W3Cschool
恭喜您成為首批注冊用戶
獲得88經(jīng)驗(yàn)值獎(jiǎng)勵(lì)
EarlyStopping
繼承自: Callback
定義在:tensorflow/python/keras/callbacks.py。
當(dāng)監(jiān)測數(shù)量停止改善時(shí)停止訓(xùn)練。
參數(shù):
__init__
__init__(
monitor='val_loss',
min_delta=0,
patience=0,
verbose=0,
mode='auto',
baseline=None,
restore_best_weights=False
)
初始化自我。
get_monitor_value
get_monitor_value(logs)
on_batch_begin
on_batch_begin(
batch,
logs=None
)
on_batch_end
on_batch_end(
batch,
logs=None
)
on_epoch_begin
on_epoch_begin(
epoch,
logs=None
)
on_epoch_end
on_epoch_end(
epoch,
logs=None
)
on_train_batch_begin
on_train_batch_begin(
batch,
logs=None
)
on_train_batch_end
on_train_batch_end(
batch,
logs=None
)
on_train_begin
on_train_begin(logs=None)
on_train_end
on_train_end(logs=None)
set_model
set_model(model)
set_params
set_params(params)
Copyright©2021 w3cschool編程獅|閩ICP備15016281號-3|閩公網(wǎng)安備35020302033924號
違法和不良信息舉報(bào)電話:173-0602-2364|舉報(bào)郵箱:jubao@eeedong.com
掃描二維碼
下載編程獅App
編程獅公眾號
聯(lián)系方式:
更多建議: