This content originally appeared on HackerNoon and was authored by Large Models (dot tech)
Table of Links
3.1. Benefits scale with model size and 3.2. Faster inference
3.3. Learning global patterns with multi-byte prediction and 3.4. Searching for the optimal n
3.5. Training for multiple epochs and 3.6. Finetuning multi-token predictors
3.7. Multi-token prediction on natural language
4. Ablations on synthetic data and 4.1. Induction capability
5. Why does it work? Some speculation and 5.1. Lookahead reinforces choice points
5.2. Information-theoretic argument
7. Conclusion, Impact statement, Environmental impact, Acknowledgements, and References
A. Additional results on self-speculative decoding
E. Additional results on model scaling behavior
F. Details on CodeContests finetuning
G. Additional results on natural language benchmarks
H. Additional results on abstractive text summarization
I. Additional results on mathematical reasoning in natural language
J. Additional results on induction learning
K. Additional results on algorithmic reasoning
L. Additional intuitions on multi-token prediction
C. Training speeds
\
:::info This paper is available on arxiv under CC BY 4.0 DEED license.
:::
:::info Authors:
(1) Fabian Gloeckle, FAIR at Meta, CERMICS Ecole des Ponts ParisTech, and contributed equally;
(2) Badr Youbi IdrissiFAIR at Meta, LISN Université Paris-Saclay, and contributed equally;
(3) Baptiste Rozière, FAIR at Meta;
(4) David Lopez-Paz, FAIR at Meta and his the last author;
(5) Gabriel Synnaeve, FAIR at Meta and the last author.
:::
\
This content originally appeared on HackerNoon and was authored by Large Models (dot tech)
Large Models (dot tech) | Sciencx (2025-06-08T15:30:03+00:00) Training Time Comparison: Multi-Token vs. Next-Token Prediction. Retrieved from https://www.scien.cx/2025/06/08/training-time-comparison-multi-token-vs-next-token-prediction/
Please log in to upload a file.
There are no updates yet.
Click the Upload button above to add an update.