Maxim Fishman, Brian Chmiel, Ron Banner, Daniel Soudry. Scaling FP8 training to trillion-token LLMs. In The Thirteenth International Conference on Learning Representations, ICLR 2025, Singapore, April 24-28, 2025. OpenReview.net, 2025. [doi]
@inproceedings{FishmanCBS25,
title = {Scaling FP8 training to trillion-token LLMs},
author = {Maxim Fishman and Brian Chmiel and Ron Banner and Daniel Soudry},
year = {2025},
url = {https://openreview.net/forum?id=E1EHO0imOb},
researchr = {https://researchr.org/publication/FishmanCBS25},
cites = {0},
citedby = {0},
booktitle = {The Thirteenth International Conference on Learning Representations, ICLR 2025, Singapore, April 24-28, 2025},
publisher = {OpenReview.net},
}