News/Open-source
Gradio 5 security audit: What developers need to know
Gradio 5 Security Audit: Enhancing Safety in Machine Learning Web Applications: Gradio, a popular Python library for building machine learning web applications, has undergone a comprehensive security audit by Trail of Bits in preparation for its version 5 release, addressing critical vulnerabilities and strengthening its security posture. The rise of Gradio in ML app development: Gradio has become the go-to solution for creating machine learning web interfaces, boasting over 6 million monthly PyPI installs and powering more than 470,000 applications on Hugging Face Spaces. Gradio allows developers to quickly build and share ML applications with just a few lines of...
read Oct 12, 2024What to know about Meta AI
Meta AI: Democratizing artificial intelligence: Meta, the parent company of Facebook, Instagram, and WhatsApp, has introduced Meta AI, a free and widely accessible AI model designed to transform user interactions across its social media platforms. Meta AI is built on the open-source Llama 3.2 model and is integrated into Meta's ecosystem, including Ray-Ban smart glasses. The AI is available as a standalone chatbot at meta.ai and within Meta's social media platforms. It offers features such as voice mode, image editing, and access to user data. Technical foundation and accessibility: Meta AI leverages advanced machine learning algorithms and natural language processing...
read Oct 12, 2024Prime Intellect launches initiative to train open model with decentralized computing
Pioneering decentralized AI training: Prime Intellect is launching INTELLECT-1, a groundbreaking initiative to train a 10-billion-parameter AI model using decentralized computing resources. INTELLECT-1 builds upon Prime Intellect's previous OpenDiLoCo work, which implemented DeepMind's Distributed Low-Communication (DiLoCo) method for distributed AI training. The project aims to enable open-source, decentralized training of large AI models, challenging the current paradigm of centralized control in AI development. Key partners contributing computing power include Hugging Face, SemiAnalysis, and Arcee, among others. Prime Intellect has opened the platform for anyone to contribute their computing resources to the project. Technological advancements: The INTELLECT-1 project incorporates several algorithmic...
read Oct 11, 2024Researchers unveil Aria, a new multimodal open-source model
Introducing Aria: A groundbreaking open-source multimodal AI model: Researchers have unveiled Aria, an innovative open-source multimodal native mixture-of-experts model that demonstrates top-tier performance across a wide range of multimodal, language, and coding tasks. Key features and capabilities: Aria represents a significant advancement in multimodal AI, offering a powerful and versatile solution for integrating diverse types of information. The model boasts 3.9 billion activated parameters per visual token and 3.5 billion activated parameters per text token, enabling it to process and understand complex multimodal inputs effectively. Aria outperforms existing models like Pixtral-12B and Llama3.2-11B, and competes with the best proprietary models...
read Oct 9, 2024‘Open-source’ has an updated definition — here’s what it is for now
Open Source AI Definition reaches release candidate stage: The Open Source Initiative (OSI) has released a Release Candidate (RC1) version of the Open Source AI Definition, marking a significant milestone in defining open-source standards for artificial intelligence systems. The RC1 version incorporates extensive community feedback gathered through town hall meetings, forum discussions, and in-person conversations across multiple countries. This release focuses on refining the definition of the "preferred form to make modifications to a machine learning system," addressing key aspects of data sharing, code completeness, and legal considerations. Key updates in the Release Candidate: Data Information requirements: The definition now...
read Oct 9, 2024Democratization of AI, open-source and AI Auditing
The democratization of AI and risk mitigation: Balancing the aspiration for democratized AI with the need to mitigate risks involves careful consideration of accessibility, transparency, and responsible development practices. Open source AI models contribute to decentralizing power in AI development, enabling a wider range of voices to be heard and facilitating research on AI safety. However, open source models do not necessarily equate to broader accessibility, as running AI models still requires significant resources and technical understanding. Most people interact with AI through user-friendly interfaces like chatbots, which can also be used to create and distribute disinformation. Open source AI...
read Oct 9, 2024The Reflection 70B saga continues with release of training data report
The Reflection 70B controversy unfolds: The AI community has been embroiled in a debate surrounding the Reflection 70B language model, with claims of exceptional performance being met with skepticism and accusations of fraud. Hyperwrite AI's CEO Matt Shumer announced Reflection 70B on September 5, 2024, touting it as "the world's top open-source model" based on benchmark results. Third-party evaluators struggled to replicate the claimed results, leading to widespread doubt and accusations within the AI community. A post-mortem reveals critical oversights: Sahil Chaudhary, founder of Glaive AI, whose data was used to train Reflection 70B, released a comprehensive report addressing the...
read Oct 8, 2024How Dynamiq differentiates itself in the crowded AI agent landscape
Dynamiq emerges as a comprehensive AI platform: Dynamiq positions itself as an end-to-end operating platform for generative AI, addressing the complex needs of modern businesses in their AI journey. Founded by Vitalii Duk, former engineering leader at Careem, Dynamiq aims to provide an enterprise-grade LLM and agent application platform. The platform takes a holistic approach, covering the entire AI development lifecycle from prototyping to deployment and fine-tuning. Key features and differentiators: Dynamiq's platform offers several unique features that set it apart in the crowded AI agent landscape. On-premise deployment capabilities ensure data privacy and security, crucial for industries dealing with...
read Oct 8, 2024SAP expands open source AI support, turns Joule into collaborative agent
SAP's AI evolution: Joule and open-source LLM integration: SAP has announced significant advancements in its generative AI capabilities, expanding the functionality of its AI copilot Joule and introducing new open-source large language model (LLM) support. Joule, SAP's generative AI copilot, can now support up to 80% of customers' most common business tasks, transforming it into a collaborative agent capable of handling complex workflows. The company is accelerating product innovation on the cloud to attract more customers to its RISE with SAP initiative, which aims to guide customers' transition from on-premises ERP systems to the cloud. SAP reported a 25% increase...
read Oct 8, 2024Hugging Face launches tool to let developers create AI apps in only minutes
Revolutionary AI development tool unveiled: Hugging Face's new "OpenAI-Gradio" Python package enables developers to create AI-powered web applications using OpenAI's language models with unprecedented ease and speed. Key features and benefits: The OpenAI-Gradio package combines OpenAI's powerful API with Gradio's user-friendly interface for machine learning applications, streamlining the development process significantly. Developers can launch functional web apps in minutes with minimal coding, requiring only installation of the package and setting of an OpenAI API key. The tool simplifies integration of OpenAI's large language models into web applications, eliminating the need for complex backend infrastructure management. It democratizes AI development, allowing...
read Oct 4, 2024Microsoft just released Drasi, and it could change how we handle big data
Microsoft unveils Drasi: A game-changer in data processing: Microsoft has launched Drasi, an open-source data processing system aimed at simplifying the detection and response to critical events in complex infrastructures, marking a significant advancement in cloud computing and event-driven architectures. The big picture: Drasi represents a new category of data processing systems, designed to address the growing complexity in event-driven architectures, particularly in scenarios like IoT edge deployments and smart building management. Mark Russinovich, CTO and Technical Fellow at Microsoft Azure, described Drasi as "the birth of a new category of data processing system" in an interview with VentureBeat. The...
read Oct 1, 2024Are community-trained AI models the future of LLM development?
The open source AI revolution: Nous Research, a pioneering organization in open source AI, is spearheading efforts to democratize AI model training and development through innovative projects like DisTrO. Nous Research, led by Bowen Peng and Jeffrey Quesnelle, is focused on accelerating open source AI research and empowering independent builders in the AI community. The organization's latest project, DisTrO, demonstrates the feasibility of training AI models across the public internet at unprecedented speeds. Nous Research is also behind other successful open source AI initiatives, including the Hermes family of "neutral" and guardrail-free language models. The DisTrO project: Addressing potential setbacks...
read Oct 1, 2024PyTorch’s releases ‘torchao’ to boost AI model performance
PyTorch Introduces torchao: Boosting Model Performance with Advanced Optimization Techniques: PyTorch has officially launched torchao, a native library designed to enhance model speed and reduce size through low-bit data types, quantization, and sparsity, offering significant improvements for both inference and training workflows. Key features and performance gains: torchao provides a toolkit of optimization techniques written primarily in PyTorch code, making it accessible and easy to implement. The library has been benchmarked on popular GenAI models, including LLama 3 and Diffusion models, with minimal accuracy loss. Impressive results for LLama 3: 97% speedup for LLama 3 8B inference using autoquant with...
read Sep 30, 2024AI-powered notebook rival built in 24 hours challenges Google
Open-source AI challenges Google's NotebookLM: A data scientist in Singapore has created an open-source alternative to Google's NotebookLM, highlighting the growing capabilities of individual developers in the AI space. Rapid development and key features: Gabriel Chua, a data scientist at Singapore's GovTech agency, built "Open NotebookLM" in just one afternoon using publicly available AI models. The tool transforms PDF documents into personalized podcasts, mirroring a key feature of Google's NotebookLM. It utilizes Meta's Llama 3.1 405B language model and MeloTTS for voice synthesis. A user-friendly interface built with Gradio and hosted on Hugging Face Spaces makes the tool accessible to...
read Sep 29, 2024AMD releases AMD-135M, its first open-source small language model
AMD's Foray into Small Language Models: AMD has unveiled its first small language model (SLM), AMD-135M, marking a significant step in the company's artificial intelligence initiatives. AMD-135M is part of the Llama family of models and was trained from scratch on AMD Instinct™ MI250 accelerators. The model comes in two variants: AMD-Llama-135M for general use and AMD-Llama-135M-code, which is fine-tuned for code-related tasks. This release aligns with AMD's commitment to an open approach to AI, aiming to foster inclusive, ethical, and innovative technological progress. Training Process and Specifications: The development of AMD-135M involved substantial computational resources and time investment to...
read Sep 27, 2024New research shows bigger AI models not always better
Llama-3 models performance in medical AI: Unexpected results and implications: A recent study comparing various Llama-3 models in medical and healthcare AI domains has revealed surprising findings, challenging assumptions about model size and performance. The Llama-3.1 70B model outperformed the larger Llama-3.2 90B model, particularly in specialized tasks like MMLU College Biology and Professional Medicine. Unexpectedly, the Meta-Llama-3.2-90B Vision Instruct and Base models showed identical performance across all datasets, an unusual occurrence for instruction-tuned models. Detailed performance breakdown: The study evaluated models using datasets such as MMLU College Biology, Professional Medicine, and PubMedQA, providing insights into their capabilities in medical...
read Sep 27, 2024AI models on Hugging Face surge past 1 million milestone
AI model explosion on Hugging Face: Hugging Face, a leading AI hosting platform, has reached a significant milestone by surpassing 1 million AI model listings, showcasing the rapid expansion and diversification of the machine learning field. The platform, which began as a chatbot app in 2016, pivoted to become an open-source hub for AI models in 2020, now offering a wide array of tools for developers and researchers. Hugging Face hosts numerous high-profile AI models, including Llama, Gemma, Phi, Flux, Mistral, Starcoder, Qwen, Stable Diffusion, Grok, Whisper, Olmo, Command, Zephyr, OpenELM, Jamba, and Yi, along with 999,984 others. Customization driving...
read Sep 26, 2024Meta’s new Llama AI model can now see and run on your device
Llama 3.2 Introduces Multimodal and On-Device Models: Meta's latest update to its Llama language model series brings significant advancements in AI capabilities, including vision processing and compact on-device models. Key Features and Enhancements: The Llama 3.2 release incorporates new multimodal vision models and smaller language models optimized for on-device applications, expanding the versatility and accessibility of AI technologies. Two sizes of vision models (11B and 90B parameters) are now available, each with base and instruction-tuned variants, enabling the processing of both text and images in tandem. New 1B and 3B parameter text-only models have been introduced, designed specifically for on-device...
read Sep 25, 2024The best open-source AI model yet is purpose built for AI agents
Breakthrough in open-source AI: The Allen Institute for AI (Ai2) has unveiled Multimodal Open Language Model (Molmo), a groundbreaking open-source AI model that combines image interpretation and conversational abilities, potentially revolutionizing AI agent development. Key capabilities and features: Molmo represents a significant advancement in open-source AI technology, offering a range of functionalities that were previously limited to proprietary models. The model can interpret images and engage in chat-based conversations, making it suitable for a variety of AI agent applications. Molmo is designed to assist AI agents in performing complex tasks such as web browsing, file navigation, and document drafting. Unlike...
read Sep 25, 2024AI2’s New Small Open-Source Model Performs as Well as Big Ones
Groundbreaking open-source AI model challenges industry giants: The Allen Institute for Artificial Intelligence (Ai2) has unveiled Molmo, a family of open-source multimodal language models that rival the performance of proprietary models from leading tech companies. Ai2 claims its largest Molmo model, with 72 billion parameters, outperforms OpenAI's GPT-4o in tests measuring image, chart, and document understanding. A smaller Molmo model with just 7 billion parameters reportedly approaches the performance of OpenAI's state-of-the-art model, highlighting Ai2's efficient data collection and training methods. Key innovations in data curation and training: Molmo's impressive performance stems from a novel approach to data collection and...
read Sep 22, 2024Mistral, Google and the Spectrum of ‘Openness’ in AI Models
The rapid advancement of artificial intelligence has sparked debates about the transparency and accessibility of AI models, highlighting the need for a clearer understanding of openness in the field. Recent developments in AI openness: Google and Mistral AI have taken divergent approaches to releasing their AI models, showcasing the varying degrees of accessibility in the industry. Google's Gemini release was accompanied by significant publicity but offered limited testing options, primarily through integration with Bard. Mistral AI quietly shared a Magnet link to one of its models, allowing skilled users to download, use, and fine-tune the model without fanfare. The contrast...
read Sep 21, 2024Open Source Maintainers are Underpaid and Skeptical of AI Coding
The state of open source maintenance: A recent survey by Tidelift reveals significant challenges and evolving trends in the open source community, highlighting issues of compensation, security, and trust among project maintainers. The majority of open source project maintainers continue to work without financial compensation, with 60% of respondents identifying as unpaid hobbyists. This lack of remuneration persists despite maintainers dedicating more time to critical aspects of project management, particularly security. The survey indicates a growing emphasis on security, with maintainers now spending 11% of their time on security-related tasks, a significant increase from 4% in 2021. Security concerns and...
read Sep 18, 2024Open-Source Model LlamaCoder Will Create an Entire App for You
AI-powered app generation revolutionizes development: Together AI's LlamaCoder app, built on the Llama 3.1 405B model, allows users to create entire applications from simple prompts, showcasing the potential of open-source AI in software development. LlamaCoder, released just over a month ago, has already garnered significant attention with more than 2,000 GitHub stars and hundreds of repository clones. The app has been used to generate over 200,000 applications, ranging from quiz apps to pomodoro timers and budgeting tools. Hassan El Mghari, a software engineer leading developer relations at Together AI, highlights the app's success as a demonstration of Llama 3.1 405B's...
read Sep 17, 2024Mistral AI Introduces Free Tier, Reduced Prices to Encourage AI Prototyping
Mistral AI expands AI accessibility: Mistral AI, a leading artificial intelligence company, has announced significant updates to its product lineup, including free offerings, reduced pricing, and enhanced model capabilities. Free tier introduction: Mistral AI has launched a free tier on its serverless platform, la Plateforme, allowing developers to experiment with and prototype AI models at no cost. Users can upgrade to a commercial tier with full data isolation and higher rate limits as needed. The platform offers deployment options across various cloud providers or on the user's own infrastructure. Pricing overhaul: The company has implemented substantial price reductions across its...
read