🤗 Optimum provides an neural_compressor package that enables you to apply magnitude pruning on many model hosted on the 🤗 hub using the Intel Neural Compressor pruning API.
( config: typing.Union[str, optimum.intel.neural_compressor.configuration.IncPruningConfig] eval_func: typing.Optional[typing.Callable] train_func: typing.Optional[typing.Callable] )