early stopping, best iteration is:
[4] training's multi_logloss: 1.38358 training's macroF1: 0.487092 valid_1's multi_logloss: 1.38315 valid_1's macroF1: 0.409667
******************** Execution ended in 00h 00m 12.30s ********************
============================== 6 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15697 training's macroF1: 0.571246 valid_1's multi_logloss: 1.18832 valid_1's macroF1: 0.384436
Early stopping, best iteration is:
[118] training's multi_logloss: 1.31517 training's macroF1: 0.542032 valid_1's multi_logloss: 1.31653 valid_1's macroF1: 0.405468
******************** Execution ended in 00h 00m 17.61s ********************
============================== 7 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.16527 training's macroF1: 0.56447 valid_1's multi_logloss: 1.15593 valid_1's macroF1: 0.438815
[1000] training's multi_logloss: 1.04307 training's macroF1: 0.59581 valid_1's multi_logloss: 1.06816 valid_1's macroF1: 0.467058
[1500] training's multi_logloss: 0.962391 training's macroF1: 0.620582 valid_1's multi_logloss: 1.03027 valid_1's macroF1: 0.464022
Early stopping, best iteration is:
[1040] training's multi_logloss: 1.03559 training's macroF1: 0.598135 valid_1's multi_logloss: 1.06401 valid_1's macroF1: 0.471391
******************** Execution ended in 00h 00m 57.97s ********************
============================== 8 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.16038 training's macroF1: 0.563742 valid_1's multi_logloss: 1.16608 valid_1's macroF1: 0.386054
Early stopping, best iteration is:
[4] training's multi_logloss: 1.38361 training's macroF1: 0.489458 valid_1's multi_logloss: 1.38324 valid_1's macroF1: 0.414727
******************** Execution ended in 00h 00m 12.39s ********************
============================== 9 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.15672 training's macroF1: 0.564723 valid_1's multi_logloss: 1.16849 valid_1's macroF1: 0.409507
Early stopping, best iteration is:
[444] training's multi_logloss: 1.17521 training's macroF1: 0.561 valid_1's multi_logloss: 1.1833 valid_1's macroF1: 0.418126
******************** Execution ended in 00h 00m 32.01s ********************
============================== 10 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 1.159 training's macroF1: 0.56456 valid_1's multi_logloss: 1.16643 valid_1's macroF1: 0.38565
Early stopping, best iteration is:
[367] training's multi_logloss: 1.20482 training's macroF1: 0.555876 valid_1's multi_logloss: 1.20274 valid_1's macroF1: 0.396325
******************** Execution ended in 00h 00m 27.79s ********************
In [112]:
# 변수중요도를 그래프로 표현
fig, ax = plt.subplots(1, 2, figsize=(20, 20))
feat_importance_df_shap = feat_importance_df.groupby('feature').mean().sort_values('shap_values', ascending=False).reset_index()
num_features = 50
sns.barplot(x=feat_importance_df_shap.shap_values[:num_features], y=feat_importance_df_shap.feature[:num_features], ax=ax[0])
ax[0].set_title('Feature importance based on shap values')
feat_importance_df = feat_importance_df.groupby('feature').mean().sort_values('feat_imp', ascending=False).reset_index()
num_features = 50
sns.barplot(x=feat_importance_df.shap_values[:num_features], y=feat_importance_df.feature[:num_features], ax=ax[1])
ax[1].set_title('Feature importance based on feature importance from lgbm')
plt.show()
In [113]:
# 예측 결과 cvs 파일로 생성
submission['Target'] = np.array(predicts_result).mean(axis=0).round().astype(int)
submission.to_csv('submission_with_new_feature_set.csv', index=False)
랜덤하게 찾기 (Randomized Search)
In [114]:
optimized_param = None
lowest_cv = 1000
total_iteration = 100
for i in range(total_iteration):
print('-'*20, 'For {} of {} iterations'.format(i+1, total_iteration), '-'*20)
learning_rate = np.random.rand() * 0.02
n_folds = 3
num_class = len(np.unique(y))
# parameter value들을 일정 범위에 맞게 random하게 설정
params = {}
params['application'] = 'multiclass'
params['metric'] = 'multi_logloss'
params['num_class'] = num_class
params['class_weight'] = 'balanced'
params['num_leaves'] = np.random.randint(24, 48)
params['max_depth'] = np.random.randint(5, 8)
params['min_child_weight'] = np.random.randint(5, 50)
params['min_split_gain'] = np.random.rand() * 0.09
params['colsample_bytree'] = np.random.rand() * (0.9 - 0.1) + 0.1
params['subsample'] = np.random.rand() * (1 - 0.8) + 0.8
params['bagging_freq'] = np.random.randint(1, 5)
params['bagging_seed'] = np.random.randint(1, 5)
params['reg_alpha'] = np.random.rand() * 2
params['reg_lambda'] = np.random.rand() * 2
params['learning_rate'] = np.random.rand() * 0.02
params['seed'] =1989
d_train = lgb.Dataset(data=new_train, label=y.values-1, categorical_feature=new_categorical_feats, free_raw_data=False)
cv_results = lgb.cv(params=params, train_set=d_train, num_boost_round=10000, categorical_feature=new_categorical_feats,
nfold=n_folds, stratified=True, shuffle=True, early_stopping_rounds=1, verbose_eval=1000)
min_cv_results = min(cv_results['multi_logloss-mean'])
# 가장 작은 평균 값의 parmas을 최적화 param으로 등록
if min_cv_results < lowest_cv:
lowest_cv = min_cv_results
optimized_param = params
-------------------- For 1 of 100 iterations --------------------
-------------------- For 2 of 100 iterations --------------------
-------------------- For 3 of 100 iterations --------------------
-------------------- For 4 of 100 iterations --------------------
-------------------- For 5 of 100 iterations --------------------
-------------------- For 6 of 100 iterations --------------------
-------------------- For 7 of 100 iterations --------------------
-------------------- For 8 of 100 iterations --------------------
-------------------- For 9 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.850644 + 0.0107643
-------------------- For 10 of 100 iterations --------------------
-------------------- For 11 of 100 iterations --------------------
-------------------- For 12 of 100 iterations --------------------
-------------------- For 13 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.844449 + 0.0113269
-------------------- For 14 of 100 iterations --------------------
-------------------- For 15 of 100 iterations --------------------
-------------------- For 16 of 100 iterations --------------------
-------------------- For 17 of 100 iterations --------------------
-------------------- For 18 of 100 iterations --------------------
-------------------- For 19 of 100 iterations --------------------
-------------------- For 20 of 100 iterations --------------------
-------------------- For 21 of 100 iterations --------------------
-------------------- For 22 of 100 iterations --------------------
-------------------- For 23 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.874355 + 0.00846433
[2000] cv_agg's multi_logloss: 0.838224 + 0.0124167
-------------------- For 24 of 100 iterations --------------------
-------------------- For 25 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.840324 + 0.0121589
-------------------- For 26 of 100 iterations --------------------
-------------------- For 27 of 100 iterations --------------------
-------------------- For 28 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.96267 + 0.0018782
[2000] cv_agg's multi_logloss: 0.930341 + 0.00400037
[3000] cv_agg's multi_logloss: 0.906175 + 0.00582314
[4000] cv_agg's multi_logloss: 0.88784 + 0.00736017
[5000] cv_agg's multi_logloss: 0.873672 + 0.0085177
[6000] cv_agg's multi_logloss: 0.862664 + 0.00952161
[7000] cv_agg's multi_logloss: 0.853967 + 0.0103318
[8000] cv_agg's multi_logloss: 0.847137 + 0.0110746
[9000] cv_agg's multi_logloss: 0.841789 + 0.0118185
[10000] cv_agg's multi_logloss: 0.837651 + 0.0124768
-------------------- For 29 of 100 iterations --------------------
-------------------- For 30 of 100 iterations --------------------
-------------------- For 31 of 100 iterations --------------------
-------------------- For 32 of 100 iterations --------------------
-------------------- For 33 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.845318 + 0.0122234
-------------------- For 34 of 100 iterations --------------------
-------------------- For 35 of 100 iterations --------------------
-------------------- For 36 of 100 iterations --------------------
-------------------- For 37 of 100 iterations --------------------
-------------------- For 38 of 100 iterations --------------------
-------------------- For 39 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.844441 + 0.0112841
-------------------- For 40 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.840559 + 0.0119178
-------------------- For 41 of 100 iterations --------------------
-------------------- For 42 of 100 iterations --------------------
-------------------- For 43 of 100 iterations --------------------
-------------------- For 44 of 100 iterations --------------------
-------------------- For 45 of 100 iterations --------------------
-------------------- For 46 of 100 iterations --------------------
-------------------- For 47 of 100 iterations --------------------
-------------------- For 48 of 100 iterations --------------------
-------------------- For 49 of 100 iterations --------------------
-------------------- For 50 of 100 iterations --------------------
-------------------- For 51 of 100 iterations --------------------
-------------------- For 52 of 100 iterations --------------------
-------------------- For 53 of 100 iterations --------------------
-------------------- For 54 of 100 iterations --------------------
-------------------- For 55 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.933994 + 0.00295004
[2000] cv_agg's multi_logloss: 0.891992 + 0.00616865
[3000] cv_agg's multi_logloss: 0.866375 + 0.0085591
[4000] cv_agg's multi_logloss: 0.850178 + 0.0105231
[5000] cv_agg's multi_logloss: 0.840013 + 0.0122332
[6000] cv_agg's multi_logloss: 0.83367 + 0.0137947
-------------------- For 56 of 100 iterations --------------------
-------------------- For 57 of 100 iterations --------------------
-------------------- For 58 of 100 iterations --------------------
-------------------- For 59 of 100 iterations --------------------
-------------------- For 60 of 100 iterations --------------------
-------------------- For 61 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.959375 + 0.00139668
[2000] cv_agg's multi_logloss: 0.925476 + 0.00324838
[3000] cv_agg's multi_logloss: 0.90067 + 0.00512656
[4000] cv_agg's multi_logloss: 0.882047 + 0.00678869
[5000] cv_agg's multi_logloss: 0.867985 + 0.00814784
[6000] cv_agg's multi_logloss: 0.857164 + 0.0093546
[7000] cv_agg's multi_logloss: 0.848876 + 0.0103916
[8000] cv_agg's multi_logloss: 0.842528 + 0.0113579
[9000] cv_agg's multi_logloss: 0.837696 + 0.0122748
[10000] cv_agg's multi_logloss: 0.834037 + 0.0132079
-------------------- For 62 of 100 iterations --------------------
-------------------- For 63 of 100 iterations --------------------
-------------------- For 64 of 100 iterations --------------------
-------------------- For 65 of 100 iterations --------------------
-------------------- For 66 of 100 iterations --------------------
-------------------- For 67 of 100 iterations --------------------
-------------------- For 68 of 100 iterations --------------------
-------------------- For 69 of 100 iterations --------------------
-------------------- For 70 of 100 iterations --------------------
-------------------- For 71 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.829763 + 0.0154857
-------------------- For 72 of 100 iterations --------------------
-------------------- For 73 of 100 iterations --------------------
-------------------- For 74 of 100 iterations --------------------
-------------------- For 75 of 100 iterations --------------------
-------------------- For 76 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.880608 + 0.00730383
[2000] cv_agg's multi_logloss: 0.843414 + 0.0122766
-------------------- For 77 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.953842 + 0.00181947
[2000] cv_agg's multi_logloss: 0.917862 + 0.00409414
[3000] cv_agg's multi_logloss: 0.892608 + 0.0061033
[4000] cv_agg's multi_logloss: 0.874444 + 0.00772182
[5000] cv_agg's multi_logloss: 0.861021 + 0.00908268
[6000] cv_agg's multi_logloss: 0.851073 + 0.0102535
[7000] cv_agg's multi_logloss: 0.843687 + 0.011348
[8000] cv_agg's multi_logloss: 0.838232 + 0.0123942
[9000] cv_agg's multi_logloss: 0.83417 + 0.0133474
[10000] cv_agg's multi_logloss: 0.831216 + 0.0142597
-------------------- For 78 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.835838 + 0.013082
-------------------- For 79 of 100 iterations --------------------
-------------------- For 80 of 100 iterations --------------------
-------------------- For 81 of 100 iterations --------------------
-------------------- For 82 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.837136 + 0.0125166
-------------------- For 83 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.882841 + 0.00739006
[2000] cv_agg's multi_logloss: 0.843938 + 0.0117782
[3000] cv_agg's multi_logloss: 0.830597 + 0.0151525
-------------------- For 84 of 100 iterations --------------------
-------------------- For 85 of 100 iterations --------------------
-------------------- For 86 of 100 iterations --------------------
-------------------- For 87 of 100 iterations --------------------
-------------------- For 88 of 100 iterations --------------------
-------------------- For 89 of 100 iterations --------------------
-------------------- For 90 of 100 iterations --------------------
-------------------- For 91 of 100 iterations --------------------
-------------------- For 92 of 100 iterations --------------------
-------------------- For 93 of 100 iterations --------------------
-------------------- For 94 of 100 iterations --------------------
-------------------- For 95 of 100 iterations --------------------
-------------------- For 96 of 100 iterations --------------------
-------------------- For 97 of 100 iterations --------------------
[1000] cv_agg's multi_logloss: 0.833212 + 0.0148136
-------------------- For 98 of 100 iterations --------------------
-------------------- For 99 of 100 iterations --------------------
-------------------- For 100 of 100 iterations --------------------
In [115]:
N_Folds = 20
SEED = 1989
# 랜덤으로 추출한 변수를 LGB 함수에 적용
predicts_result, feat_importance_df = LGB_OOF(optimized_param, new_categorical_feats, N_Folds, SEED=1989)
============================== 1 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.643909 training's macroF1: 0.767897 valid_1's multi_logloss: 1.01731 valid_1's macroF1: 0.419311
Early stopping, best iteration is:
[455] training's multi_logloss: 0.66926 training's macroF1: 0.756981 valid_1's multi_logloss: 1.02123 valid_1's macroF1: 0.424335
******************** Execution ended in 00h 00m 31.75s ********************
============================== 2 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.642988 training's macroF1: 0.772115 valid_1's multi_logloss: 0.973459 valid_1's macroF1: 0.441129
Early stopping, best iteration is:
[246] training's multi_logloss: 0.836124 training's macroF1: 0.70146 valid_1's multi_logloss: 1.00646 valid_1's macroF1: 0.460188
******************** Execution ended in 00h 00m 22.09s ********************
============================== 3 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.646518 training's macroF1: 0.764841 valid_1's multi_logloss: 0.90879 valid_1's macroF1: 0.425241
Early stopping, best iteration is:
[301] training's multi_logloss: 0.785451 training's macroF1: 0.71514 valid_1's multi_logloss: 0.941626 valid_1's macroF1: 0.43586
******************** Execution ended in 00h 00m 24.43s ********************
============================== 4 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.642282 training's macroF1: 0.776681 valid_1's multi_logloss: 1.00217 valid_1's macroF1: 0.454198
[1000] training's multi_logloss: 0.448314 training's macroF1: 0.847268 valid_1's multi_logloss: 0.976682 valid_1's macroF1: 0.466582
Early stopping, best iteration is:
[722] training's multi_logloss: 0.536986 training's macroF1: 0.814864 valid_1's multi_logloss: 0.985533 valid_1's macroF1: 0.475647
******************** Execution ended in 00h 00m 42.88s ********************
============================== 5 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.642159 training's macroF1: 0.772177 valid_1's multi_logloss: 0.95837 valid_1's macroF1: 0.421447
Early stopping, best iteration is:
[30] training's multi_logloss: 1.24539 training's macroF1: 0.586624 valid_1's multi_logloss: 1.24092 valid_1's macroF1: 0.438978
******************** Execution ended in 00h 00m 11.48s ********************
============================== 6 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.642027 training's macroF1: 0.767923 valid_1's multi_logloss: 1.06385 valid_1's macroF1: 0.415905
Early stopping, best iteration is:
[355] training's multi_logloss: 0.734352 training's macroF1: 0.730868 valid_1's multi_logloss: 1.06958 valid_1's macroF1: 0.419991
******************** Execution ended in 00h 00m 27.10s ********************
============================== 7 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.65071 training's macroF1: 0.769133 valid_1's multi_logloss: 0.944704 valid_1's macroF1: 0.407626
[1000] training's multi_logloss: 0.454591 training's macroF1: 0.841963 valid_1's multi_logloss: 0.921269 valid_1's macroF1: 0.411259
Early stopping, best iteration is:
[820] training's multi_logloss: 0.509099 training's macroF1: 0.827846 valid_1's multi_logloss: 0.925363 valid_1's macroF1: 0.427033
******************** Execution ended in 00h 00m 47.39s ********************
============================== 8 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.641123 training's macroF1: 0.773204 valid_1's multi_logloss: 0.970383 valid_1's macroF1: 0.396786
[1000] training's multi_logloss: 0.449023 training's macroF1: 0.853571 valid_1's multi_logloss: 0.94899 valid_1's macroF1: 0.398541
Early stopping, best iteration is:
[968] training's multi_logloss: 0.457699 training's macroF1: 0.847806 valid_1's multi_logloss: 0.949025 valid_1's macroF1: 0.414443
******************** Execution ended in 00h 00m 51.53s ********************
============================== 9 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.63771 training's macroF1: 0.773218 valid_1's multi_logloss: 0.987514 valid_1's macroF1: 0.439933
Early stopping, best iteration is:
[115] training's multi_logloss: 1.01131 training's macroF1: 0.6531 valid_1's multi_logloss: 1.08709 valid_1's macroF1: 0.469625
******************** Execution ended in 00h 00m 15.76s ********************
============================== 10 of 10 folds ==============================
Training until validation scores don't improve for 500 rounds.
[500] training's multi_logloss: 0.640941 training's macroF1: 0.777775 valid_1's multi_logloss: 1.03424 valid_1's macroF1: 0.404788
[1000] training's multi_logloss: 0.445307 training's macroF1: 0.848409 valid_1's multi_logloss: 1.0132 valid_1's macroF1: 0.420932
[1500] training's multi_logloss: 0.348561 training's macroF1: 0.882879 valid_1's multi_logloss: 1.01566 valid_1's macroF1: 0.411427
Early stopping, best iteration is:
[1068] training's multi_logloss: 0.4287 training's macroF1: 0.853714 valid_1's multi_logloss: 1.01169 valid_1's macroF1: 0.42216
******************** Execution ended in 00h 00m 56.61s ********************
In [ ]:
submission['Target'] = np.array(predicts_result).mean(axis=0).round().astype(int)
submission.to_csv('submission_shap_randomized_search.csv', index = False)
'Competition > Kaggle' 카테고리의 다른 글
| [kaggle][필사] Statoil/C-CORE Iceberg Classifier Challenge (0) | 2020.09.25 |
|---|---|
| [kaggle][필사] Costa Rican Household Proverty (2) (0) | 2020.09.20 |
| [kaggle][필사] Costa Rican Household Proverty (1) (0) | 2020.09.19 |
'Competition/Kaggle'의 다른글
- 현재글[kaggle][필사] Costa Rican Household Proverty (3)