cerebras.modelzoo.data_preparation.data_preprocessing.data_preprocessor

This module implements a generic data preprocessor called DataPreprocessor. It internally uses DataFrame and DataReader to read and process data.

Functions

check_and_create_dir	Ensures a directory exists, optionally handling a subdirectory.
default_hook
format_time	Format seconds into a human-readable string showing hours:minutes:seconds, minutes:seconds, or seconds.
get_available_memory	Returns available memory in bytes.
get_compression_factor	Calculate and return the compression factor based on a file's extension.
save_image_locally
update_progress	Update the progress bar based on the current progress.

Classes

DataPreprocessor

Initialize the class with given parameters.

previous

cerebras.modelzoo.data_preparation.data_preprocessing.data_dedup.to_hash.to_minhash

next

cerebras.modelzoo.data_preparation.data_preprocessing.data_preprocessor.check_and_create_dir