LLM Training

This repository is intended to use for training HuggingFace models for the ReCo projects.

Installation with Docker

Build docker and start training as follows:

Docker Building:

docker build -t llm_training:latest .

run the following commmand if you can not build its docker, because we have a limited space and don't want to occupy unncessary spaces

sudo docker rmi $(docker images -f "dangling=true" -q) --force

Fine-tuning

Before training the model, you need to create .env file since we need to use GPU 1. Add GPU_DEVICE=1 to .env.

Model training:

screen -L -Logfile {LOGFILE_NAME} sudo docker run --rm --gpus all -v /reco/llm-training/:/app --env-file .env --name llm_training llm_training:latest bash scripts/spot/train_{MODEL_PREFIX}.sh

Model testing, you need to remove --train from the script and to run the above script.

To fine-tune Llama2, make sure that you add HF credentials in .env as follows:

HF_INFERENCE_TOKEN=YOUR_TOKEN

Zero-shot Learning

Zero-shot learning codes for ChatGPT and Llama2 are located under app/nshot/{model_name}_zero_shot.py.

Run the following command to get predictions from ChatGPT

bash scripts/{TASK_NAME}/zero_shot_chatgpt.sh

Run the following command to get predictions from Llama2

bash scripts/{TASK_NAME}/zero_shot_llama.sh