Practical overview
When it fits
Use it as a notebook-based reference or prototype when investigating how user rating sequences can inform movie-rating predictions.
Who it is for
- Developers exploring sequential recommendation with PyTorch.
- Researchers or learners studying transformer-based movie rating prediction.
- Practitioners prototyping recommendations from timestamped rating histories.
Documented uses
- Prepare MovieLens data for a sequence-based recommendation experiment.
- Inspect a notebook implementation of the BST model architecture.
- Explore prediction of a target movie’s rating from a user’s historical ratings.
Requirements & limits
The repository is public and lists an MIT license. The README links to a Colab notebook and documents a local notebook workflow.
- The local workflow separates data processing and model architecture; the data-preparation notebook is specified to run first.
- The README reports test MAE of 0.74 and RMSE of 0.93 for a split holding out each user’s latest rating, and notes overfitting around epochs 12–15.
- The implementation is notebook-based; the README does not provide a separate packaged service.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Prepare data and open the model notebook locally
Follow the repository’s documented local notebook sequence for the MovieLens-based recommender.
Before you start
- The repository’s Jupyter notebooks
- MovieLens 1M dataset
Steps
Run the data-processing notebook,
prepare_data.ipynb, first.Open the model architecture notebook,
pytorch-best.ipynb.
Expected outcome
The repository’s documented notebook sequence for preparing data and accessing the BST model architecture.
Guide limitations
- The README does not specify environment setup or package installation steps.
- The reported evaluation is tied to the README’s described rating split.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 17 October 2026
- Original website title
- jiwidi/Behavior-Sequence-Transformer-Pytorch · Official homepage title · retained in its source language
Related options
Keras MovieLens recommendations transformer example — A cited Keras implementation reference for transformer-based MovieLens recommendations.
DeepCTR BST implementation — A cited TensorFlow implementation reference for the Behavior Sequence Transformer.