Skip to content

Latest commit

 

History

History
49 lines (35 loc) · 1.61 KB

INSTALL.md

File metadata and controls

49 lines (35 loc) · 1.61 KB
Raw

Installation

Acknowledgement: This readme file for installing datasets is modified from PromptSRC's official repository.

This codebase is tested on Ubuntu 20.04.2 LTS with python 3.8. Follow the below steps to create environment and install dependencies.

  • Setup conda environment (recommended).
# Create a conda environment
conda create -y -n protext python=3.8

# Activate the environment
conda activate protext

# Install torch (requires version >= 1.8.1) and torchvision
# Please refer to https://pytorch.org/ if you need a different cuda version
pip install torch==1.9.0+cu111 torchvision==0.10.0+cu111 torchaudio==0.9.0 -f https://download.pytorch.org/whl/torch_stable.html
  • Clone ProText code repository and install requirements
# Clone PromptSRC code base
git clone https://github.com/muzairkhattak/ProText.git

cd ProText/
# Install requirements

pip install -r requirements.txt

# Update setuptools package 
pip install setuptools==59.5.0
  • Install dassl library.

Note: We have modified original dassl library in order to configure text-based data-loaders for training. Therefore it is strongly recommended to directly utilize Dassl library provided in this repository.

# Instructions borrowed from https://github.com/KaiyangZhou/Dassl.pytorch#installation

# No need to clone dassl library as it is already present in this repository
cd Dassl.pytorch/

# Install dependencies
pip install -r requirements.txt

# Install this library (no need to re-build if the source code is modified)
python setup.py develop
cd ..