paper-with-me

Papers

Dynamic Layer Tying for Parameter-Efficient Transformers

2024-01-23 · Tamir David Hay, Lior Wolf

In the pursuit of reducing the number of trainable parameters in deep transformer networks, we employ Reinforcement Learning to dynamically select layers during training and tie them together. Every few iterations, the RL agent is asked whether to train each layer $i$ independently or to copy the weights of a previous layer $j<i$. This facilitates weight sharing, reduces the number of trainable parameters, and also serves as an effective regularization technique. Experimental evaluations validate that our model modestly outperforms the baseline transformer model with regard to perplexity and drastically reduces the number of trainable parameters. In particular, the memory consumption during training is up to one order of magnitude less than the conventional training method.

📄 PDF Abstract BibTeX arXiv:2401.12819

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Relaxed Recursive Transformers: Effective Parameter Sharing with Layer-wise LoRA

2024-10-28 · Sangmin Bae, Adam Fisch, Hrayr Harutyunyan, Ziwei Ji 외

Large language models (LLMs) are expensive to deploy. Parameter sharing offers a possible path towards reducing their size and cost, but its effectiveness in modern LLMs remains fairly limited. In this work, we revisit "…

Knowledge Distillation

Tying the Loop -- Tied Expert Layers in Mixture-of-Experts Language Models

2026-06-15 · Martin Jaggi arxiv

Mixture-of-Experts (MoE) architectures efficiently scale Large Language Models (LLMs) by activating only a small fraction of their experts per token, yet the full parameter count - dominated by the expert parameters - mu…

Beyond Weight Tying: Learning Joint Input-Output Embeddings for Neural Machine Translation

2018-08-31 · WS 2018 10 · Nikolaos Pappas, Lesly Miculicich Werlen, James Henderson

Tying the weights of the target word embeddings with the target word classifiers of neural machine translation models leads to faster training and often to better translation quality. Given the success of this parameter …

DecoderMachine TranslationTranslationWord Embeddings

Improving Recursive Transformers with Mixture of LoRAs

2025-12-14 · Mohammadmahdi Nouriborji, Morteza Rohanian, Omid Rohanian arxiv

Parameter sharing in recursive transformers reduces model size but collapses layer-wise expressivity. We propose Mixture of LoRAs (MoL), a lightweight conditional-computation mechanism that inserts Low-Rank Adaptation (L…

Hybrid Perception and Equivariant Diffusion for Robust Multi-Node Rebar Tying

2025-08-26 · Zhitao Wang, Yirong Xiong, Roberto Horowitz, Yanke Wang 외 arxiv

Rebar tying is a repetitive but critical task in reinforced concrete construction, typically performed manually at considerable ergonomic risk. Recent advances in robotic manipulation hold the potential to automate the t…

Collision AvoidanceMotion Planning