ClassCat® AI Research

Keras Core 0.1 : 開発者ガイド : 独自のコールバックを書く

Posted on by Sales Information

Keras Core 0.1 : 開発者ガイド : 独自のコールバックを書く (翻訳/解説)

翻訳 : (株)クラスキャット セールスインフォメーション
作成日時 : 07/17/2023 (0.1.0)

* 本ページは、Keras Core の以下のドキュメントを翻訳した上で適宜、補足説明したものです：

  • Writing your own callbacks
    (Author: Rick Chao, Francois Chollet ; Date created: 2019/03/20 ; Last modified: 2023/06/25)
    Description: Complete guide to writing new Keras callbacks.

* サンプルコードの動作確認はしておりますが、必要な場合には適宜、追加改変しています。
* ご自由にリンクを張って頂いてかまいませんが、sales-info@classcat.com までご一報いただけると嬉しいです。

 

クラスキャット 人工知能 研究開発支援サービス

◆ クラスキャット は人工知能・テレワークに関する各種サービスを提供しています。お気軽にご相談ください :

◆ 人工知能とビジネスをテーマに WEB セミナーを定期的に開催しています。スケジュール。
  • お住まいの地域に関係なく Web ブラウザからご参加頂けます。事前登録 が必要ですのでご注意ください。

◆ お問合せ : 本件に関するお問い合わせ先は下記までお願いいたします。

  • 株式会社クラスキャット セールス・マーケティング本部 セールス・インフォメーション
  • sales-info@classcat.com  ;  Web: www.classcat.com  ;   ClassCatJP

 

Keras Core 0.1 : 開発者ガイド : 独自のコールバックを書く

イントロダクション

コールバックは 訓練、評価や推論の間の Keras モデルの動作をカスタマイズするための強力なツールです。サンプルは訓練進捗と結果を TensorBoard で可視化できる keras.callbacks.TensorBoard や、訓練の間にモデルを定期的にセーブする keras.callbacks.ModelCheckpoint を含みます。

このガイドでは、Keras コールバックが何か、それは何ができるか、そして独自のものをどのように構築できるかを学習します。始めるために単純なコールバック・アプリケーションの幾つかのデモを提供します。

 

セットアップ

import numpy as np
import keras_core as keras

Using TensorFlow backend

 

Keras コールバック概要

総てのコールバックは keras.callbacks.Callback クラスをサブクラス化して、そして訓練、テストと予測の様々なステージで呼び出されるメソッドのセットを override します。コールバックは訓練の間のモデルの内部状態と統計上のビューを得るために有用です。

コールバックのリストを (キーワード引数 callbacks として) 以下のモデル・メソッドに渡すことができます :

  • keras.Model.fit()
  • keras.Model.evaluate()
  • keras.Model.predict()

 

コールバック・メソッドの概要

グローバル・メソッド

on_(train|test|predict)_begin(self, logs=None)

fit/evaluate/predict の最初に呼び出されます。

on_(train|test|predict)_end(self, logs=None)

fit/evaluate/predict の最後に呼び出されます。

 

訓練/テスト/予測のためのバッチレベル・メソッド

on_(train|test|predict)_batch_begin(self, batch, logs=None)

訓練/テスト/予測の間のバッチを処理するすぐ前に呼び出されます。

on_(train|test|predict)_batch_end(self, batch, logs=None)

バッチを訓練/テスト/予測する最後に呼び出されます。このメソッド内では、logs は metrics 結果を含む辞書です。

 

エポックレベル・メソッド (訓練 only)

on_epoch_begin(self, epoch, logs=None)

訓練の間にエポックの最初に呼び出されます。

on_epoch_end(self, epoch, logs=None)

訓練の間にエポックの最後に呼び出されます。

 

基本的なサンプル

具体的なサンプルを見ましょう。始めるために、tensorflow をインポートして単純な Sequential Keras モデルを定義しましょう :

# Define the Keras model to add callbacks to
def get_model():
    model = keras.Sequential()
    model.add(keras.layers.Dense(1))
    model.compile(
        optimizer=keras.optimizers.RMSprop(learning_rate=0.1),
        loss="mean_squared_error",
        metrics=["mean_absolute_error"],
    )
    return model

それから訓練とテストのために Keras datasets API から MNIST データをロードします :

# Load example MNIST data and pre-process it
(x_train, y_train), (x_test, y_test) = keras.datasets.mnist.load_data()
x_train = x_train.reshape(-1, 784).astype("float32") / 255.0
x_test = x_test.reshape(-1, 784).astype("float32") / 255.0

# Limit the data to 1000 samples
x_train = x_train[:1000]
y_train = y_train[:1000]
x_test = x_test[:1000]
y_test = y_test[:1000]

次に、以下をログ出力する単純なカスタム・コールバックを定義します :

  • fit/evaluate/predict が開始 & 終了する時
  • 各エポックが開始 & 終了する時
  • 各訓練バッチが開始 & 終了する時
  • 各評価 (テスト) バッチが開始 & 終了する時
  • 各推論 (予測) バッチが開始 & 終了する時 
class CustomCallback(keras.callbacks.Callback):
    def on_train_begin(self, logs=None):
        keys = list(logs.keys())
        print("Starting training; got log keys: {}".format(keys))

    def on_train_end(self, logs=None):
        keys = list(logs.keys())
        print("Stop training; got log keys: {}".format(keys))

    def on_epoch_begin(self, epoch, logs=None):
        keys = list(logs.keys())
        print("Start epoch {} of training; got log keys: {}".format(epoch, keys))

    def on_epoch_end(self, epoch, logs=None):
        keys = list(logs.keys())
        print("End epoch {} of training; got log keys: {}".format(epoch, keys))

    def on_test_begin(self, logs=None):
        keys = list(logs.keys())
        print("Start testing; got log keys: {}".format(keys))

    def on_test_end(self, logs=None):
        keys = list(logs.keys())
        print("Stop testing; got log keys: {}".format(keys))

    def on_predict_begin(self, logs=None):
        keys = list(logs.keys())
        print("Start predicting; got log keys: {}".format(keys))

    def on_predict_end(self, logs=None):
        keys = list(logs.keys())
        print("Stop predicting; got log keys: {}".format(keys))

    def on_train_batch_begin(self, batch, logs=None):
        keys = list(logs.keys())
        print("...Training: start of batch {}; got log keys: {}".format(batch, keys))

    def on_train_batch_end(self, batch, logs=None):
        keys = list(logs.keys())
        print("...Training: end of batch {}; got log keys: {}".format(batch, keys))

    def on_test_batch_begin(self, batch, logs=None):
        keys = list(logs.keys())
        print("...Evaluating: start of batch {}; got log keys: {}".format(batch, keys))

    def on_test_batch_end(self, batch, logs=None):
        keys = list(logs.keys())
        print("...Evaluating: end of batch {}; got log keys: {}".format(batch, keys))

    def on_predict_batch_begin(self, batch, logs=None):
        keys = list(logs.keys())
        print("...Predicting: start of batch {}; got log keys: {}".format(batch, keys))

    def on_predict_batch_end(self, batch, logs=None):
        keys = list(logs.keys())
        print("...Predicting: end of batch {}; got log keys: {}".format(batch, keys))

Let’s try it out:

model = get_model()
model.fit(
    x_train,
    y_train,
    batch_size=128,
    epochs=1,
    verbose=0,
    validation_split=0.5,
    callbacks=[CustomCallback()],
)

res = model.evaluate(
    x_test, y_test, batch_size=128, verbose=0, callbacks=[CustomCallback()]
)

res = model.predict(x_test, batch_size=128, callbacks=[CustomCallback()])

Starting training; got log keys: []
Start epoch 0 of training; got log keys: []
...Training: start of batch 0; got log keys: []
...Training: end of batch 0; got log keys: ['loss', 'mean_absolute_error']
...Training: start of batch 1; got log keys: []
...Training: end of batch 1; got log keys: ['loss', 'mean_absolute_error']
...Training: start of batch 2; got log keys: []
...Training: end of batch 2; got log keys: ['loss', 'mean_absolute_error']
...Training: start of batch 3; got log keys: []
...Training: end of batch 3; got log keys: ['loss', 'mean_absolute_error']
Start testing; got log keys: []
...Evaluating: start of batch 0; got log keys: []
...Evaluating: end of batch 0; got log keys: ['loss', 'mean_absolute_error']
...Evaluating: start of batch 1; got log keys: []
...Evaluating: end of batch 1; got log keys: ['loss', 'mean_absolute_error']
...Evaluating: start of batch 2; got log keys: []
...Evaluating: end of batch 2; got log keys: ['loss', 'mean_absolute_error']
...Evaluating: start of batch 3; got log keys: []
...Evaluating: end of batch 3; got log keys: ['loss', 'mean_absolute_error']
Stop testing; got log keys: ['loss', 'mean_absolute_error']
End epoch 0 of training; got log keys: ['loss', 'mean_absolute_error', 'val_loss', 'val_mean_absolute_error']
Stop training; got log keys: ['loss', 'mean_absolute_error', 'val_loss', 'val_mean_absolute_error']
Start testing; got log keys: []
...Evaluating: start of batch 0; got log keys: []
...Evaluating: end of batch 0; got log keys: ['loss', 'mean_absolute_error']
...Evaluating: start of batch 1; got log keys: []
...Evaluating: end of batch 1; got log keys: ['loss', 'mean_absolute_error']
...Evaluating: start of batch 2; got log keys: []
...Evaluating: end of batch 2; got log keys: ['loss', 'mean_absolute_error']
...Evaluating: start of batch 3; got log keys: []
...Evaluating: end of batch 3; got log keys: ['loss', 'mean_absolute_error']
...Evaluating: start of batch 4; got log keys: []
...Evaluating: end of batch 4; got log keys: ['loss', 'mean_absolute_error']
...Evaluating: start of batch 5; got log keys: []
...Evaluating: end of batch 5; got log keys: ['loss', 'mean_absolute_error']
...Evaluating: start of batch 6; got log keys: []
...Evaluating: end of batch 6; got log keys: ['loss', 'mean_absolute_error']
...Evaluating: start of batch 7; got log keys: []
...Evaluating: end of batch 7; got log keys: ['loss', 'mean_absolute_error']
Stop testing; got log keys: ['loss', 'mean_absolute_error']
Start predicting; got log keys: []
...Predicting: start of batch 0; got log keys: []
...Predicting: end of batch 0; got log keys: ['outputs']
 1/8 ━━━━━━━━━━━━━━━━━━━━  0s 36ms/step...Predicting: start of batch 1; got log keys: []
...Predicting: end of batch 1; got log keys: ['outputs']
...Predicting: start of batch 2; got log keys: []
...Predicting: end of batch 2; got log keys: ['outputs']
...Predicting: start of batch 3; got log keys: []
...Predicting: end of batch 3; got log keys: ['outputs']
...Predicting: start of batch 4; got log keys: []
...Predicting: end of batch 4; got log keys: ['outputs']
...Predicting: start of batch 5; got log keys: []
...Predicting: end of batch 5; got log keys: ['outputs']
...Predicting: start of batch 6; got log keys: []
...Predicting: end of batch 6; got log keys: ['outputs']
...Predicting: start of batch 7; got log keys: []
...Predicting: end of batch 7; got log keys: ['outputs']
Stop predicting; got log keys: []
 8/8 ━━━━━━━━━━━━━━━━━━━━ 0s 3ms/step

 

logs 辞書の使用方法

logs 辞書はバッチやエポックの最後の損失値と総てのメトリクスを含みます。サンプルは損失と平均絶対誤差 (MAE) を含みます。

class LossAndErrorPrintingCallback(keras.callbacks.Callback):
    def on_train_batch_end(self, batch, logs=None):
        print(
            "Up to batch {}, the average loss is {:7.2f}.".format(batch, logs["loss"])
        )

    def on_test_batch_end(self, batch, logs=None):
        print(
            "Up to batch {}, the average loss is {:7.2f}.".format(batch, logs["loss"])
        )

    def on_epoch_end(self, epoch, logs=None):
        print(
            "The average loss for epoch {} is {:7.2f} "
            "and mean absolute error is {:7.2f}.".format(
                epoch, logs["loss"], logs["mean_absolute_error"]
            )
        )


model = get_model()
model.fit(
    x_train,
    y_train,
    batch_size=128,
    epochs=2,
    verbose=0,
    callbacks=[LossAndErrorPrintingCallback()],
)

res = model.evaluate(
    x_test,
    y_test,
    batch_size=128,
    verbose=0,
    callbacks=[LossAndErrorPrintingCallback()],
)

Up to batch 0, the average loss is   33.48.
Up to batch 1, the average loss is  466.14.
Up to batch 2, the average loss is  318.55.
Up to batch 3, the average loss is  241.21.
Up to batch 4, the average loss is  194.39.
Up to batch 5, the average loss is  163.04.
Up to batch 6, the average loss is  140.47.
Up to batch 7, the average loss is  123.61.
The average loss for epoch 0 is  123.61 and mean absolute error is    5.94.
Up to batch 0, the average loss is    4.82.
Up to batch 1, the average loss is    5.23.
Up to batch 2, the average loss is    4.90.
Up to batch 3, the average loss is    4.81.
Up to batch 4, the average loss is    4.85.
Up to batch 5, the average loss is    4.92.
Up to batch 6, the average loss is    4.86.
Up to batch 7, the average loss is    4.71.
The average loss for epoch 1 is    4.71 and mean absolute error is    1.74.
Up to batch 0, the average loss is    4.84.
Up to batch 1, the average loss is    4.55.
Up to batch 2, the average loss is    4.53.
Up to batch 3, the average loss is    4.46.
Up to batch 4, the average loss is    4.57.
Up to batch 5, the average loss is    4.54.
Up to batch 6, the average loss is    4.46.
Up to batch 7, the average loss is    4.40.

 

self.model 属性の使用方法

これらのメソッドの一つが呼び出されるとき、ログ情報を受け取ることに加えて、コールバックは訓練/評価/推論の現在のラウンドに関連するモデルへのアクセスを持ちます : self.model です。

コールバック内で self.model で行なうことができることの幾つかがあります :

  • 訓練を直ちに中断するために self.model.stop_training = True を設定します。
  • self.model.optimizer.learning_rate のような、(self.model.optimizer として利用可能な) optimizer のハイパーパラメータを変化させます。
  • 定期的な間隔でモデルをセーブします。
  • 訓練の間の正当性チェックとして使用するため、各エポックの最後で幾つかのテストサンプル上 model.predict() の出力を記録します。
  • 時間につれてモデルが何を学習しているか監視するため、各エポックの最後に中間特徴の可視化を抽出します。
  • 等々。

これを 2, 3 のサンプルで実際に見ましょう。

 

Keras コールバック・アプリケーションのサンプル

最小損失における Early Stopping

最初の例はコールバックの作成を示します、これは損失の最小に達したとき、属性 self.model.stop_training (boolean) を設定することにより訓練を停止します。オプションで、局所的な最小に達した後で停止する前に幾つのエポックを待つべきかを指定するための引数 patience を供給できます。

keras.callbacks.EarlyStopping はより完全で一般的な実装を提供します。

class EarlyStoppingAtMinLoss(keras.callbacks.Callback):
    """Stop training when the loss is at its min, i.e. the loss stops decreasing.

    Arguments:
        patience: Number of epochs to wait after min has been hit. After this
        number of no improvement, training stops.
    """

    def __init__(self, patience=0):
        super().__init__()
        self.patience = patience
        # best_weights to store the weights at which the minimum loss occurs.
        self.best_weights = None

    def on_train_begin(self, logs=None):
        # The number of epoch it has waited when loss is no longer minimum.
        self.wait = 0
        # The epoch the training stops at.
        self.stopped_epoch = 0
        # Initialize the best as infinity.
        self.best = np.Inf

    def on_epoch_end(self, epoch, logs=None):
        current = logs.get("loss")
        if np.less(current, self.best):
            self.best = current
            self.wait = 0
            # Record the best weights if current results is better (less).
            self.best_weights = self.model.get_weights()
        else:
            self.wait += 1
            if self.wait >= self.patience:
                self.stopped_epoch = epoch
                self.model.stop_training = True
                print("Restoring model weights from the end of the best epoch.")
                self.model.set_weights(self.best_weights)

    def on_train_end(self, logs=None):
        if self.stopped_epoch > 0:
            print(f"Epoch {self.stopped_epoch + 1}: early stopping")


model = get_model()
model.fit(
    x_train,
    y_train,
    batch_size=64,
    epochs=30,
    verbose=0,
    callbacks=[LossAndErrorPrintingCallback(), EarlyStoppingAtMinLoss()],
)

Up to batch 0, the average loss is   26.59.
Up to batch 1, the average loss is  499.28.
Up to batch 2, the average loss is  344.22.
Up to batch 3, the average loss is  260.05.
Up to batch 4, the average loss is  209.62.
Up to batch 5, the average loss is  175.54.
Up to batch 6, the average loss is  151.26.
Up to batch 7, the average loss is  133.00.
Up to batch 8, the average loss is  118.77.
Up to batch 9, the average loss is  107.52.
Up to batch 10, the average loss is   98.29.
Up to batch 11, the average loss is   90.49.
Up to batch 12, the average loss is   83.88.
Up to batch 13, the average loss is   78.21.
Up to batch 14, the average loss is   73.27.
Up to batch 15, the average loss is   68.84.
The average loss for epoch 0 is   68.84 and mean absolute error is    3.96.
Up to batch 0, the average loss is    3.25.
Up to batch 1, the average loss is    3.54.
Up to batch 2, the average loss is    3.44.
Up to batch 3, the average loss is    3.76.
Up to batch 4, the average loss is    4.03.
Up to batch 5, the average loss is    5.05.
Up to batch 6, the average loss is    7.64.
Up to batch 7, the average loss is   13.50.
Up to batch 8, the average loss is   25.10.
Up to batch 9, the average loss is   35.96.
Up to batch 10, the average loss is   40.17.
Up to batch 11, the average loss is   38.74.
Up to batch 12, the average loss is   36.79.
Up to batch 13, the average loss is   34.97.
Up to batch 14, the average loss is   33.31.
Up to batch 15, the average loss is   31.61.
The average loss for epoch 1 is   31.61 and mean absolute error is    4.14.
Up to batch 0, the average loss is    6.76.
Up to batch 1, the average loss is    6.35.
Up to batch 2, the average loss is    7.28.
Up to batch 3, the average loss is   11.00.
Up to batch 4, the average loss is   15.26.
Up to batch 5, the average loss is   17.59.
Up to batch 6, the average loss is   20.06.
Up to batch 7, the average loss is   21.12.
Up to batch 8, the average loss is   22.65.
Up to batch 9, the average loss is   25.84.
Up to batch 10, the average loss is   28.86.
Up to batch 11, the average loss is   30.41.
Up to batch 12, the average loss is   31.02.
Up to batch 13, the average loss is   31.43.
Up to batch 14, the average loss is   31.11.
Up to batch 15, the average loss is   30.72.
The average loss for epoch 2 is   30.72 and mean absolute error is    4.80.
Up to batch 0, the average loss is   27.88.
Up to batch 1, the average loss is   24.97.
Up to batch 2, the average loss is   21.35.
Up to batch 3, the average loss is   19.31.
Up to batch 4, the average loss is   18.35.
Up to batch 5, the average loss is   18.01.
Up to batch 6, the average loss is   19.03.
Up to batch 7, the average loss is   21.70.
Up to batch 8, the average loss is   25.97.
Up to batch 9, the average loss is   30.15.
Up to batch 10, the average loss is   32.60.
Up to batch 11, the average loss is   34.73.
Up to batch 12, the average loss is   35.49.
Up to batch 13, the average loss is   35.09.
Up to batch 14, the average loss is   34.38.
Up to batch 15, the average loss is   33.64.
The average loss for epoch 3 is   33.64 and mean absolute error is    5.05.
Restoring model weights from the end of the best epoch.
Epoch 4: early stopping

<keras_core.src.callbacks.history.History at 0x16c79ded0>

 

学習率スケジューリング

この例では、カスタム・コールバックが訓練の過程において optimizer の学習率を動的に変更するためにどのように使用できるかを示しています。

より一般的な実装については callbacks.LearningRateScheduler をご覧ください。

class CustomLearningRateScheduler(keras.callbacks.Callback):
    """Learning rate scheduler which sets the learning rate according to schedule.

    Arguments:
        schedule: a function that takes an epoch index
            (integer, indexed from 0) and current learning rate
            as inputs and returns a new learning rate as output (float).
    """

    def __init__(self, schedule):
        super().__init__()
        self.schedule = schedule

    def on_epoch_begin(self, epoch, logs=None):
        if not hasattr(self.model.optimizer, "learning_rate"):
            raise ValueError('Optimizer must have a "learning_rate" attribute.')
        # Get the current learning rate from model's optimizer.
        lr = self.model.optimizer.learning_rate
        # Call schedule function to get the scheduled learning rate.
        scheduled_lr = self.schedule(epoch, lr)
        # Set the value back to the optimizer before this epoch starts
        self.model.optimizer.learning_rate = scheduled_lr
        print(f"\nEpoch {epoch}: Learning rate is {float(np.array(scheduled_lr))}.")


LR_SCHEDULE = [
    # (epoch to start, learning rate) tuples
    (3, 0.05),
    (6, 0.01),
    (9, 0.005),
    (12, 0.001),
]


def lr_schedule(epoch, lr):
    """Helper function to retrieve the scheduled learning rate based on epoch."""
    if epoch < LR_SCHEDULE[0][0] or epoch > LR_SCHEDULE[-1][0]:
        return lr
    for i in range(len(LR_SCHEDULE)):
        if epoch == LR_SCHEDULE[i][0]:
            return LR_SCHEDULE[i][1]
    return lr


model = get_model()
model.fit(
    x_train,
    y_train,
    batch_size=64,
    epochs=15,
    verbose=0,
    callbacks=[
        LossAndErrorPrintingCallback(),
        CustomLearningRateScheduler(lr_schedule),
    ],
)

Epoch 0: Learning rate is 0.10000000149011612.
Up to batch 0, the average loss is   23.52.
Up to batch 1, the average loss is  420.81.
Up to batch 2, the average loss is  288.41.
Up to batch 3, the average loss is  218.74.
Up to batch 4, the average loss is  176.27.
Up to batch 5, the average loss is  148.04.
Up to batch 6, the average loss is  127.75.
Up to batch 7, the average loss is  112.55.
Up to batch 8, the average loss is  100.66.
Up to batch 9, the average loss is   91.02.
Up to batch 10, the average loss is   83.14.
Up to batch 11, the average loss is   76.86.
Up to batch 12, the average loss is   71.88.
Up to batch 13, the average loss is   67.90.
Up to batch 14, the average loss is   65.49.
Up to batch 15, the average loss is   64.70.
The average loss for epoch 0 is   64.70 and mean absolute error is    4.47.

Epoch 1: Learning rate is 0.10000000149011612.
Up to batch 0, the average loss is   38.88.
Up to batch 1, the average loss is   37.12.
Up to batch 2, the average loss is   38.05.
Up to batch 3, the average loss is   42.01.
Up to batch 4, the average loss is   41.40.
Up to batch 5, the average loss is   37.24.
Up to batch 6, the average loss is   33.52.
Up to batch 7, the average loss is   30.67.
Up to batch 8, the average loss is   28.53.
Up to batch 9, the average loss is   27.19.
Up to batch 10, the average loss is   26.16.
Up to batch 11, the average loss is   25.73.
Up to batch 12, the average loss is   25.83.
Up to batch 13, the average loss is   27.42.
Up to batch 14, the average loss is   30.41.
Up to batch 15, the average loss is   33.43.
The average loss for epoch 1 is   33.43 and mean absolute error is    4.92.

Epoch 2: Learning rate is 0.10000000149011612.
Up to batch 0, the average loss is   49.76.
Up to batch 1, the average loss is   34.67.
Up to batch 2, the average loss is   27.66.
Up to batch 3, the average loss is   22.61.
Up to batch 4, the average loss is   19.57.
Up to batch 5, the average loss is   17.36.
Up to batch 6, the average loss is   16.26.
Up to batch 7, the average loss is   16.09.
Up to batch 8, the average loss is   17.32.
Up to batch 9, the average loss is   21.83.
Up to batch 10, the average loss is   26.74.
Up to batch 11, the average loss is   30.42.
Up to batch 12, the average loss is   31.12.
Up to batch 13, the average loss is   30.43.
Up to batch 14, the average loss is   29.68.
Up to batch 15, the average loss is   28.90.
The average loss for epoch 2 is   28.90 and mean absolute error is    4.42.

Epoch 3: Learning rate is 0.05.
Up to batch 0, the average loss is   23.39.
Up to batch 1, the average loss is   14.21.
Up to batch 2, the average loss is   10.66.
Up to batch 3, the average loss is    8.70.
Up to batch 4, the average loss is    7.49.
Up to batch 5, the average loss is    7.02.
Up to batch 6, the average loss is    6.53.
Up to batch 7, the average loss is    6.26.
Up to batch 8, the average loss is    5.98.
Up to batch 9, the average loss is    5.69.
Up to batch 10, the average loss is    5.43.
Up to batch 11, the average loss is    5.19.
Up to batch 12, the average loss is    5.05.
Up to batch 13, the average loss is    4.93.
Up to batch 14, the average loss is    4.91.
Up to batch 15, the average loss is    4.82.
The average loss for epoch 3 is    4.82 and mean absolute error is    1.65.

Epoch 4: Learning rate is 0.05000000074505806.
Up to batch 0, the average loss is    2.08.
Up to batch 1, the average loss is    3.01.
Up to batch 2, the average loss is    3.54.
Up to batch 3, the average loss is    4.26.
Up to batch 4, the average loss is    5.16.
Up to batch 5, the average loss is    6.44.
Up to batch 6, the average loss is    7.52.
Up to batch 7, the average loss is    8.42.
Up to batch 8, the average loss is    8.88.
Up to batch 9, the average loss is    9.54.
Up to batch 10, the average loss is   10.30.
Up to batch 11, the average loss is   10.46.
Up to batch 12, the average loss is   10.34.
Up to batch 13, the average loss is   10.14.
Up to batch 14, the average loss is    9.89.
Up to batch 15, the average loss is    9.68.
The average loss for epoch 4 is    9.68 and mean absolute error is    2.56.

Epoch 5: Learning rate is 0.05000000074505806.
Up to batch 0, the average loss is    5.41.
Up to batch 1, the average loss is    5.36.
Up to batch 2, the average loss is    5.97.
Up to batch 3, the average loss is    6.72.
Up to batch 4, the average loss is    7.56.
Up to batch 5, the average loss is    8.67.
Up to batch 6, the average loss is    9.58.
Up to batch 7, the average loss is   10.40.
Up to batch 8, the average loss is   10.62.
Up to batch 9, the average loss is   10.22.
Up to batch 10, the average loss is    9.63.
Up to batch 11, the average loss is    9.40.
Up to batch 12, the average loss is    9.43.
Up to batch 13, the average loss is    9.83.
Up to batch 14, the average loss is   10.75.
Up to batch 15, the average loss is   11.05.
The average loss for epoch 5 is   11.05 and mean absolute error is    2.75.

Epoch 6: Learning rate is 0.01.
Up to batch 0, the average loss is    7.07.
Up to batch 1, the average loss is    6.06.
Up to batch 2, the average loss is    5.31.
Up to batch 3, the average loss is    4.69.
Up to batch 4, the average loss is    4.44.
Up to batch 5, the average loss is    4.29.
Up to batch 6, the average loss is    4.06.
Up to batch 7, the average loss is    3.94.
Up to batch 8, the average loss is    3.88.
Up to batch 9, the average loss is    3.77.
Up to batch 10, the average loss is    3.61.
Up to batch 11, the average loss is    3.62.
Up to batch 12, the average loss is    3.54.
Up to batch 13, the average loss is    3.56.
Up to batch 14, the average loss is    3.58.
Up to batch 15, the average loss is    3.47.
The average loss for epoch 6 is    3.47 and mean absolute error is    1.43.

Epoch 7: Learning rate is 0.009999999776482582.
Up to batch 0, the average loss is    3.26.
Up to batch 1, the average loss is    3.40.
Up to batch 2, the average loss is    3.50.
Up to batch 3, the average loss is    3.38.
Up to batch 4, the average loss is    3.47.
Up to batch 5, the average loss is    3.36.
Up to batch 6, the average loss is    3.36.
Up to batch 7, the average loss is    3.31.
Up to batch 8, the average loss is    3.19.
Up to batch 9, the average loss is    3.18.
Up to batch 10, the average loss is    3.18.
Up to batch 11, the average loss is    3.16.
Up to batch 12, the average loss is    3.10.
Up to batch 13, the average loss is    3.06.
Up to batch 14, the average loss is    2.99.
Up to batch 15, the average loss is    2.90.
The average loss for epoch 7 is    2.90 and mean absolute error is    1.30.

Epoch 8: Learning rate is 0.009999999776482582.
Up to batch 0, the average loss is    3.65.
Up to batch 1, the average loss is    3.04.
Up to batch 2, the average loss is    3.05.
Up to batch 3, the average loss is    3.06.
Up to batch 4, the average loss is    2.93.
Up to batch 5, the average loss is    2.87.
Up to batch 6, the average loss is    2.81.
Up to batch 7, the average loss is    2.90.
Up to batch 8, the average loss is    2.88.
Up to batch 9, the average loss is    2.95.
Up to batch 10, the average loss is    2.96.
Up to batch 11, the average loss is    2.90.
Up to batch 12, the average loss is    2.95.
Up to batch 13, the average loss is    3.01.
Up to batch 14, the average loss is    2.97.
Up to batch 15, the average loss is    2.94.

Epoch 9: Learning rate is 0.005.
Up to batch 0, the average loss is    3.17.
Up to batch 1, the average loss is    3.22.
Up to batch 2, the average loss is    3.00.
Up to batch 3, the average loss is    2.87.
Up to batch 4, the average loss is    2.89.
Up to batch 5, the average loss is    2.84.
Up to batch 6, the average loss is    2.78.
Up to batch 7, the average loss is    2.85.
Up to batch 8, the average loss is    2.77.
Up to batch 9, the average loss is    2.78.
Up to batch 10, the average loss is    2.74.
Up to batch 11, the average loss is    2.72.
Up to batch 12, the average loss is    2.70.
Up to batch 13, the average loss is    2.68.
Up to batch 14, the average loss is    2.72.
Up to batch 15, the average loss is    2.83.
The average loss for epoch 9 is    2.83 and mean absolute error is    1.28.

Epoch 10: Learning rate is 0.004999999888241291.
Up to batch 0, the average loss is    2.72.
Up to batch 1, the average loss is    2.64.
Up to batch 2, the average loss is    2.72.
Up to batch 3, the average loss is    2.67.
Up to batch 4, the average loss is    2.62.
Up to batch 5, the average loss is    2.60.
Up to batch 6, the average loss is    2.60.
Up to batch 7, the average loss is    2.65.
Up to batch 8, the average loss is    2.59.
Up to batch 9, the average loss is    2.59.
Up to batch 10, the average loss is    2.58.
Up to batch 11, the average loss is    2.62.
Up to batch 12, the average loss is    2.65.
Up to batch 13, the average loss is    2.71.
Up to batch 14, the average loss is    2.77.
Up to batch 15, the average loss is    2.80.
The average loss for epoch 10 is    2.80 and mean absolute error is    1.27.

Epoch 11: Learning rate is 0.004999999888241291.
Up to batch 0, the average loss is    3.07.
Up to batch 1, the average loss is    2.80.
Up to batch 2, the average loss is    2.80.
Up to batch 3, the average loss is    2.79.
Up to batch 4, the average loss is    2.81.
Up to batch 5, the average loss is    2.80.
Up to batch 6, the average loss is    2.64.
Up to batch 7, the average loss is    2.62.
Up to batch 8, the average loss is    2.67.
Up to batch 9, the average loss is    2.68.
Up to batch 10, the average loss is    2.83.
Up to batch 11, the average loss is    2.77.
Up to batch 12, the average loss is    2.80.
Up to batch 13, the average loss is    2.82.
Up to batch 14, the average loss is    2.80.
Up to batch 15, the average loss is    2.76.
The average loss for epoch 11 is    2.76 and mean absolute error is    1.27.

Epoch 12: Learning rate is 0.001.
Up to batch 0, the average loss is    3.08.
Up to batch 1, the average loss is    2.42.
Up to batch 2, the average loss is    2.42.
Up to batch 3, the average loss is    2.34.
Up to batch 4, the average loss is    2.60.
Up to batch 5, the average loss is    2.56.
Up to batch 6, the average loss is    2.54.
Up to batch 7, the average loss is    2.59.
Up to batch 8, the average loss is    2.65.
Up to batch 9, the average loss is    2.76.
Up to batch 10, the average loss is    2.71.
Up to batch 11, the average loss is    2.76.
Up to batch 12, the average loss is    2.79.
Up to batch 13, the average loss is    2.71.
Up to batch 14, the average loss is    2.64.
Up to batch 15, the average loss is    2.64.
The average loss for epoch 12 is    2.64 and mean absolute error is    1.23.

Epoch 13: Learning rate is 0.0010000000474974513.
Up to batch 0, the average loss is    1.69.
Up to batch 1, the average loss is    2.27.
Up to batch 2, the average loss is    2.48.
Up to batch 3, the average loss is    2.42.
Up to batch 4, the average loss is    2.48.
Up to batch 5, the average loss is    2.57.
Up to batch 6, the average loss is    2.48.
Up to batch 7, the average loss is    2.55.
Up to batch 8, the average loss is    2.63.
Up to batch 9, the average loss is    2.68.
Up to batch 10, the average loss is    2.68.
Up to batch 11, the average loss is    2.66.
Up to batch 12, the average loss is    2.61.
Up to batch 13, the average loss is    2.64.
Up to batch 14, the average loss is    2.62.
Up to batch 15, the average loss is    2.60.
The average loss for epoch 13 is    2.60 and mean absolute error is    1.23.

Epoch 14: Learning rate is 0.0010000000474974513.
Up to batch 0, the average loss is    3.36.
Up to batch 1, the average loss is    2.91.
Up to batch 2, the average loss is    2.76.
Up to batch 3, the average loss is    2.80.
Up to batch 4, the average loss is    2.81.
Up to batch 5, the average loss is    2.82.
Up to batch 6, the average loss is    2.80.
Up to batch 7, the average loss is    2.76.
Up to batch 8, the average loss is    2.79.
Up to batch 9, the average loss is    2.65.
Up to batch 10, the average loss is    2.67.
Up to batch 11, the average loss is    2.62.
Up to batch 12, the average loss is    2.63.
Up to batch 13, the average loss is    2.61.
Up to batch 14, the average loss is    2.58.
Up to batch 15, the average loss is    2.62.
The average loss for epoch 14 is    2.62 and mean absolute error is    1.23.

<keras_core.src.callbacks.history.History at 0x16c1ed300>

 

組込み Keras コールバック

API doc を読むことにより既存の Keras コールバックを確実に調べてください。アプリケーションは CSV へのロギング、モデルのセーブ、TensorBoard 内のメトリクスの可視化、そしてそれ以上を含みます！

 

以上