train Namespace Reference

Module caffe2.experiments.python.train. More...

Functions
def	FakeData (args, model)
def	NNLoaderData (args, model)
def	sparse_transform (model)
def	train (model_gen, data_gen, args)
def	mlp_model (args, model, input_data)
def	mlp_decomp_model (args, model, input_data)
def	mlp_prune_model (args, model, input_data)
def	sparse_mlp_model (args, model, input_data)
def	debug_sparse_mlp_model (args, model, input_data)
def	debug_sparse_mlp_decomposition_model (args, model, input_data)
def	debug_sparse_mlp_prune_model (args, model, input_data)
def	Iter (model, blob_out, kwargs)
def	Accuracy (model, blob_in, blob_out, kwargs)

Variables
	logger = logging.getLogger(__name__)
	InputData
dictionary	MODEL_TYPE_FUNCTIONS
	parser = pyinit.FbcodeArgumentParser(description='Ads NN trainer')
	type
	int
	default
	help
	str
	float
	choices
	action
	args = parser.parse_args()
def	data_gen = NNLoaderData if args.loaderConfig else FakeData

Detailed Description

Module caffe2.experiments.python.train.

Module caffe2.python.helpers.train.

Variable Documentation

InputData

train.InputData

Initial value:

=  namedtuple(
    'InputData', ['data', 'label', 'weight', 'prod_pred', 'sparse_segments'])

Definition at line 61 of file train.py.

MODEL_TYPE_FUNCTIONS

dictionary train.MODEL_TYPE_FUNCTIONS

Initial value:

=  {
    'mlp': mlp_model,
    'mlp_decomp': mlp_decomp_model,
    'mlp_prune': mlp_prune_model,
    'sparse_mlp': sparse_mlp_model,
    'debug_sparse_mlp': debug_sparse_mlp_model,
    'debug_sparse_mlp_decomposition': debug_sparse_mlp_decomposition_model,
    'debug_sparse_mlp_prune': debug_sparse_mlp_prune_model,
    # Add more model_type functions here.
}

Definition at line 259 of file train.py.