Prefix Tuning

Files:

.
├── gpt2                          # Code for GPT2 style autoregressive LM
│   ├── train_e2e.py              # high-level scripts to train.
│   ├── train_control.py          # code that implements prefix-tuning.
│   ├── trainer_prefix.py         # trainer code for the training loop. 
│   ├── run_language_modeling.py  # training code (contains data loading, model loading, and calls trainer)
│   ├── gen.py                    # high-level scripts to decode. 
│   └── run_generation.py         # decoding code. 
│
├── seq2seq                       # Code for encoder-decoder architecture
│   ├── train_bart.py             # high-level scripts to train.
│   ├── prefixTuning.py           # code that implements prefix-tuning.
│   ├── finetune.py               # training code (contains data loading, model loading, and calls trainer)   
│   ├── lightning_base.py         # helper code
│   ├── utils.py                  # helper code
│   └── callbacks.py              # helper code
└── ...

To run the code for GPT2 style autoregressive LM, the code is in gpt2/. This corresponds to the table-to-text experiments in the paper.

To run the code for encoder-decoder architecture like BART, the code is in seq2seq. This corresponds to the summarization experiments in the paper.

The two primary scripts I used to run my codes are gpt2/train_e2e.py (for table-to-text) and seq2seq/train_bart.py(for summarization). they are set to default of good hyperparameters, and can be used to tune hyperparameter :)

Setup:

cd transformer; pip install -e .

Train via prefix-tuning:

cd gpt2;

python train_e2e.py --optim_prefix yes --preseqlen 5 --epoch 5 --learning_rate 0.00005 --mode webnlg --bsz 5 --seed 101

cd seq2seq; 

python train_bart.py --mode xsum --preseqlen 200 --do_train yes --fp16 yes --bsz 16  --epoch 30  --gradient_accumulation_step 3 --learning_rate 0.00005  --mid_dim 800

Other baseline approaches

cd gpt2;

python train_e2e.py --tuning_mode {finetune/adaptertune} --epoch 5 --learning_rate 0.00005 --mode webnlg --bsz 5 --seed 101

cd seq2seq;

python train_e2e.py --tuning_mode finetune --epoch 5 --learning_rate 0.00005 --mode webnlg --bsz 5 --seed 101

Decode:

cd gpt2;

python gen.py {data2text/webnlg/...} yes test {checkpoint_path} no

cd seq2seq; 

python train_bart.py --mode xsum --do_train no --prefix_model_path {checkpoint_path} --preseqlen {same as training} --mid_dim {same as training}

For details of the methods and results, please refer to our paper.

@misc{li2021prefixtuning,
      title={Prefix-Tuning: Optimizing Continuous Prompts for Generation}, 
      author={Xiang Lisa Li and Percy Liang},
      year={2021},
      eprint={2101.00190},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

Prefix-Tuning: Optimizing Continuous Prompts for Generation

Related tags

Overview

Prefix Tuning

Files:

Setup:

Train via prefix-tuning:

Decode:

Owner

Learning Off-Policy with Online Planning, CoRL 2021

A Closer Look at Reference Learning for Fourier Phase Retrieval

Code for Motion Representations for Articulated Animation paper

Finding Biological Plausibility for Adversarially Robust Features via Metameric Tasks

Data, notebooks, and articles associated with the RSNA AI Deep Learning Lab at RSNA 2021

Tensorflow Implementation of ECCV'18 paper: Multimodal Human Motion Synthesis

Evaluation and Benchmarking of Speech Super-resolution Methods

Equivariant Imaging: Learning Beyond the Range Space

Official code release for: EditGAN: High-Precision Semantic Image Editing

Generative Autoregressive, Normalized Flows, VAEs, Score-based models (GANVAS)

A heterogeneous entity-augmented academic language model based on Open Academic Graph (OAG)

Reinforcement learning library(framework) designed for PyTorch, implements DQN, DDPG, A2C, PPO, SAC, MADDPG, A3C, APEX, IMPALA ...

The implementation of the CVPR2021 paper "Structure-Aware Face Clustering on a Large-Scale Graph with 10^7 Nodes"

Learnable Boundary Guided Adversarial Training (ICCV2021)

Pip-package for trajectory benchmarking from "Be your own Benchmark: No-Reference Trajectory Metric on Registered Point Clouds", ECMR'21

Official code for CVPR2022 paper: Depth-Aware Generative Adversarial Network for Talking Head Video Generation

CLIPImageClassifier wraps clip image model from transformers

Beyond Image to Depth: Improving Depth Prediction using Echoes (CVPR 2021)

[ICCV 2021] A Simple Baseline for Semi-supervised Semantic Segmentation with Strong Data Augmentation

Codebase for "Revisiting spatio-temporal layouts for compositional action recognition" (Oral at BMVC 2021).