Skip to content

Latest commit

 

History

History
16 lines (16 loc) · 670 Bytes

INSTALL.md

File metadata and controls

16 lines (16 loc) · 670 Bytes

Installation

Step 1: Create a new conda environment:

conda create -n prune_llm python=3.9
conda activate prune_llm

Step 2: Install relevant packages

conda install pytorch==1.10.1 torchvision==0.11.2 torchaudio==0.10.1 cudatoolkit=11.3 -c pytorch -c conda-forge
pip install transformers==4.28.0 datasets==2.11.0 evaluate wandb sentencepiece
pip install accelerate==0.18.0
pip install lm_eval==0.3.0
pip install ftfy
pip install peft

There are known issues with the transformers library on loading the LLaMA tokenizer correctly. Please follow the mentioned suggestions to resolve this issue.