Implementation of "The Power of Scale for Parameter-Efficient Prompt Tuning"

Last update: Dec 19, 2022

Related tags

Overview

Prompt-Tuning

Implementation of "The Power of Scale for Parameter-Efficient Prompt Tuning" Currently, we support the following huggigface models:

BartForConditionalGeneration

Setup

conda create -n prompt-tuning python==3.7.0
pip install -r requirements.txt

Data

The task of summarization supports custom CSV and JSONLINES formats.

You can use data2json.py to transformer data to JSONLINES formats.

Custom CSV Files

If it's a csv file the training and validation files should have a column for the inputs texts and a column for the summaries.

If the csv file has just two columns as in the following example:

text,summary
"I'm sitting here in a boring room. It's just another rainy Sunday afternoon. I'm wasting my time I got nothing to do. I'm hanging around I'm waiting for you. But nothing ever happens. And I wonder","I'm sitting in a room where I'm waiting for something to happen"
"I see trees so green, red roses too. I see them bloom for me and you. And I think to myself what a wonderful world. I see skies so blue and clouds so white. The bright blessed day, the dark sacred night. And I think to myself what a wonderful world.","I'm a gardener and I'm a big fan of flowers."
"Christmas time is here. Happiness and cheer. Fun for all that children call. Their favorite time of the year. Snowflakes in the air. Carols everywhere. Olden times and ancient rhymes. Of love and dreams to share","It's that time of year again."

The first column is assumed to be for text and the second is for summary.

If the csv file has multiple columns, you can then specify the names of the columns to use:

    --text_column text_column_name \
    --summary_column summary_column_name \

For example if the columns were:

id,date,text,summary

and you wanted to select only text and summary, then you'd pass these additional arguments:

    --text_column text \
    --summary_column summary \

Custom JSONLINES Files

The second supported format is jsonlines. Here is an example of a jsonlines custom data file.

{"text": "I'm sitting here in a boring room. It's just another rainy Sunday afternoon. I'm wasting my time I got nothing to do. I'm hanging around I'm waiting for you. But nothing ever happens. And I wonder", "summary": "I'm sitting in a room where I'm waiting for something to happen"}
{"text": "I see trees so green, red roses too. I see them bloom for me and you. And I think to myself what a wonderful world. I see skies so blue and clouds so white. The bright blessed day, the dark sacred night. And I think to myself what a wonderful world.", "summary": "I'm a gardener and I'm a big fan of flowers."}
{"text": "Christmas time is here. Happiness and cheer. Fun for all that children call. Their favorite time of the year. Snowflakes in the air. Carols everywhere. Olden times and ancient rhymes. Of love and dreams to share", "summary": "It's that time of year again."}

Same as with the CSV files, by default the first value will be used as the text record and the second as the summary record. Therefore you can use any key names for the entries, in this example text and summary were used.

And as with the CSV files, you can specify which values to select from the file, by explicitly specifying the corresponding key names. In our example this again would be:

    --text_column text \
    --summary_column summary \

Train

bash run_train.sh

You can adjust the values for the arguments --train_file, --validation_file in run_train.sh

To control the prompt length, you can adjust the values for the arguments --pre_seq_len in run_train.sh.

Other setting, such as learning rate, batch_size, you can also adjust in run_train.sh.

Test

bash run_test.sh

You can adjust the values for the arguments --test_file in run_test.sh

Other setting, you can also adjust in run_test.sh. The generated summary is in output_dir/generated_predictions.txt

Citation

@misc{lester2021power,
      title={The Power of Scale for Parameter-Efficient Prompt Tuning}, 
      author={Brian Lester and Rami Al-Rfou and Noah Constant},
      year={2021},
      eprint={2104.08691},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

Implementation of "The Power of Scale for Parameter-Efficient Prompt Tuning"

Related tags

Overview

Prompt-Tuning

Setup

Data

Custom CSV Files

Custom JSONLINES Files

Train

Test

Citation

Owner

Andrew Zeng

Official Matlab Implementation for "Tiny Obstacle Discovery by Occlusion-aware Multilayer Regression", TIP 2020

This is the official source code of "BiCAT: Bi-Chronological Augmentation of Transformer for Sequential Recommendation".

Deep Learning and Reinforcement Learning Library for Scientists and Engineers 🔥

A cross-lingual COVID-19 fake news dataset

Stacked Hourglass Network with a Multi-level Attention Mechanism: Where to Look for Intervertebral Disc Labeling

Encoding Causal Macrovariables

View model summaries in PyTorch!

AquaTimer - Programmable Timer for Aquariums based on ATtiny414/814/1614

Fast mesh denoising with data driven normal filtering using deep variational autoencoders

A simple, fast, and efficient object detector without FPN

Code for ACL2021 long paper: Knowledgeable or Educated Guess? Revisiting Language Models as Knowledge Bases

Implementation for the IJCAI2021 work "Beyond the Spectrum: Detecting Deepfakes via Re-synthesis"

Lux AI environment interface for RLlib multi-agents

Code and hyperparameters for the paper "Generative Adversarial Networks"

[3DV 2021] Channel-Wise Attention-Based Network for Self-Supervised Monocular Depth Estimation

Distributed DataLoader For Pytorch Based On Ray

Automated Hyperparameter Optimization Competition

RINDNet: Edge Detection for Discontinuity in Reflectance, Illumination, Normal and Depth, in ICCV 2021 (oral)

3D-printable hand-strapped keyboard

An Ensemble of CNN (Python 3.5.1 Tensorflow 1.3 numpy 1.13)