ContRec

This is an early access release of the code and data for the paper "Diffusion Generative Recommendation with Continuous Tokens."

ContRec represents users\&items as latent vector representations using a not-quantized tokenizer and leverages the exceptional continuous-valued generation capability of diffusion models to operate within continuous spaces and generate implicit user preferences conditioned on the reasoning content of LLMs.

Setup Environment

You can install the useful wheels by:

pip install -r requirements.txt

Please download the any open-source LLM, such as LLaMA-3.2-1B-Instruct, to the path: "ContRec/code/huggingface_path/".

A simple Example of Implementation

All code for a complete implementation of ContRec (including finetuning, validation, and testing) is included in the "code" folder.

Please change your path to the "code" folder.

cd code

Also replace input a correct "hf_token" in Line 228 of the "model_interface.py" file

hf_token = "Your_HF_TOKEN"

We can simply run the "main.py" file to go through the pipeline:

python main.py

More configurations can be found in the "parse.py" file.