To get started with this repository, follow the instructions below:
- Clone the repository.
- In the root folder of the repository, run the following command to initialize the submodules:
git submodule update --init --recursive
Before running the code in this repository, you need to set up the necessary environments.
-
Install Anaconda by following the instructions here. You can choose either Anaconda or Miniconda.
-
For GRU4REC-pytorch and Torch-GRU4Rec, you need to install PyTorch. The required packages for the PyTorch versions can be found in
gru4rec_conda_pytorch_env.yaml. To create the environment, run the following command:conda env create -f gru4rec_conda_pytorch_env.yaml
-
For GRU4Rec_TensorFlow and KerasGRU4Rec, you need to install TensorFlow. The required packages for the TensorFlow versions can be found in
gru4rec_conda_tensorflow_env.yaml. To create the environment, run the following command:conda env create -f gru4rec_conda_tensorflow_env.yaml
-
For recpack, you need to install PyTorch. The required packages for this version can be found in
gru4rec_conda_recpack_env.yaml. To create the environment, run the following command:conda env create -f gru4rec_conda_recpack_env.yaml
-
Once the environments are created, you can activate them by running one of the following commands:
- For TensorFlow environment:
conda activate gru4rec_tensorflow
- For PyTorch environment:
conda activate gru4rec_pytorch
- For recpack environment:
conda activate gru4rec_recpack
- For TensorFlow environment:
-
To deactivate the environment, run the following command:
conda deactivate
Please refer to the requirements mentioned in the repository for more details.
This repository provides an example Jupyter Notebook for each model, along with descriptions on how to run the experiments. Refer to the notebooks for detailed instructions.
- Yoochoose:
- Rees46:
- Coveo:
- Diginetica:
- Retailrocket: