Aguru

𝐂𝐚𝐬𝐞 𝐒𝐭𝐮𝐝𝐲 𝐨𝐧 𝐋𝐋𝐌 𝐀𝐩𝐩𝐥𝐢𝐜𝐚𝐭𝐢𝐨𝐧𝐬: 𝐑𝐞𝐧𝐨𝐯𝐚𝐭𝐢𝐧𝐠 𝐏𝐮𝐛𝐥𝐢𝐜 𝐒𝐜𝐡𝐨𝐨𝐥 𝐈𝐧𝐟𝐫𝐚𝐬𝐭𝐫𝐮𝐜𝐭𝐮𝐫𝐞 𝐢𝐧 𝐅𝐫𝐚𝐧𝐜𝐞. 🏦 Banque des Territoires, a French public financial institution, has launched EduRénov, a program aiming to support and finance the energy renovation of 10,000 public school buildings with 2 billion euros in loans. The goal is to achieve 40% energy savings in 5 years as part of France's national strategy for energy transformation. ❓ A typical question from a local community official might be, "𝐼𝑛 2025, 𝑜𝑢𝑟 𝑐𝑖𝑡𝑦 𝑤𝑜𝑢𝑙𝑑 𝑙𝑖𝑘𝑒 𝑡𝑜 𝑖𝑛𝑖𝑡𝑖𝑎𝑡𝑒 𝑡ℎ𝑒 𝑟𝑒𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 𝑖𝑡𝑠 𝑠𝑐ℎ𝑜𝑜𝑙. 𝑊ℎ𝑎𝑡 𝑡𝑦𝑝𝑒 𝑜𝑓 𝑙𝑜𝑎𝑛𝑠 𝑐𝑎𝑛 𝑤𝑒 𝑔𝑒𝑡 𝑡𝑜 𝑓𝑢𝑛𝑑 𝑡ℎ𝑒 𝑝𝑟𝑜𝑗𝑒𝑐𝑡?". Now, generative AI can provide accurate answers to such queries. 🤝 Indeed, Hugging Face collaborates with Banque des Territoires and Polyconseil on an AI solution that facilitates communication between local officials and Banque des Territoires representatives. This solution aims to increase the number of initiated renovation projects. 📖 Read our blog post to learn more about how open-source LLMs, RAG, and sovereign cloud solutions are used in this initiative: https://lnkd.in/e-9FPPqq 👨🏾🏫 I am proud to contribute to this project through Hugging Face’s Expert Support Program (https://lnkd.in/eNgDgTeQ) and share expertise in ML solutions development with Polyconseil and Banque des Territoires. 🎉 Kudos to Henri Jouhaud , Anthony Truchet, Jeremy Cailton, Emma de Corbière, Johnny CHEN, Sébastien Ehling, Amaury Dreux, Nicolas T., Hakim Lahlou, Baptiste Bontoux, Marie Lattelais, Jean-Christophe Bernigaud, David Buchner and 🤗 Adrien Dufayard 🤗!

I would like to share a fantastic visual representation of the #AI Universe. This diagram beautifully encapsulates the intricate ecosystem of AI and its various subfields. AI: At the outermost layer, we have AI, the broadest and most encompassing term. AI refers to machines and systems designed to perform tasks that typically require human intelligence. Some of these tasks include: Natural Language Processing: Enabling machines to understand and respond to human language. Computer Vision: Allowing machines to interpret and process visual data. Knowledge Representation: Storing information about the world in a form that a computer system can utilize. AI Ethics: Ensuring AI systems are developed and used responsibly. Cognitive Computing: Simulating human thought processes in a computerized model. Machine Learning (ML): Moving one layer in, we find ML. This subset of AI involves systems that learn from data to make decisions and predictions. Key concepts include: Dimensionality Reduction: Simplifying data without losing significant information. Unsupervised Learning: Finding patterns in data without pre-labeled outcomes. Reinforcement Learning: Learning optimal actions through trial and error. Ensemble Learning: Combining multiple models to improve performance. Neural Networks: Delving deeper, we encounter Neural Networks, which are inspired by the human brain's structure. These are essential for many advanced AI capabilities. Components include: Perceptrons: The simplest type of neural network. Convolutional Neural Networks: Specialize in processing visual data. Recurrent Neural Networks: Handle sequential data, like time series. Multi-Layer Perceptrons: Networks with multiple layers between input and output. Activation Functions: Functions that determine the output of a neural network. Backpropagation: The method for training neural networks. Deep Learning: Within neural networks, we have the realm of Deep Learning. This subset involves networks with many layers (hence "deep") and includes: Deep Neural Networks: Networks with multiple hidden layers. Generative Adversarial Networks: Networks that generate new data similar to the input data. Deep Reinforcement Learning: Combining deep learning with reinforcement learning. Generative AI: At the core, we find Generative AI, which is about creating new content. This includes: Language Modeling: Predicting the next word in a sequence. Transformer Architecture: A model that handles sequential data efficiently, crucial for NLP. Self-Attention Mechanism: Allows models to focus on different parts of the input sequence. Natural Language Understanding: Comprehending and generating human language. Dialogue Systems: AI systems that can converse with humans. Transfer Learning: Using knowledge from one task to improve performance on another. By understanding these layers, we gain insight into the capabilities and potential of AI technologies. #machinelearning #python #datascience

Just published - our brand new 2024 Hype Cycle for APIs! - AI, of course, is a big factor this year. AI Gateways are new on the Hype Cycle, both from specialists like Aguru and Portkey, and from established API Gateway vendors like Kong Inc. and IBM. It will be very interesting to see how this particular market plays out. We also include AI used to create APIs, and AI used to consume APIs, on this year's Hype Cycle. - API security is well represented on the Hype Cycle. This year that includes Composable security APIs, including vendors like Pangea and Skyflow, which is climbing up the Hype Cycle - API aggregators, such as Knit and Merge, continue to gain momentum - GraphQL APIs are headed towards the Trough of Disillusionment, where Service Mesh still sits. - Many industry API initiatives now feature on the Hype Cycle, with finance furthest along, then industries like insurance and healthcare, and finally supply chain APIs which are most nascent. Check out the Hype Cycle for APIs here: https://lnkd.in/ejQNTQ3h . And a big thanks to my co-author John Santoro!

Open LLM Leaderboard 2 released! Evaluating LLMs is not easy. Finding new ways to compare LLM fairly, transparently, and reproducibly is important! Benchmarks are not perfect, but they give us a first understanding of how well models perform and where their strengths are. What's new?! 📈 New benchmarks with MMLU-Pro, GPQA, MuSR, MATH, IFEval and BBH. 📊 Improved ranking with normalized scores adjusted to baselines 🏆 Qwen2 72B Instruct > Meta Llama 3 70B Instruct > Cohere Command R+ ⚡ Faster, simpler Interface with a new Gradio component. 🛠️ Enhanced reproducibility with support for delta weights and chat templates ⭐ Introduction of "maintainer's highlight" and “community voting system” Leaderboard: https://lnkd.in/eeJP696h Blog: https://lnkd.in/esdDPWPi Kudos to the Hugging Face team, especially Clémentine Fourrier, Nathan HABIB, Alina Lozovskaya, Konrad Szafer, and Thomas Wolf.

If 2023 was the year the world discovered generative AI, 2024 is the year organizations began deriving real business value from it. According to our latest State of AI in early 2024 survey by QuantumBlack, a McKinsey company , 65% of respondents report regular use of gen AI, nearly double from ten months ago. Learn more about the significant benefits organizations are already seeing, including cost reductions and revenue increases in our new report ➡️ https://mck.co/4aGSIZw #AIbyMcKinsey #GenAI

If 2023 was the year the world discovered generative AI, 2024 is the year organizations began deriving real business value from it. According to our latest State of AI in early 2024 survey by QuantumBlack, a McKinsey company , 65% of respondents report regular use of gen AI, nearly double from ten months ago. Learn more about the significant benefits organizations are already seeing, including cost reductions and revenue increases in our new report ➡️ https://mck.co/4aGSIZw #AIbyMcKinsey #GenAI