all convolutional blocks 4 kinds of conv blocks, which #feature_map are 64,128,256,512 Depth: 9 17 29 49 --------

Format evaluation string. Args: category: Category evaluation to format. Returns:

For masking out the padding part of key sequence.

Use dict to convert all vocabs to ids

Init confusion matrix. Args: label_map: Label map. Returns: confusion_matrix.

Init count dict. Args: label_map: Label map. Returns: count_dict.

Insert vocab to dict

Convolutional Block of VDCNN Convolutional block contains 2 conv layers, and can be repeated Temp Conv-->Batch Norm-->ReLU-->T

Log debug message msg: Message to log

Log error message msg: Message to log

Args: logits: model's output, shape of [batch_size, num_cls] target: ground truth labels, shape of [batch_size]

Args: inputs: seq_lengths: init_state: ori_state: If true, will return ori state generate by

Update lr

Update lr

Update lr

Update lr