This repo contains code to fine-tune GPT-J-6B with a famous quotes dataset. Originally, the repo downloaded and converted the model weights when GPTJ was not yet added to huggingface transformer package. That code can still be seen under the branch original_youtube
/quotes_dataset
contains the dataset properly formatted for fine-tuning. See repo for making this dataset here
/finetuning_repo
contains code orginally from the repo here that I have modified to work with GPT-J-6B
See the video for orignal repo code here for a video tutorial
- First create a conda envrionment and enter the environment
- Run the
./install_requirements.sh
script - Then you want to copy the data from
train.csv
andvalidation.csv
from/quotes_dataset
to the/finetuning_repo
folder - Run the finetuning code with appropriate flags to fine tune the model. See
example_run.txt