A single model scale challenges the flexibility of a production retrieval system: some settings need it faster, others need a smaller index, and the right trade-off changes with the workload. In the context of information retrieval (IR), a transformer-based model can be made smaller in three ways---using fewer layers,...
Yu Wang, Shengyao Zhuang, Xue-Guang Ma et al.· 1 citation
Speculative decoding reduces large language model inference latency by drafting multiple tokens before target-model verification, but its effectiveness depends on both the proposal mechanism and draft depth. Fixed configurations cannot respond to changes in predictability, repetition, and acceptance during generation,...
Manvi Jha, Zach Zhang, Zhi-Chao Xu et al.· 0 citations
Combining capabilities of multiple expert models trained starting from the same base checkpoint has become increasingly common in frontier language-model post-training. Recent trends suggest that multi-teacher on-policy distillation (MOPD) outperforms conventional off-policy methods. However, despite the higher inferen...
Roy Xie, Dan Friedman, Feng Nan et al.· 0 citations
CLEAR first employs a reflection agent to perform contrastive analysis over past execution trajectories and summarize useful context for each observed task to train a context augmentation model (CAM), which further optimize CAM using reinforcement learning, where the reward signal is obtained by running the task execut...
Lin-Bo Liu, Guan Wu, Han Ding et al.· arXiv.org· 0 citations
A novel simulation approach that combines categorical judgments with evaluator-specific auxiliary data--retrospective reasoning traces and interface telemetry--to enable LLM-based simulation of individual evaluators via in-context learning is proposed.
Zeyu He, Xuan Qi, Subramanian Chidambaram et al.· SIGDIAL Conferences· 0 citations
Tevatron 3.0 is presented, which integrates a Megatron-Core training backend into Tevatron while preserving its data pipeline, evaluation workflow, and Hugging Face-compatible checkpoints, and supports both LoRA and full-parameter fine-tuning.
Zhichao Xu, Xueguang Ma, Shengyao Zhuang et al.· 2 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
