README

how to run

In order to avoid excessive memory usage, we recommend that you modify the config file and run it in the following order.The config file is ./data/config/cloth.json

download dataset

from https://mcauleylab.ucsd.edu/public_datasets/data/amazon_v2/categoryFiles/ to download dataset and move the file to ./data/cloth/

download Llama

from transformers to download llama and replace the llama_path in ./work/all.py from YOUR_LLAMA_PATH to the actual location of your Llama model

first pass

change skip.part4, skip.part5 to False and run python ./work/all.py

second pass

change skip.part1, skip.part2, skip.part3 to False , skip.part4 to True and run python ./work/all.py

third pass

change skip.part4 to False , skip.part5 to True and run python ./work/all.py