This project uses the following versions of Python and Torch:
- Python: 3.8.19
- Torch: 1.13.0+cu117
For a more detailed list of dependencies, please refer to the requirements.txt file.
Follow the steps below to run the Semantic Retrieval Augmented Contrastive Learning (SRA-CL):
cd build_datasets&prompts
# Replace <dataset> with the name of the selected dataset
jupyter notebook <dataset_name>.ipynb cd get_llmResponse&semanticEmb
# Obtain LLM's description for items
python obtain_response_item.py
# Obtain LLM's description for users
python obtain_response_user.py
# Transform items' textual descriptions into embeddings
python obtain_text_emb_item.py
# Transform users' textual descriptions into embeddings
python obtain_text_emb_user.pycd recommender_code
sh train.sh