dmx.compressor.modeling.model.DmxModelMixin

class dmx.compressor.modeling.model.DmxModelMixin

__init__()

Methods

__init__()
check_dim_consistency()	A function that checks format dimension consistency and sparseness dimension consistency for all applicable dmx modules in the model
configure(config, *rules)	Configure Dmx-specific numerics/sparsity/logics
counting_flops([zero])
create_submod_transform_forward(model, ...)	Only supported for fx path, submodule forward can be directly called in export path
deepcopy_args(args)
fold_weights_and_biases()	A function that applies the ops the weights and biases using the corresponding formats.
forward_weight_hypernets()
freeze([config_file])	A function that stores the state and ops format of the model to a config file
get_monitoring_records([submodules_to_monitor])
get_runtime_records()
is_same_signature(_model, args, kwargs)
keep_dmx_config()
measure_runtimes(device[, submodules_to_measure])
monitoring([submodules_to_monitor, ...])
named_dmx_modules()	"Returns a list of named modules that are dmx configurable
post_process_gm(_model, args, kwargs)
print_model_tree([include_type])	A function that prints out the tree structure of a model
process_inputs_for_export(model, args, kwargs)
thaw([config_file])	A function that transforms the model in place from a config file.
to_baseline_mode()
to_basic_mode([sbfp_weight_storage])	Configures a transformed DmxModel to the BASIC mode on dmx hardware.
to_fp8_mode()	Configures a transformed DmxModel to the FP8 mode on dmx hardware.
to_old_forward(_m)
to_signature_key(_m, _args, _kwargs)
to_transformed_forward(_m)
transform(config, *rules)	Configure Dmx-specific numerics/sparsity/logics

Attributes

dmx_config	"Returns the DmxConfig object for the model
dmx_module_names	"Returns a list of module names listed in a dmx_config
op_set	Returns a set of unique ops present in the model
transformed
additional_dmx_aware_mappings