examples

Examples

Rich examples are included to demonstrate the use of Texar. The implementations of cutting-edge models/algorithms also provide references for reproducibility and comparisons.

More examples are continuously added...

Examples by Models/Algorithms

RNN / Seq2seq

• language_model_ptb: Basic RNN language model
• distributed_gpu: Basic RNN language model with distributed training
• seq2seq_attn: Attentional seq2seq
• seq2seq_configs: Seq2seq implemented with Texar model template
• seq2seq_rl: Attentional seq2seq trained with policy gradient
• seq2seq_exposure_bias: Various algorithms tackling exposure bias in sequence generation
• hierarchical_dialog: Hierarchical recurrent encoder-decoder model for conversation response generation
• torchtext: Use of torchtext data loader

Transformer (Self-attention)

• transformer: Transformer for machine translation
• bert: Pre-trained BERT model for text representation
• gpt-2: Pre-trained OpenAI GPT-2 language model
• vae_text: VAE with a transformer decoder for improved language modeling

Variational Autoencoder (VAE)

• vae_text: VAE language model

GANs / Discriminiator-supervision

• seqGAN: GANs for text generation
• text_style_transfer: Discriminator supervision for controlled text generation

Reinforcement Learning

• seq2seq_rl: Attentional seq2seq trained with policy gradient.
• seqGAN: Policy gradient for sequence generation
• rl_gym: Various RL algoritms for games on OpenAI Gym

Memory Network

• memory_network_lm: End-to-end memory network for language modeling

Classifier / Sequence Prediction

• bert: Pre-trained BERT model for text representation
• sentence_classifier: Basic CNN-based sentence classifier
• sequence_tagging: BiLSTM-CNN model for Named Entity Recognition (NER)

Reward Augmented Maximum Likelihood (RAML)

• seq2seq_exposure_bias: RAML and other learning algorithms for sequence generation

---

Examples by Tasks

Language Modeling

• gpt-2: Pre-trained OpenAI GPT-2 language model
• language_model_ptb: Basic RNN language model
• vae_text: VAE language model
• seqGAN: GAN + policy gradient
• memory_network_lm: End-to-end memory network for language modeling

Machine Translation

• seq2seq_attn: Attentional seq2seq
• seq2seq_configs: Seq2seq implemented with Texar model template.
• seq2seq_rl: Attentional seq2seq trained with policy gradient.
• seq2seq_exposure_bias: Various algorithms tackling exposure bias in sequence generation (MT and summarization as examples).
• transformer: Transformer for machine translation

Dialog

• hierarchical_dialog: Hierarchical recurrent encoder-decoder model for conversation response generation.

Text Summarization

• seq2seq_exposure_bias: Various algorithms tackling exposure bias in sequence generation (MT and summarization as examples).

Text Style Transfer

• text_style_transfer: Discriminator supervision for controlled text generation

Classification

• bert: Pre-trained BERT model for text representation
• sentence_classifier: Basic CNN-based sentence classifier

Sequence Tagging

• sequence_tagging: BiLSTM-CNN model for Named Entity Recognition (NER)

Games

• rl_gym: Various RL algoritms for games on OpenAI Gym

---

MISC

Distributed training

• distributed_gpu: Basic example of distributed training.
• bert: Distributed training of BERT.