DynaMo: Accelerating Language Model Inference with Dynamic Multi-Token Sampling

Shikhar Tuli, Chi-Heng Lin, Yen-Chang Hsu, Niraj Jha, Yilin Shen, Hongxia Jin


Abstract
Traditional language models operate autoregressively, i.e., they predict one token at a time. Rapid explosion in model sizes has resulted in high inference times. In this work, we propose DynaMo, a suite of multi-token prediction language models that reduce net inference times. Our models *dynamically* predict multiple tokens based on their confidence in the predicted joint probability distribution. We propose a lightweighttechnique to train these models, leveraging the weights of traditional autoregressive counterparts. Moreover, we propose novel ways to enhance the estimated joint probability to improve text generation quality, namely co-occurrence weighted masking and adaptive thresholding. We also propose systematic qualitative and quantitative methods to rigorously test the quality of generated text for non-autoregressive generation. One of the models in our suite, DynaMo-7.3B-T3, achieves same-quality generated text as the baseline (Pythia-6.9B) while achieving 2.57× speed-up with only 5.87% and 2.67% parameter and training time overheads, respectively.
Anthology ID:
2024.naacl-long.182
Volume:
Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)
Month:
June
Year:
2024
Address:
Mexico City, Mexico
Editors:
Kevin Duh, Helena Gomez, Steven Bethard
Venue:
NAACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
3322–3345
Language:
URL:
https://aclanthology.org/2024.naacl-long.182
DOI:
10.18653/v1/2024.naacl-long.182
Bibkey:
Cite (ACL):
Shikhar Tuli, Chi-Heng Lin, Yen-Chang Hsu, Niraj Jha, Yilin Shen, and Hongxia Jin. 2024. DynaMo: Accelerating Language Model Inference with Dynamic Multi-Token Sampling. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), pages 3322–3345, Mexico City, Mexico. Association for Computational Linguistics.
Cite (Informal):
DynaMo: Accelerating Language Model Inference with Dynamic Multi-Token Sampling (Tuli et al., NAACL 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.naacl-long.182.pdf