Training-method researcherSystems researcher

Kan Wu

Worked on FP8-LM / FP8 training, showing large models can be trained in 8-bit floating point to cut compute and memory.

Works in the archive (1)

“Lead” marks the first-listed author — a byline fact, not a measure of influence; co-authorship is shown as recorded.

Co-authors in the archive

Houwen Peng, Han Hu, Peng Cheng, Paulius Micikevicius

View co-author network (5 people)

Loading graph…