 Visualizza su TensorFlow.org |  Esegui in Google Colab |  Visualizza l'origine su GitHub |  Scarica quaderno |
La tolleranza ai guasti si riferisce a un meccanismo di salvataggio periodico degli stati degli oggetti tracciabili, come parametri e modelli. Ciò consente di recuperarli in caso di guasto del programma/macchina durante l'allenamento.
Questa guida mostra innanzitutto come aggiungere la tolleranza agli errori all'addestramento con tf.estimator.Estimator in TensorFlow 1 specificando il salvataggio della metrica con tf.estimator.RunConfig . Quindi imparerai come implementare la tolleranza agli errori per l'addestramento in Tensorflow 2 in due modi:
- Se utilizzi l'API Keras
Model.fit, puoi passarle il callbacktf.keras.callbacks.BackupAndRestore. - Se utilizzi un ciclo di addestramento personalizzato (con
tf.GradientTape), puoi salvare arbitrariamente i checkpoint utilizzando le APItf.train.Checkpointetf.train.CheckpointManager.
Entrambi questi metodi eseguiranno il backup e ripristineranno gli stati di addestramento nei file di checkpoint .
Impostare
import tensorflow.compat.v1 as tf1
import tensorflow as tf
import numpy as np
import tempfile
import time
mnist = tf.keras.datasets.mnist
(x_train, y_train),(x_test, y_test) = mnist.load_data()
x_train, x_test = x_train / 255.0, x_test / 255.0
TensorFlow 1: salva i checkpoint con tf.estimator.RunConfig
In TensorFlow 1, puoi configurare un tf.estimator per salvare i checkpoint in ogni passaggio configurando tf.estimator.RunConfig .
In questo esempio, inizia scrivendo un hook che genera artificialmente un errore durante il quinto checkpoint:
class InterruptHook(tf1.train.SessionRunHook):
# A hook for artificially interrupting training.
def begin(self):
self._step = -1
def before_run(self, run_context):
self._step += 1
def after_run(self, run_context, run_values):
if self._step == 5:
raise RuntimeError('Interruption')
Quindi, configura tf.estimator.Estimator per salvare ogni checkpoint e utilizzare il set di dati MNIST:
feature_columns = [tf1.feature_column.numeric_column("x", shape=[28, 28])]
config = tf1.estimator.RunConfig(save_summary_steps=1,
save_checkpoints_steps=1)
path = tempfile.mkdtemp()
classifier = tf1.estimator.DNNClassifier(
feature_columns=feature_columns,
hidden_units=[256, 32],
optimizer=tf1.train.AdamOptimizer(0.001),
n_classes=10,
dropout=0.2,
model_dir=path,
config = config
)
train_input_fn = tf1.estimator.inputs.numpy_input_fn(
x={"x": x_train},
y=y_train.astype(np.int32),
num_epochs=10,
batch_size=50,
shuffle=True,
)
INFO:tensorflow:Using config: {'_model_dir': '/tmp/tmpv15yxr9g', '_tf_random_seed': None, '_save_summary_steps': 1, '_save_checkpoints_steps': 1, '_save_checkpoints_secs': None, '_session_config': allow_soft_placement: true
graph_options {
rewrite_options {
meta_optimizer_iterations: ONE
}
}
, '_keep_checkpoint_max': 5, '_keep_checkpoint_every_n_hours': 10000, '_log_step_count_steps': 100, '_train_distribute': None, '_device_fn': None, '_protocol': None, '_eval_distribute': None, '_experimental_distribute': None, '_experimental_max_worker_delay_secs': None, '_session_creation_timeout_secs': 7200, '_checkpoint_save_graph_def': True, '_service': None, '_cluster_spec': ClusterSpec({}), '_task_type': 'worker', '_task_id': 0, '_global_id_in_cluster': 0, '_master': '', '_evaluation_master': '', '_is_chief': True, '_num_ps_replicas': 0, '_num_worker_replicas': 1}
WARNING:tensorflow:From /tmp/ipykernel_20837/314197976.py:17: The name tf.estimator.inputs is deprecated. Please use tf.compat.v1.estimator.inputs instead.
WARNING:tensorflow:From /tmp/ipykernel_20837/314197976.py:17: The name tf.estimator.inputs.numpy_input_fn is deprecated. Please use tf.compat.v1.estimator.inputs.numpy_input_fn instead.
Inizia ad addestrare il modello. Un'eccezione artificiale verrà sollevata dall'hook che hai definito in precedenza.
try:
classifier.train(input_fn=train_input_fn,
hooks=[InterruptHook()],
max_steps=10)
except Exception as e:
print(f'{type(e).__name__}:{e}')
WARNING:tensorflow:From /tmpfs/src/tf_docs_env/lib/python3.7/site-packages/tensorflow/python/training/training_util.py:397: Variable.initialized_value (from tensorflow.python.ops.variables) is deprecated and will be removed in a future version. Instructions for updating: Use Variable.read_value. Variables in 2.X are initialized automatically both in eager and graph (inside tf.defun) contexts. WARNING:tensorflow:From /tmpfs/src/tf_docs_env/lib/python3.7/site-packages/tensorflow_estimator/python/estimator/inputs/queues/feeding_queue_runner.py:65: QueueRunner.__init__ (from tensorflow.python.training.queue_runner_impl) is deprecated and will be removed in a future version. Instructions for updating: To construct input pipelines, use the `tf.data` module. WARNING:tensorflow:From /tmpfs/src/tf_docs_env/lib/python3.7/site-packages/tensorflow_estimator/python/estimator/inputs/queues/feeding_functions.py:491: add_queue_runner (from tensorflow.python.training.queue_runner_impl) is deprecated and will be removed in a future version. Instructions for updating: To construct input pipelines, use the `tf.data` module. INFO:tensorflow:Calling model_fn. INFO:tensorflow:Done calling model_fn. INFO:tensorflow:Create CheckpointSaverHook. INFO:tensorflow:Graph was finalized. INFO:tensorflow:Running local_init_op. INFO:tensorflow:Done running local_init_op. WARNING:tensorflow:From /tmpfs/src/tf_docs_env/lib/python3.7/site-packages/tensorflow/python/training/monitored_session.py:914: start_queue_runners (from tensorflow.python.training.queue_runner_impl) is deprecated and will be removed in a future version. Instructions for updating: To construct input pipelines, use the `tf.data` module. INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 0... INFO:tensorflow:Saving checkpoints for 0 into /tmp/tmpv15yxr9g/model.ckpt. INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 0... INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 1... INFO:tensorflow:Saving checkpoints for 1 into /tmp/tmpv15yxr9g/model.ckpt. INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 1... INFO:tensorflow:loss = 118.92192, step = 0 INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 2... INFO:tensorflow:Saving checkpoints for 2 into /tmp/tmpv15yxr9g/model.ckpt. INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 2... INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 3... INFO:tensorflow:Saving checkpoints for 3 into /tmp/tmpv15yxr9g/model.ckpt. INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 3... INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 4... INFO:tensorflow:Saving checkpoints for 4 into /tmp/tmpv15yxr9g/model.ckpt. INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 4... INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 5... INFO:tensorflow:Saving checkpoints for 5 into /tmp/tmpv15yxr9g/model.ckpt. WARNING:tensorflow:From /tmpfs/src/tf_docs_env/lib/python3.7/site-packages/tensorflow/python/training/saver.py:1054: remove_checkpoint (from tensorflow.python.training.checkpoint_management) is deprecated and will be removed in a future version. Instructions for updating: Use standard file APIs to delete files with this prefix. INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 5... INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 6... INFO:tensorflow:Saving checkpoints for 6 into /tmp/tmpv15yxr9g/model.ckpt. INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 6... RuntimeError:Interruption
Ricostruisci tf.estimator.Estimator utilizzando l'ultimo checkpoint salvato e continua la formazione:
classifier = tf1.estimator.DNNClassifier(
feature_columns=feature_columns,
hidden_units=[256, 32],
optimizer=tf1.train.AdamOptimizer(0.001),
n_classes=10,
dropout=0.2,
model_dir=path,
config = config
)
classifier.train(input_fn=train_input_fn,
max_steps = 10)
INFO:tensorflow:Using config: {'_model_dir': '/tmp/tmpv15yxr9g', '_tf_random_seed': None, '_save_summary_steps': 1, '_save_checkpoints_steps': 1, '_save_checkpoints_secs': None, '_session_config': allow_soft_placement: true
graph_options {
rewrite_options {
meta_optimizer_iterations: ONE
}
}
, '_keep_checkpoint_max': 5, '_keep_checkpoint_every_n_hours': 10000, '_log_step_count_steps': 100, '_train_distribute': None, '_device_fn': None, '_protocol': None, '_eval_distribute': None, '_experimental_distribute': None, '_experimental_max_worker_delay_secs': None, '_session_creation_timeout_secs': 7200, '_checkpoint_save_graph_def': True, '_service': None, '_cluster_spec': ClusterSpec({}), '_task_type': 'worker', '_task_id': 0, '_global_id_in_cluster': 0, '_master': '', '_evaluation_master': '', '_is_chief': True, '_num_ps_replicas': 0, '_num_worker_replicas': 1}
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Create CheckpointSaverHook.
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/tmpv15yxr9g/model.ckpt-6
WARNING:tensorflow:From /tmpfs/src/tf_docs_env/lib/python3.7/site-packages/tensorflow/python/training/saver.py:1161: get_checkpoint_mtimes (from tensorflow.python.training.checkpoint_management) is deprecated and will be removed in a future version.
Instructions for updating:
Use standard file utilities to get mtimes.
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 6...
INFO:tensorflow:Saving checkpoints for 6 into /tmp/tmpv15yxr9g/model.ckpt.
INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 6...
INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 7...
INFO:tensorflow:Saving checkpoints for 7 into /tmp/tmpv15yxr9g/model.ckpt.
INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 7...
INFO:tensorflow:loss = 105.44863, step = 6
INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 8...
INFO:tensorflow:Saving checkpoints for 8 into /tmp/tmpv15yxr9g/model.ckpt.
INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 8...
INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 9...
INFO:tensorflow:Saving checkpoints for 9 into /tmp/tmpv15yxr9g/model.ckpt.
INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 9...
INFO:tensorflow:Calling checkpoint listeners before saving checkpoint 10...
INFO:tensorflow:Saving checkpoints for 10 into /tmp/tmpv15yxr9g/model.ckpt.
INFO:tensorflow:Calling checkpoint listeners after saving checkpoint 10...
INFO:tensorflow:Loss for final step: 100.47882.
<tensorflow_estimator.python.estimator.canned.dnn.DNNClassifier at 0x7fcfe8165150>
TensorFlow 2: backup e ripristino con callback e Model.fit
In TensorFlow 2, se utilizzi l'API Keras Model.fit per l'addestramento, puoi fornire il callback tf.keras.callbacks.BackupAndRestore per aggiungere la funzionalità di tolleranza agli errori.
Per aiutare a dimostrarlo, iniziamo innanzitutto definendo una classe di callback che genera artificialmente un errore durante il quinto checkpoint:
class InterruptingCallback(tf.keras.callbacks.Callback):
# A callback for artificially interrupting training.
def on_epoch_end(self, epoch, log=None):
if epoch == 4:
raise RuntimeError('Interruption')
Quindi, definisci e istanzia un semplice modello Keras, definisci la funzione di perdita, chiama Model.compile e imposta un callback tf.keras.callbacks.BackupAndRestore che salverà i checkpoint in una directory temporanea:
def create_model():
return tf.keras.models.Sequential([
tf.keras.layers.Flatten(input_shape=(28, 28)),
tf.keras.layers.Dense(512, activation='relu'),
tf.keras.layers.Dropout(0.2),
tf.keras.layers.Dense(10)
])
loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True)
model = create_model()
model.compile(optimizer='adam',
loss=loss,
metrics=['accuracy'],
steps_per_execution=10)
log_dir = tempfile.mkdtemp()
backup_restore_callback = tf.keras.callbacks.BackupAndRestore(
backup_dir = log_dir
)
Ora, inizia ad addestrare il modello con Model.fit . Durante l'allenamento, i checkpoint verranno salvati grazie al backup_restore_callback sopra definito, mentre InterruptingCallback solleverà un'eccezione artificiale per simulare un errore.
try:
model.fit(x=x_train,
y=y_train,
epochs=10,
validation_data=(x_test, y_test),
callbacks=[backup_restore_callback, InterruptingCallback()])
except Exception as e:
print(f'{type(e).__name__}:{e}')
Epoch 1/10 1875/1875 [==============================] - 3s 2ms/step - loss: 0.2186 - accuracy: 0.9352 - val_loss: 0.1267 - val_accuracy: 0.9615 Epoch 2/10 1875/1875 [==============================] - 2s 1ms/step - loss: 0.0967 - accuracy: 0.9700 - val_loss: 0.0910 - val_accuracy: 0.9718 Epoch 3/10 1875/1875 [==============================] - 2s 1ms/step - loss: 0.0687 - accuracy: 0.9784 - val_loss: 0.0679 - val_accuracy: 0.9797 Epoch 4/10 1875/1875 [==============================] - 2s 1ms/step - loss: 0.0527 - accuracy: 0.9829 - val_loss: 0.0623 - val_accuracy: 0.9814 Epoch 5/10 1860/1875 [============================>.] - ETA: 0s - loss: 0.0434 - accuracy: 0.9857RuntimeError:Interruption
Quindi, crea un'istanza del modello Keras, chiama Model.compile e continua ad addestrare il modello con Model.fit da un checkpoint salvato in precedenza:
model = create_model()
model.compile(optimizer='adam',
loss=loss,
metrics=['accuracy'],
steps_per_execution=10)
model.fit(x=x_train,
y=y_train,
epochs=10,
validation_data=(x_test, y_test),
callbacks=[backup_restore_callback])
Epoch 6/10 1875/1875 [==============================] - 3s 2ms/step - loss: 0.0370 - accuracy: 0.9879 - val_loss: 0.0732 - val_accuracy: 0.9791 Epoch 7/10 1875/1875 [==============================] - 2s 1ms/step - loss: 0.0306 - accuracy: 0.9898 - val_loss: 0.0601 - val_accuracy: 0.9827 Epoch 8/10 1875/1875 [==============================] - 2s 1ms/step - loss: 0.0259 - accuracy: 0.9913 - val_loss: 0.0655 - val_accuracy: 0.9819 Epoch 9/10 1875/1875 [==============================] - 2s 1ms/step - loss: 0.0244 - accuracy: 0.9918 - val_loss: 0.0746 - val_accuracy: 0.9812 Epoch 10/10 1875/1875 [==============================] - 2s 1ms/step - loss: 0.0221 - accuracy: 0.9923 - val_loss: 0.0818 - val_accuracy: 0.9813 <keras.callbacks.History at 0x7fcfe0647350>
TensorFlow 2: scrivi checkpoint manuali con un ciclo di addestramento personalizzato
Se utilizzi un ciclo di addestramento personalizzato in TensorFlow 2, puoi implementare un meccanismo di tolleranza agli errori con le API tf.train.Checkpoint e tf.train.CheckpointManager .
Questo esempio mostra come:
- Utilizzare un oggetto
tf.train.Checkpointper creare manualmente un checkpoint, in cui gli oggetti tracciabili che si desidera salvare sono impostati come attributi. - Utilizzare un
tf.train.CheckpointManagerper gestire più checkpoint.
Inizia definendo e istanziando il modello Keras, l'ottimizzatore e la funzione di perdita. Quindi, crea un Checkpoint che gestisca due oggetti con stati tracciabili (il modello e l'ottimizzatore), nonché un CheckpointManager per la registrazione e il mantenimento di diversi checkpoint in una directory temporanea.
model = create_model()
optimizer = tf.keras.optimizers.SGD(learning_rate=0.001)
loss_fn = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True)
log_dir = tempfile.mkdtemp()
epochs = 5
steps_per_epoch = 5
checkpoint = tf.train.Checkpoint(model=model, optimizer=optimizer)
checkpoint_manager = tf.train.CheckpointManager(
checkpoint, log_dir, max_to_keep=2)
Ora, implementa un ciclo di addestramento personalizzato in cui dopo la prima epoca ogni volta che inizia una nuova epoca viene caricato l'ultimo checkpoint:
for epoch in range(epochs):
if epoch > 0:
tf.train.load_checkpoint(save_path)
print(f"\nStart of epoch {epoch}")
for step in range(steps_per_epoch):
with tf.GradientTape() as tape:
logits = model(x_train, training=True)
loss_value = loss_fn(y_train, logits)
grads = tape.gradient(loss_value, model.trainable_weights)
optimizer.apply_gradients(zip(grads, model.trainable_weights))
save_path = checkpoint_manager.save()
print(f"Checkpoint saved to {save_path}")
print(f"Training loss at step {step}: {loss_value}")
Start of epoch 0 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-1 Training loss at step 0: 2.3636362552642822 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-2 Training loss at step 1: 2.3626415729522705 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-3 Training loss at step 2: 2.3613197803497314 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-4 Training loss at step 3: 2.360600233078003 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-5 Training loss at step 4: 2.3589422702789307 Start of epoch 1 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-6 Training loss at step 0: 2.3563339710235596 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-7 Training loss at step 1: 2.3568854331970215 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-8 Training loss at step 2: 2.354109287261963 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-9 Training loss at step 3: 2.3532731533050537 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-10 Training loss at step 4: 2.351112127304077 Start of epoch 2 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-11 Training loss at step 0: 2.348905563354492 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-12 Training loss at step 1: 2.349478006362915 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-13 Training loss at step 2: 2.3487260341644287 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-14 Training loss at step 3: 2.345991611480713 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-15 Training loss at step 4: 2.3451104164123535 Start of epoch 3 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-16 Training loss at step 0: 2.3441312313079834 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-17 Training loss at step 1: 2.341529130935669 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-18 Training loss at step 2: 2.342329263687134 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-19 Training loss at step 3: 2.340449571609497 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-20 Training loss at step 4: 2.3367927074432373 Start of epoch 4 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-21 Training loss at step 0: 2.3366076946258545 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-22 Training loss at step 1: 2.335028886795044 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-23 Training loss at step 2: 2.3338520526885986 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-24 Training loss at step 3: 2.3345272541046143 Checkpoint saved to /tmp/tmpnr4ss2g8/ckpt-25 Training loss at step 4: 2.332385301589966
Prossimi passi
Per ulteriori informazioni sulla tolleranza agli errori e sul checkpoint in TensorFlow 2, prendere in considerazione la seguente documentazione:
- I documenti dell'API di callback
tf.keras.callbacks.BackupAndRestore. - I documenti dell'API
tf.train.Checkpointetf.train.CheckpointManager. - La Guida ai punti di controllo della formazione , inclusa la sezione Scrittura dei punti di controllo .
Potresti anche trovare utile il seguente materiale relativo alla formazione distribuita :
- La sezione Tolleranza agli errori nell'esercitazione Formazione multi-lavoratore con Keras .
- La sezione Gestione dell'errore dell'attività nell'esercitazione di addestramento del server dei parametri .