Chain Dimension Scaling: Communication-Efficient Loop-Style Scaling for Industrial Short-Video Ranking
Industrial recommendation rankers keep scaling in width and depth to billions of parameters, but this increasingly incurs costs in parameter growth, distributed communication overhead, and training inefficiency.
ProofPaper ↗
Key points
- We introduce \underline{C}hain \underline{D}imension \underline{S}caling (CDS), a scaling strategy orthogonal to width and depth.
- In the compression regime, shallow blocks with a longer chain match deeper baselines with fewer parameters and faster training; in the quality-scaling regime, increasing chain length under a fixed parameter budget yields consistent offline and online gains, suggesting the chain dimension adds modeling capacity beyond parameter reuse.
- We evaluate CDS in a large-scale multi-task short-video ranking system in production.
- These results position parameter-shared iterative computation as a practical, communication-efficient scaling axis for industrial ranking systems.
Sources (1)
- [1]Chain Dimension Scaling: Communication-Efficient Loop-Style Scaling for Industrial Short-Video RankingarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 12, 08:49 AM
Industrial recommendation rankers keep scaling in width and depth to billions of parameters, but this increasingly incurs costs in parameter growth, distributed communication overhead, and training inefficiency.
We introduce \underline{C}hain \underline{D}imension \underline{S}caling (CDS), a scaling strategy orthogonal to width and depth.
Extractive summary: sentences quoted from the sources.