Search Results for "llm"
NVIDIA's NeMo Framework Enables Weekend Training of Reasoning-Capable LLMs
NVIDIA introduces an efficient method to train reasoning-capable language models over a weekend using the NeMo framework, leveraging the Llama Nemotron dataset and LoRA adapters.
NVIDIA's ProRL v2 Advances LLM Reinforcement Learning with Extended Training
NVIDIA unveils ProRL v2, a significant leap in reinforcement learning for large language models (LLMs), enhancing performance through extended training and innovative algorithms.
Enhancing LLM Inference with CPU-GPU Memory Sharing
NVIDIA introduces a unified memory architecture to optimize large language model inference, addressing memory constraints and improving performance.
NVIDIA's Run:ai Model Streamer Enhances LLM Inference Speed
NVIDIA introduces the Run:ai Model Streamer, significantly reducing cold start latency for large language models in GPU environments, enhancing user experience and scalability.
Enhancing LLM Inference with NVIDIA Run:ai and Dynamo Integration
NVIDIA's Run:ai v2.23 integrates with Dynamo to address large language model inference challenges, offering gang scheduling and topology-aware placement for efficient, scalable deployments.
ATLAS: Revolutionizing LLM Inference with Adaptive Learning
Together.ai introduces ATLAS, a system enhancing LLM inference speed by adapting to workloads, achieving 500 TPS on DeepSeek-V3.1.
Unsloth Simplifies LLM Training on NVIDIA Blackwell GPUs
Unsloth's open-source framework enables efficient LLM training on NVIDIA Blackwell GPUs, democratizing AI development with faster throughput and reduced VRAM usage.
AutoJudge Revolutionizes LLM Inference with Enhanced Token Processing
AutoJudge introduces a novel method to accelerate large language model inference by optimizing token processing, reducing human annotation needs, and improving processing speed with minimal accuracy loss.
NVIDIA's Breakthrough in LLM Memory: Test-Time Training for Enhanced Context Learning
NVIDIA introduces a novel approach to LLM memory using Test-Time Training (TTT-E2E), offering efficient long-context processing with reduced latency and loss, paving the way for future AI advancements.
Open-Source AI Judges Beat GPT-5.2 at 15x Lower Cost Using DPO Fine-Tuning
Together AI demonstrates fine-tuned open-source LLMs can outperform GPT-5.2 as evaluation judges using just 5,400 preference pairs, slashing costs dramatically.
Is Conversational Diagnostic AI like AMIE Feasible?
AMIE, an AI system developed by Google Research and DeepMind, demonstrates superior diagnostic accuracy compared to human physicians in a groundbreaking study, signaling a new era in AI-driven healthcare.
Deceptive AI: The Hidden Dangers of LLM Backdoors
Recent studies reveal large language models can deceive, challenging AI safety training methods. They can hide dangerous behaviors, creating false safety impressions, necessitating the development of robust protocols.