Machine Learning
Golden Gemini Revolutionizes Speech AI with Enhanced Efficiency
Golden Gemini introduces a novel method in Speech AI, improving accuracy and reducing computational needs by addressing fundamental flaws in traditional speech processing models.
NVIDIA Enhances AI Inference with Full-Stack Solutions
NVIDIA introduces full-stack solutions to optimize AI inference, enhancing performance, scalability, and efficiency with innovations like the Triton Inference Server and TensorRT-LLM.
Sei Blockchain Empowers Secure ML Data Pipelines with Props
Sei blockchain introduces 'props' to enhance machine learning data security and authenticity, enabling ethical use of private data for AI models.
NVIDIA Hackathon Showcases Strategies for RAPIDS-Accelerated Machine Learning
The NVIDIA Hackathon at ODSC West highlighted innovative strategies using RAPIDS for accelerated machine learning, with participants developing efficient models under time constraints.
Enhancing GPU Analytics with RAPIDS and Ray Integration
Explore how RAPIDS and Ray can accelerate GPU analytics, focusing on Ray's Actor model and integration with RAPIDS libraries for optimized data processing.
NVIDIA NeMo-Aligner Enhances Supervised Fine-Tuning with Data-Efficient Knowledge Distillation
NVIDIA NeMo-Aligner introduces a data-efficient approach to knowledge distillation for supervised fine-tuning, enhancing performance and efficiency in neural models.
Enhancing Action Recognition Models Using Synthetic Data
NVIDIA explores the use of synthetic data to improve action recognition models, highlighting the benefits and applications across industries such as retail and healthcare.
Together AI Enhances Fine-Tuning API with Long-Context and Conversational Data Support
Together AI introduces significant updates to its Fine-Tuning API, including long-context training, conversation data support, and improved configuration options for better model customization and performance.
Accelerating Causal Inference with NVIDIA RAPIDS and cuML
Discover how NVIDIA RAPIDS and cuML enhance causal inference by leveraging GPU acceleration for large datasets, offering significant speed improvements over traditional CPU-based methods.
IBM's LoRA Technique Revolutionizes AI Model Specialization
IBM Research explores how Low-Rank Adaptation (LoRA) can quickly and efficiently specialize AI models, offering a cost-effective and flexible alternative to traditional fine-tuning methods.