Module: tf.keras.config

DO NOT EDIT.

This file was autogenerated. Do not edit it by hand, since your modifications would be overwritten.

Functions

backend(...): Publicly accessible method for determining the current backend.

disable_flash_attention(...): Disable flash attention.

disable_interactive_logging(...): Turn off interactive logging.

disable_traceback_filtering(...): Turn off traceback filtering.

dtype_policy(...): Returns the current default dtype policy object.

enable_flash_attention(...): Enable flash attention.

enable_interactive_logging(...): Turn on interactive logging.

enable_traceback_filtering(...): Turn on traceback filtering.

enable_unsafe_deserialization(...): Disables safe mode globally, allowing deserialization of lambdas.

epsilon(...): Return the value of the fuzz factor used in numeric expressions.

floatx(...): Return the default float type, as a string.

image_data_format(...): Return the default image data format convention.

is_flash_attention_enabled(...): Checks whether flash attention is globally enabled in Keras.

is_interactive_logging_enabled(...): Check if interactive logging is enabled.

is_nnx_enabled(...): Checks whether NNX specific features are enabled for the JAX backend.

is_traceback_filtering_enabled(...): Check if traceback filtering is enabled.

max_epochs(...): Get the maximum number of epochs for any call to fit.

max_steps_per_epoch(...): Get the maximum number of steps for any call to fit/evaluate/predict.

set_backend(...): Reload the backend (and the Keras package).

set_dtype_policy(...): Sets the default dtype policy globally.

set_epsilon(...): Set the value of the fuzz factor used in numeric expressions.

set_floatx(...): Set the default float dtype.

set_image_data_format(...): Set the value of the image data format convention.

set_max_epochs(...): Limit the maximum number of epochs for any call to fit.

set_max_steps_per_epoch(...): Limit the maximum number of steps for any call to fit/evaluate/predict.