CODEWITHSUNDEEP
google-colaboratory
BIPUL MANDOL
BIPUL MANDOL

Reputation: 289

Keras model.fit Buffered data was truncated in google colab during model training

When training a model using Keras.model.fit, I run into "Buffered data was truncated after reaching the output size limit." in the output, and the output does not update anymore.

Epoch 14/30
    51/721 [=>............................] - ETA: 27s - loss: 0.7874 - dense_6_loss: 0.4921 - dense_7_loss: 0.1440 - dense_8_loss: 0.1513 - dense_6_accuracy: 0.8556 - dense_7_accuracy: 0.9532 - dense_8_accuracy: 0.9511Buffered data was truncated after reaching the output size limit.

This seems to be some sort of memory error Buffered data was truncated after reaching the output size limit

Even though the amount of text remains the same.

Upvotes: 2

Views: 1124

Answers (1)

Emad Helmi
Emad Helmi

Reputation: 695

I aslo have the same problem with epochs=100 and batch_size=1. When I set the keras verbosity to 2 the issue was solved. For save the results of the learning as you got in stdout you can simply use a logger just like this:

from keras.callbacks import CSVLogger
csv_logger = CSVLogger(
        "log.csv",
        append=True,
        separator=','
    )
history = model.fit(
        x_train,
        y_train,
        batch_size=batch_size,
        epochs=epochs,
        verbose=2,
        validation_data=(x_test, y_test),
        callbacks=[csv_logger]
    )

Upvotes: 2

Related Questions