Is Sliding Window All You Need? An Open Framework for Long-Sequence Recommendation
For researchers in recommender systems, this paper provides an open, extensible framework and empirical evidence that long-sequence training is feasible on modest hardware, addressing a gap between industrial practice and academic research.
This work demonstrates that long-sequence training with sliding windows is practical and effective at academic scale, releasing a complete framework that achieves up to +6.04% MRR and +6.34% Recall@10 on Retailrocket with ~4x training-time overheads, while introducing a k-shift embedding layer for million-scale vocabularies on commodity GPUs.
Long interaction histories are central to modern recommender systems, yet training with long sequences is often dismissed as impractical under realistic memory and latency budgets. This work demonstrates that it is not only practical but also effective-at academic scale. We release a complete, end-to-end framework that implements industrial-style long-sequence training with sliding windows, including all data processing, training, and evaluation scripts. Beyond reproducing prior gains, we contribute two capabilities missing from earlier reports: (i) a runtime-aware ablation study that quantifies the accuracy-compute frontier across windowing regimes and strides, and (ii) a novel k-shift embedding layer that enables million-scale vocabularies on commodity GPUs with negligible accuracy loss. Our implementation trains reliably on modest university clusters while delivering competitive retrieval quality (e.g., up to +6.04% MRR and +6.34% Recall@10 on Retailrocket) with $\sim 4 \times $ training-time overheads. By packaging a robust pipeline, reporting training time costs, and introducing an embedding mechanism tailored for low-resource settings, we transform long-sequence training from a closed, industrial technique into a practical, open, and extensible methodology for the community.