Quartet: Native FP4 Training Can Be Optimal for Large Language Models
Paper • 2505.14669 • Published • 79
Oliver Sieberling
OliverSieberling
·
AI & ML interests
None yet
Recent Activity
upvoted a paper 14 days ago
Variable-Width Transformers upvoted a paper 5 months ago
Quartet II: Accurate LLM Pre-Training in NVFP4 by Improved Unbiased Gradient EstimationOrganizations
authored a
paper about 1 year ago
authored 3
papers over 1 year ago
EvoPress: Towards Optimal Dynamic Model Compression via Evolutionary Search
Paper • 2410.14649 • Published • 8
Plus Strategies are Exponentially Slower for Planted Optima of Random Height
Paper • 2404.09687 • Published
Hardest Monotone Functions for Evolutionary Algorithms
Paper • 2311.07438 • Published