Dual-Objective Language Models: Efficient Training, No Overfitting
Discover how dual-objective language models enhance training efficiency while preventing overfitting using autoregressive and masked-diffusion methods.
PathFinder: Enhanced Path Loss Prediction for Multi-Transmitters
Discover PathFinder, a novel model improving path loss prediction in multi-transmitter 5G scenarios with advanced environmental modeling.
SonicMoE: Boost MoE Efficiency with IO & Tile Optimizations
Discover how SonicMoE enhances Mixture of Experts models with IO and tile-aware optimizations, cutting memory use and boosting GPU training speed.
Nemotron-Cascade: Scalable Cascade RL for Reasoning Models
Discover Nemotron-Cascade, a scalable cascaded reinforcement learning model enhancing general-purpose reasoning with state-of-the-art performance and effic...
Particulate: Fast Feed-Forward 3D Object Articulation
Discover Particulate, a feed-forward model for rapid, accurate 3D object articulation using transformer networks and AI-generated assets.