Multi-task Learning with TensorFlow

Tensorflow implementation of multi-task learning. (Language model & Text classification)

Model

Requirements

• Python3
• TensorFlow
• pip install -r requirements.txt

Usage

$ python train.py

Hyperparameters

$ python train.py -h
usage: train.py [-h] [--embedding_size EMBEDDING_SIZE]
                [--num_layers NUM_LAYERS] [--num_hidden NUM_HIDDEN]
                [--keep_prob KEEP_PROB] [--learning_rate LEARNING_RATE]
                [--batch_size BATCH_SIZE] [--num_epochs NUM_EPOCHS]
                [--max_document_len MAX_DOCUMENT_LEN]

optional arguments:
  -h, --help            show this help message and exit
  --embedding_size EMBEDDING_SIZE
                        embedding size.
  --num_layers NUM_LAYERS
                        RNN network depth.
  --num_hidden NUM_HIDDEN
                        RNN network size.
  --keep_prob KEEP_PROB
                        dropout keep prob.
  --learning_rate LEARNING_RATE
                        learning rate.
  --batch_size BATCH_SIZE
                        batch size.
  --num_epochs NUM_EPOCHS
                        number of epochs.
  --max_document_len MAX_DOCUMENT_LEN
                        max document length.

Experimental Results

Language Model Training Loss

Text Classification Training Loss