CSIT5530 Final Project

• HUANG Yanzhen
• LI Kaiwen
• ZHENG Haoyu

Dataset

Process Dataset:

python -m data.dataset

A demo of our processed dataset and model evaluation statistics can be found in demo.

Model Training

Model training is done using llamafactory. Training config yaml can be found in demo/llamafactory_configs.

Model Evaluation

Our fine-tuned model (along with baseline) is uploaded to huggingface.

Download them by:

hf download YzHuangYanzhen/CSIT5530-FinalProject-HuangLiAndZheng --local-dir models

Evaluate model:

python -m eval