Research & Education

BertViz is an interactive, open-source tool designed for visualizing attention mechanisms within Transformer language models. It can be seamlessly integrated and run inside Jupyter or Colab notebooks through a simple Python API, offering compatibility with most Huggingface models. BertViz extends the functionality of the Tensor2Tensor visualization tool by Llion Jones, providing multiple distinct views: the head view for single or multiple attention heads, the model view for an overview across all layers and heads, and the neuron view for visualizing individual neurons in query and key vectors. This tool is invaluable for researchers and developers seeking to understand and interpret the complex internal workings of Transformer models.

awesome-self-supervised-gnn is a comprehensive repository featuring a curated list of academic papers focused on self-supervised learning within the domain of Graph Neural Networks (GNNs). The collection is meticulously organized by publication year, providing a structured overview of advancements in the field. This resource is invaluable for researchers, academics, and practitioners who need to explore, understand, and implement the latest self-supervised learning techniques for GNNs. It helps users quickly identify influential papers, indicated by a '🔥' for highly cited works, and often includes direct links to both the paper and its associated code, facilitating deeper engagement with the research.

ai-cookbook is an open-source resource designed for developers seeking to build AI systems. It offers a collection of examples and tutorials, complete with copy/paste code snippets, to facilitate easy integration into various projects. The cookbook is maintained by Dave Ebbelaar, an AI engineer and founder of Datalumina, who also shares practical tutorials on his YouTube channel. Beyond the cookbook, Dave provides additional resources including a free five-hour Python course for AI beginners and a program for developers looking to build and deploy end-to-end GenAI solutions. This tool is ideal for those who want practical, real-world applications of AI development.

Diffusion-Probabilistic-Models offers a reference implementation for deep unsupervised learning, specifically focusing on methods described in the paper "Deep Unsupervised Learning using Nonequilibrium Thermodynamics." This tool enables the construction of generative models by training a Gaussian diffusion process to transform a noise distribution into a data distribution over a fixed number of time steps. The mean and covariance of this diffusion process are parameterized using deep supervised learning. The resulting models are designed to be tractable for training, easy to sample from, allow for efficient probability evaluation of data points, and facilitate straightforward computation of conditional and posterior distributions. It includes features for training on datasets like MNIST and CIFAR-10, outputting objective function values, and generating samples, inpaintings, and denoised images.

The deep-reinforcement-learning repository serves as a comprehensive resource for Udacity's Deep Reinforcement Learning Nanodegree program. It offers a collection of tutorials that guide users through the implementation of various reinforcement learning algorithms, including Dynamic Programming, Monte Carlo, Temporal-Difference, Deep Q-Networks, and Deep Deterministic Policy Gradients, all utilizing PyTorch. Additionally, the repository features labs and projects that leverage rich simulation environments from Unity ML-Agents, allowing users to train agents in tasks like navigation, continuous control, and collaborative competition. It also includes benchmarks for classic control and Box2d environments, along with detailed setup instructions for the Python environment.

AI Hub Frankfurt is dedicated to unlocking the potential of data and AI for businesses within the Frankfurt and Rhine-Main region. It serves as a central hub for AI events, consulting, training, and innovation, aiming to strengthen the region's competitiveness. The initiative focuses on the practical application of AI in economic environments, supporting companies in generating value from data and AI through various offerings. It emphasizes collaboration, bringing together diverse businesses and initiatives to facilitate knowledge transfer and foster a strong local AI community.

dmls-book is a GitHub repository offering summaries and resources for the 'Designing Machine Learning Systems' book by Chip Huyen (O'Reilly 2022). This resource focuses on the key design decisions for developing and deploying machine learning systems, emphasizing reliability, scalability, maintainability, and adaptability. It is not a tutorial book with extensive code snippets but provides valuable content such as chapter summaries, MLOps tools, and a review of basic ML concepts. The book itself has been translated into over 10 languages and is available on major platforms like Amazon and O'Reilly. The repository is ideal for ML engineers, data scientists, data engineers, ML platform engineers, and engineering managers looking to leverage ML to solve real-world problems at scale.

Opeton is an AI language tutor designed to help users achieve fluency in a second language through conversational practice. Instead of traditional grammar drills or flashcards, Opeton focuses on real-time speaking sessions with an AI tutor that adapts to the user's proficiency level. These sessions can range from 2 to 30 minutes, covering various topics from casual chats to debates. The platform offers a judgment-free environment where users can make mistakes and even switch to English if needed, ensuring a challenging yet never overwhelming learning experience. Opeton tracks conversation progress and automatically adjusts topics and difficulty, making every session fresh and engaging. It's available on both the App Store and Google Play, with the first call offered for free.

papers-notebook is an open-source GitHub repository dedicated to compiling reading notes on academic papers, primarily in the fields of distributed systems, virtualization, and machine learning. The project aims to document the author's thoughts on papers, including their core ideas, implementation methods, and personal evaluations, with each note ideally under 1000 words. While initially maintained as Markdown files, the project has transitioned to using GitHub issues for ongoing maintenance and welcomes community contributions for paper suggestions. The repository covers a wide range of topics within its focus areas, including schedulers like Mesos and Borg, consensus algorithms like Raft and Paxos, storage systems, virtualization technologies, and security aspects.

Open Source Ai Year In Review 2024 is an interactive web application hosted on Hugging Face, designed to summarize the key developments in open-source AI throughout 2024. It presents a calendar of 25 cards, each representing a significant day of AI highlights. Users can click on any card to access a pop-up window displaying detailed articles, insightful charts, and embedded visualizations related to that specific highlight. This tool serves as a comprehensive resource for anyone interested in understanding the trajectory and future of open-source artificial intelligence, offering a curated overview of the year's most impactful events and innovations.

Guide.AI is an innovative AI audio guide generator designed to help users create, publish, and monetize audio guides effortlessly. The platform leverages advanced AI text-to-speech technology, eliminating the need for manual audio recordings or professional translation services. This enables quick and easy creation of guides in multiple languages, boosting accessibility, inclusivity, and engagement for various audiences. Guide.AI aims to provide a free solution for content creators to generate passive income, making it an environmentally friendly and sustainable option for enhancing visitor experiences in sectors like travel, history, and culture.

Next Server Example App is an AI application hosted on Hugging Face, designed for generating responses based on user-provided text input. This tool offers a straightforward interface where users can type their input into a text box and receive a corresponding output. It serves as a practical example of a server-side AI application, demonstrating basic text generation capabilities. The application is suitable for those looking to experiment with AI text generation or understand the fundamental interaction between user input and AI-generated responses in a web environment. Its simplicity makes it accessible for quick demonstrations and testing.

TRAIL - TRusted AI Labs aims to mobilize AI research and innovation capabilities in Walloon and Brussels regions for socio-economic development. It acts as a bridge between academic research and industrial needs, facilitating the transfer of expertise and tools developed by universities and accredited research centers. TRAIL organizes workshops, events, and initiatives to engage industrial players, researchers, and international stakeholders, addressing grand challenges in AI. The organization focuses on creating a trusted AI ecosystem, promoting ethical and impactful AI solutions through public and private funded projects.

Transformer-in-Computer-Vision is a comprehensive and regularly updated paper list focusing on recent Transformer-based works in the field of Computer Vision. This GitHub repository serves as a valuable resource for researchers, academics, and students interested in the latest advancements in this rapidly evolving area. The list is meticulously organized by various computer vision tasks, including classification, detection, segmentation, generative models, and more, making it easy to navigate and find relevant papers. Each entry, where available, includes links to the paper and its corresponding code implementation. Users are encouraged to contribute by opening issues or pull requests for any overlooked papers, fostering a collaborative environment for knowledge sharing in the CV community.

PTE APEUni is a comprehensive, free platform designed to help students prepare for the PTE Academic and PTE Core exams. It offers advanced AI scoring for various sections, including speaking (Read Aloud, Repeat Sentence, Describe Image, Re-tell Lecture, Answer Short Question) and writing (Summarize Written Text, Write Essay). Users can practice with real AI scores, which evaluate pronunciation, fluency, grammar, and spelling. The platform synchronizes practice records across web and app versions, allowing for flexible study. Additionally, PTE APEUni provides study materials, a vocabulary book with 90% exam vocabs, a shadowing feature to improve pronunciation, and AI-powered score report analysis. It also includes weekly predictions and study guides for all PTE sections.

Vowpal Wabbit is an open-source machine learning system designed for advanced online learning. It incorporates techniques like hashing, allreduce, reductions, learning2search, active, and interactive learning. A key focus is on reinforcement learning, offering several contextual bandit algorithms. The system is built for performance, with a specific emphasis on speed and scalability, ensuring its memory footprint remains bounded regardless of data size. It supports flexible input formats, including free-form text features with multiple namespaces, and allows for feature interaction to optimize ranking problems. Vowpal Wabbit is a destination for implementing and maturing state-of-the-art algorithms efficiently.

vsepp is an open-source PyTorch implementation for enhancing visual-semantic embeddings, specifically designed for image-caption retrieval tasks. It provides the code for methods detailed in the paper "VSE++: Improving Visual-Semantic Embeddings with Hard Negatives" presented at BMVC 2018. The repository includes scripts for evaluation of pre-trained models and training new models, with options for different arguments like `max_violation` and `measure order`. It supports Python 2.7 (with a Python 3 branch available) and PyTorch, along with other dependencies like NumPy and TensorBoard. The project also provides instructions for downloading datasets and pre-trained models, making it a valuable resource for researchers and developers working on visual-semantic embedding problems.

Huggingface Model leaderboard is a comprehensive AI search engine designed to help users navigate the vast ecosystem of models available on Hugging Face. This tool enables users to easily browse and discover trending models, offering a streamlined way to find specific AI models by searching through names, owners, or tags. Beyond simple discovery, the leaderboard provides detailed information and statistics on top model creators, giving users insights into the performance and popularity of various models. It serves as an invaluable resource for researchers, developers, and enthusiasts looking to stay updated on the latest advancements and identify suitable models for their projects.

Latent Navigation is an AI tool hosted on Hugging Face Spaces, designed to help users explore and visualize the latent space of a model. By providing a text prompt and two contrasting concepts (e.g., "winter" and "summer"), the application computes a directional path within the text-image space. It then generates a sequence of images that smoothly transition from one concept to the other, illustrating the model's understanding and representation of these ideas. This tool is particularly useful for researchers and engineers seeking to understand how AI models interpret and connect different data points in their internal representations. The Space is currently paused, requiring users to request a restart from the author.

Time Machine is an enterprise AI sales training platform designed to significantly reduce sales onboarding time and improve performance. It leverages DARPA-proven AI technology to create personalized learning paths and offers unlimited AI role-play practice, allowing sales reps to master complex products and sales scenarios quickly. The platform integrates with existing content, transforming it into optimal learning modules, and provides real-time analytics to track individual and team progress. Time Machine is built for scale, offering solutions for startups, mid-market, and large enterprises, and ensures enterprise-grade security. It aims to democratize world-class sales training, making advanced AI accessible to organizations looking to accelerate pipeline and increase quota attainment.

Multimodal VLM Thinking is a Hugging Face Space designed for AI research, enabling users to interact with various vision-language models (VLMs). Users can upload an image, input a question or instruction, and select from models like Lumian-VLR, VisionThink, MiniCPM-V, Typhoon-OCR, or olmOCR to process the request. The application provides written responses, capable of describing image content, extracting text via OCR, or performing other image-based reasoning tasks. This tool is particularly useful for researchers and engineers focused on advancing AI capabilities in understanding and processing both visual and textual information.

Multiview Diffusion 3d is an AI tool hosted on Hugging Face Spaces that enables users to generate multiple perspectives of an object. By providing either a text prompt describing the desired object or uploading an image, the tool processes the input to produce a grid displaying various views of that object. This capability is particularly useful for visualizing objects from different angles without manual rendering. While the live website indicates a runtime error, the tool's core functionality is designed for creating diverse 3D representations, making it suitable for research, experimentation, and educational purposes in 3D content generation and diffusion techniques.

Murder.Ai is an interactive AI Agents & Automation tool hosted on Hugging Face Spaces, designed to simulate and solve murder cases. Users can select a case file, configure game settings, and engage with various interactive tools to progress through the investigation. Key features include location mapping to visualize crime scenes, evidence collection mechanisms, and suspect interviews to gather information. The platform offers a unique way to explore narrative-driven AI interactions, allowing users to choose between different gameplay experiences. It serves as an experimental environment for understanding how AI can be applied to complex problem-solving scenarios within a fictional context.

Music Arena Leaderboard is an AI tool designed to compare and rank AI-generated songs from various platforms, including Suno, Udio, Google, and Meta. Users can visit the Music Arena to view an interactive leaderboard of top tracks, allowing them to explore and discover the best AI-generated music without needing to provide any input. The platform serves as a community-driven space where AI-generated songs are ranked, offering insights into the performance and quality of different AI music generators. It's a valuable resource for anyone interested in the evolving landscape of AI music creation.