Tapered Language Models
Tapered language models allocate more parameters to earlier layers and fewer to later layers, improving performance without increasing total parameters or compute costs.
Hugging Face · Daily Papers
·Reza Bayat, Ali Behrouz
·
·▲ 3 upvotes
Este artigo está em destaque na seleção diária de papers do Hugging Face, curada pela comunidade de pesquisa em IA.
Autores: Reza Bayat, Ali Behrouz, Aaron Courville
- 3 upvotes da comunidade
- Temas: transformer, recurrent, memory-based variants, identical layers, parameter-efficient fine-tuning, perplexity
Resumo
Resumo original (em inglês), extraído do paper:
Tapered language models allocate more parameters to earlier layers and fewer to later layers, improving performance without increasing total parameters or compute costs.
