caffe2.python.crf.CRFWithLoss Class Reference

Inheritance diagram for caffe2.python.crf.CRFWithLoss:

Public Member Functions
def	__init__ (self, model, num_classes, transitions_blob=None)
def	crf_loss (self, predictions, labels, seq_lengths=None)
def	build_crf_net (self, input_blob, initial_state, transitions)
def	update_predictions (self, classes)

Static Public Member Functions
def	pad_labels (labels, init_net, net, num_classes)
def	pad_predictions (predictions, init_net, net, num_classes)

Public Attributes
	model
	num_classes
	num_classes_padded
	transitions

Detailed Description

Definition at line 16 of file crf.py.

Member Function Documentation

def caffe2.python.crf.CRFWithLoss.build_crf_net	(		self,
			input_blob,
			initial_state,
			transitions
	)

    Adds the crf_net recurrent operator to the model.

    model: model_helper.ModelHelper object new operators would be added
    to

    input_blob: the input sequence in a format T x N x D
    where T is sequence size, N - batch size and D - input dimention
    ##Only supports batch-size 1##

    seq_lengths: blob containing sequence lengths (unused)

Definition at line 136 of file crf.py.

---

The documentation for this class was generated from the following file:

• caffe2/python/crf.py