cerebras.modelzoo.config_manager.config_loader

Utility to load the config from yaml or .py config file

Functions

flatten_optimizer_params	Config classes package optimizer related params in a sub dict.
flatten_sparsity_params	Config classes package sparsity related params in a sub dict.
get_config_from_class	Read the config class file and returns a config object
get_config_from_yaml	Get the config object after reading input yaml.
process_config	Perform config mapping and validation :param config: The config class object :param config_class: The clas the config belongs to :param params_conf: Dictionary of params
read_from_config_class_file	Read the config class file and call the config generator function to get config object
validate_config_params	Load the config class and run validation check on the config based on parameter constraints :param params_conf: The config params passed as a dict :param model_name: The model key name used by config map to check what class of config to use

previous

cerebras.modelzoo.config_manager.config_classes.base.sparsity_config.SparsityBaseConfig

next

cerebras.modelzoo.config_manager.config_loader.flatten_optimizer_params