AI TRAINING
SkyRL Adds Vision-Language RL Support for Multimodal Models
SkyRL introduces vision-language reinforcement learning, enabling scalable training for multimodal tasks. Learn how this impacts AI development.
Google's Decoupled DiLoCo Redefines Distributed AI Training
Google's Decoupled DiLoCo architecture enables faster, resilient AI training across data centers, leveraging mixed-generation hardware for efficiency.
NVIDIA Megatron Boosts LLM Training With Muon Optimizer
NVIDIA integrates Muon and advanced optimizers into Megatron to enhance large-scale LLM training with near-parity throughput to AdamW.
NVIDIA NeMo RL Achieves 48% Speedup with End-to-End FP8 Precision Training
NVIDIA's new FP8 recipe for reinforcement learning delivers 48% faster training while matching BF16 accuracy, cutting AI infrastructure costs significantly.
NVIDIA nvCOMP Cuts AI Training Checkpoint Costs by $56K Monthly
New GPU compression library reduces LLM training checkpoint sizes by 25-40%, saving teams up to $222K monthly on large-scale model training infrastructure.
Mistral AI Launches Forge for Enterprise Custom AI Model Training
Mistral AI debuts Forge platform enabling enterprises to train frontier-grade AI models on proprietary data. ASML, Ericsson, ESA among launch partners.
OpenAI Launches €500K Grant and SME Training Program in EU Push
OpenAI's EU Economic Blueprint 2.0 targets 20,000 SMEs for AI training across six countries, backed by Booking.com partnership and youth safety research funding.
NVIDIA NVFP4 Training Delivers 1.59x Speed Boost Without Accuracy Loss
NVIDIA's NVFP4 4-bit training format achieves 59% faster AI model training than BF16 while matching accuracy on Llama 3 8B benchmarks, per new research.
Google Launches AI Professional Certificate to Close Workforce Skills Gap
Google's new AI Professional Certificate targets the 86% of workers lacking AI training, with Walmart, Deloitte, and Verizon already adopting the program.
NVIDIA Isaac Lab Hits 150K FPS for Robot Training at Scale
NVIDIA's Isaac Lab framework achieves 135,000+ FPS for humanoid training, attracting Agility Robotics and Skild AI as enterprise adopters.
NVIDIA Releases Open Source Tools for License-Safe AI Model Training
NVIDIA's NeMo Data Designer enables developers to build synthetic data pipelines for AI distillation without licensing headaches or massive datasets.
NVIDIA Achieves 36% Training Speedup for 256K Token AI Models
NVIDIA's NVSHMEM integration with XLA compiler delivers up to 36% faster training for long-context LLMs, enabling efficient 256K token sequence processing on JAX.
NVIDIA Hybrid-EP Slashes MoE AI Training Communication Overhead by 14%
NVIDIA's new Hybrid-EP communication library achieves up to 14% faster training for DeepSeek-V3 and other MoE models on Grace Blackwell hardware.
NVIDIA Blackwell Leads MLPerf Training v5.1 with Record-Breaking Performance
NVIDIA's Blackwell architecture achieves top performance across all MLPerf Training v5.1 benchmarks, highlighting advancements in AI training efficiency and precision.
NVIDIA Dominates MLPerf Training v5.1 with Blackwell Ultra GPUs
NVIDIA swept the MLPerf Training v5.1 benchmarks, showcasing superior AI training performance with its Blackwell Ultra GPU architecture across multiple AI model categories.