InfoQ Homepage AI, ML & Data Engineering Content on InfoQ
-
Google Uses AutoML to Discover More Efficient AI Training Algorithm
Researchers at Google have open-sourced EvoLved sIgn mOmeNtum (Lion), an optimization algorithm for training neural networks, which was discovered using an automated machine learning (AutoML) evolutionary algorithm. Models trained with Lion can achieve better accuracy on several benchmarks than models trained with other optimizers, while requiring fewer compute cycles to converge.
-
The Sentence Similarity Scenario in ML.NET Model Builder
Microsoft recently published information about adding the Sentence Similarity scenario in Model Builder. This scenario allows the training of custom sentence similarity models. Together with the addition of this scenario to the Model Builder, it is no longer necessary to install the Model Builder GPU extension.
-
Meta AI’s Large Language Model with 10x Fewer Parameters
Meta AI recently released a new large language model called Language Large Models Meta AI (LLaMA) that outperforms foundational models such as GPT-3 and is competitive with PaLM, despite having 10 times fewer parameters. LLaMA has better performance in language tasks such as natural questions, common-sense reasoning and mathematical reasoning.
-
Google AI Updates Universal Speech Model to Scale Automatic Speech Recognition beyond 100 Languages
Google AI has recently unveiled a new update for their Universal Speech Model (USM), to support the 1,000 Languages Initiative. The new model performs better than OpenAI Whisper for all segments of automation speech recognition.
-
ChatGPT Now Available in Preview on Microsoft’s Azure OpenAI Service
ChatGPT is now available in preview on Microsoft’s Azure OpenAI service allowing developers to integrate ChatGPT directly into a host of different enterprise and end-user applications using a token-based pricing system.
-
Microsoft Open-Sources Weather Forecasting Deep Learning Model ClimaX
Researchers from Microsoft's Autonomous Systems and Robotics Research group have open-sourced ClimaX, a deep learning foundation model for weather and climate modeling. ClimaX can be fine-tuned for a variety of prediction tasks and performs as well as or better than state-of-the-art models on several benchmarks.
-
Zero-Copy In-Memory Sharing of Large Distributed Data: V6d
Zero-copy and in-memory data manager Vineyard (v6d) is maintained as a CNCF sandbox project and provides distributed operators that can be utilized to share immutable data within or across cluster nodes. V6d is of interest particularly for deep network training on big (sharded) datasets such as large language and graph models.
-
High-Performance Computing for Researchers and Students with Amazon Lightsail for Research
AWS recently announced the general availability (GA) of Amazon Lightsail for Research, a new offering designed to enable researchers and students to easily create and manage high-performance CPU or GPU research computers on the cloud.
-
NVIDIA Open-Sources Robot Learning Framework Orbit
A team of researchers from NVIDIA, ETH Zurich, and the University of Toronto open-sourced Orbit, a simulation-based robot learning framework. Orbit includes wrappers for four learning libraries, a suite of benchmark tasks, and simulation for several robot platforms, as well as interfaces for deploying trained agents on physical robots.
-
OpenAI Unleashes ChatGPT and Whisper APIs for Next-Gen Language Capabilities
OpenAI has announced that it’s now letting third-party developers integrate ChatGPT and Whisper into their apps and services via API, offering access to AI-powered language and speech-to-text capabilities. As compared to using the company's current language approach, these APIs will make it easier for businesses to integrate ChatGPT and Whisper into their platforms.
-
Microsoft Launches New Cognitive Speech Services Features to Accelerate Language Learning
Microsoft recently launched new features for its Cognitive Speech Service to accelerate language learning with pronunciation assessment, new speech-to-text (STT) languages, and prebuilt and custom neural voice enhancements.
-
GitHub Enhanced Copilot with New AI Model and Security-Oriented Capabilities
GitHub Copilot adopted a new AI model which is both faster and more accurate than the previous one, says GitHub. Additionally, GitHub has started using AI to detect vulnerabilities in Copilot suggestion by blocking insecure coding patterns in real-time.
-
DeepMind Open-Sources AI Interpretability Research Tool Tracr
Researchers at DeepMind have open-sourced TRAnsformer Compiler for RASP (Tracr), a compiler that translates programs into neural network models. Tracr is intended for research in mechanistic interpretability of Transformer AI models such as GPT-3.
-
Hugging Face and AWS Join Forces to Democratize AI
Hugging Face, a top supplier of open-source machine learning tools, and AWS have joined together to increase the access to artificial intelligence (AI). Hugging Face's cutting-edge transformers and natural language processing (NLP) models will be made available to AWS customers as a result of the cooperation, making it simpler for them to develop and deploy AI applications.
-
Microsoft Wants to Use ChatGPT to Control Robots through Language
In a recent paper, researchers at Microsoft Autonomous Systems and Robotics Group showed how OpenAI's ChatGPT can be used for robotics applications, including how to design prompts and how to direct ChatGPT to use specific robotic libraries to program the task at hand.