الملخص
Industrial sequential recommender systems operate over massive item catalogs (e.g., 10^5--10^7 items). Multi-label recommendation models are trained with Binary Cross-Entropy (BCE) loss over the full vocabulary, but standard BCE materializes a dense [B, N, V] logits tensor in High Bandwidth Memory (HBM), incurring prohibitive $O(BNV)$ memory and fatal Out-Of-Memory (OOM) errors. While chunked loss optimizations exist for Softmax Cross-Entropy in LLMs, large-scale multi-label BCE optimization remains unexplored across deep learning ecosystems. We propose CutBCE, an exact, hardware-accelerated BCE loss and gradient operator implemented in JAX and Pallas for large-vocabulary workloads. CutBCE introduces (1) an exact fused reformulation evaluating dense background loss and sparse target corrections; (2) a custom Vector-Jacobian Product (VJP) with a dedicated Pallas TPU backward kernel computing logit tiles on-chip in both passes so logits and their gradients never reside in HBM; (3) dynamic VMEM budgeting and sharding-aware collective hoisting for distributed meshes; and (4) count-based zero-overhead training metrics. On single-chip TPU v5e/v6e mini-benchmarks, CutBCE eliminates OOM errors with up to 91.9% speedup. On 8-chip TPU slice training for multi-label SASRec with 876k items (Yambda-50M), CutBCE reduces peak HBM by 65.7% (>14 GiB saved per chip) and increases training speed by 225.9% with comparable accuracy. CutBCE is open-sourced at https://github.com/AI-Hypercomputer/RecML/blob/main/recml/core/ops/binary_cross_entropy_ops.py.
الكلمات المفتاحية
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Li, Y., Ning, H., & Hammad, M. (2026). Cut Binary Cross Entropy: Efficient Large-Vocabulary Loss and Gradient Kernels for Sequential Recommendation. https://omanscience.com/ar/articles/cut-binary-cross-entropy-efficient-large-vocabulary-loss-and-gradient-kernels-for-sequential-recommendation
MLA 9
Li, Yaoyiran, et al. "Cut Binary Cross Entropy: Efficient Large-Vocabulary Loss and Gradient Kernels for Sequential Recommendation." https://omanscience.com/ar/articles/cut-binary-cross-entropy-efficient-large-vocabulary-loss-and-gradient-kernels-for-sequential-recommendation.
شيكاغو (المؤلف–التاريخ)
Li, Yaoyiran, Haowen Ning, and Mohamed Hammad. 2026. "Cut Binary Cross Entropy: Efficient Large-Vocabulary Loss and Gradient Kernels for Sequential Recommendation." https://omanscience.com/ar/articles/cut-binary-cross-entropy-efficient-large-vocabulary-loss-and-gradient-kernels-for-sequential-recommendation.
هارفارد
Li, Y., Ning, H. and Hammad, M. (2026) 'Cut Binary Cross Entropy: Efficient Large-Vocabulary Loss and Gradient Kernels for Sequential Recommendation', Available at: https://omanscience.com/ar/articles/cut-binary-cross-entropy-efficient-large-vocabulary-loss-and-gradient-kernels-for-sequential-recommendation.
فانكوفر
Li Y, Ning H, Hammad M. Cut Binary Cross Entropy: Efficient Large-Vocabulary Loss and Gradient Kernels for Sequential Recommendation. https://omanscience.com/ar/articles/cut-binary-cross-entropy-efficient-large-vocabulary-loss-and-gradient-kernels-for-sequential-recommendation
IEEE
Y. Li, H. Ning, and M. Hammad, "Cut Binary Cross Entropy: Efficient Large-Vocabulary Loss and Gradient Kernels for Sequential Recommendation," https://omanscience.com/ar/articles/cut-binary-cross-entropy-efficient-large-vocabulary-loss-and-gradient-kernels-for-sequential-recommendation.