Context-Aware Subtitle Translation on CPU: Deployment-Driven Training, Evaluation, and Latency Design
Read the original on arXiv Computation and Language →The paper presents a study on English‑to‑Traditional‑Chinese subtitle translation for a free browser extension that must run on a viewer’s CPU within about 500 ms per subtitle. Two models based on the LMT‑60 architecture, Sotto‑1.7B and a smaller Sotto‑0.6B, were trained with task‑specific vocabulary, continued pretraining, supervised fine‑tuning, weight interpolation, logit distillation, and on‑policy preference optimization. Evaluation via pairwise comparison against Google Translate shows Sotto‑1.7B achieving a 0.576 win rate and outperforming four open translation models, while latency optimizations such as idle‑time prefill and a compact vocabulary reduce end‑to‑end latency by 31% and 28% respectively, keeping the deployed model within the 500 ms target.
Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computation and Language.