Efficient-LLMs-Survey

Efficient-LLMs-Survey copied to clipboard

Suggest for incorporating a speculative decoding paper

2

Thanks for your great work! I wanted to bring to your attention our recent work [PEARL](https://arxiv.org/pdf/2408.11850), a parallel sepculative decoding framework to achieve adaptive draft length. It has shown significant...

smart-lty

This Paper: Perplexed by Perplexity: Perplexity-Based Data Pruning With Small Reference Models, arXiv, 2024 [[Paper](https://arxiv.org/abs/2405.20541)] I believe is using small model to pruning dataset in order to have a better...

Michael-jze