Research & Education

Lumina-mGPT 2.0 is an open-source, stand-alone, decoder-only autoregressive model designed for a broad spectrum of image generation tasks. Trained from scratch, it supports functionalities such as text-to-image generation, image pair generation, subject-driven generation, multi-turn image editing, controllable generation, and dense prediction. The project provides inference code for image-to-image tasks and all-in-one model checkpoints on HuggingFace. It also offers acceleration strategies like Speculative Jacobi Decoding and Model Quantization to optimize inference time and GPU memory usage. Lumina-mGPT 2.0 is ideal for AI researchers and machine learning engineers looking to explore and implement advanced image modeling techniques.

Luotuo-Chinese-LLM is an open-source initiative focused on advancing Chinese large language models. The project, developed by researchers from Huazhong Normal University and SenseTime, encompasses a range of models, datasets, pipelines, and applications. Key sub-projects include ChatHaruhi for character-based conversational AI, Luotuo Embedding for generative text embedding, Luotuo QA for conversational question answering, and Mini Luotuo for distilled instruction-following models. It also features Silk Road for building Chinese LLM data foundations and Silk Magic Book for collecting effective prompts. The project emphasizes practical applications and research into cross-language data tuning.

llm-universe offers a comprehensive tutorial for beginner developers interested in large language model (LLM) application development. The project is designed to be highly practical, guiding users through the creation of a personal knowledge base assistant on an Alibaba Cloud server. It covers essential topics such as LLM introductions, API calling methods for various models (including GPT, Baidu Wenxin, iFlytek Spark, and Zhipu AI), knowledge base construction, and building RAG (Retrieval Augmented Generation) applications. The tutorial emphasizes hands-on learning, simplifying complex concepts and focusing on core skills needed to develop LLM-powered applications, making it accessible even for those without a strong AI or algorithm background.

ICLR2025-Papers-with-Code is a comprehensive GitHub repository dedicated to compiling research papers and their corresponding open-source projects from the International Conference on Learning Representations (ICLR). The collection spans from ICLR 2021 to the upcoming ICLR 2025, with a particular emphasis on advancements in Large Language Models (LLMs) and various subfields within Natural Language Processing (NLP). This resource serves as a valuable hub for researchers, academics, and developers looking to stay updated on the latest research trends and access practical code implementations. The repository is actively maintained and updated, encouraging community contributions through watching, forking, and starring the project.

iir is an open-source project hosted on GitHub, offering a collection of algorithms and functionalities for machine learning, natural language processing, and information retrieval. Developed primarily in Python, Ruby, C++, and R, it serves as a valuable resource for researchers and developers in AI-related fields. The repository includes implementations for tasks such as active learning, clustering, natural language detection, LDA, PCA, perceptron, and various neural network components. Its modular structure allows users to explore and integrate different techniques for their specific AI projects, making it suitable for both academic research and practical application development.

Handy-Multi-Agent is a comprehensive tutorial designed for developers interested in understanding and implementing multi-agent systems. Based on the CAMEL-AI framework, this guide starts with basic Agent development and progresses to complex Multi Agent applications. It emphasizes practical application and hands-on building, combining necessary theoretical knowledge with real-world examples. The project includes detailed documentation in the 'docs' directory and executable code in the 'code' directory, allowing users to run examples directly. It covers topics such as RAG, Memory, and Multi Agent techniques, aiming to enhance skills in building and managing intelligent agents and applying them to solve practical problems.

The McGill Artificial Intelligence Society (MAIS) is a student-run organization dedicated to making artificial intelligence more accessible to students. They achieve this by hosting a variety of initiatives, including bootcamps like MAIS 202 for ML fundamentals, workshops on applied machine learning topics, and Canada's largest AI hackathon, MAIS Hacks. MAIS also fosters community through events like the Learnathon, an undergraduate AI research conference, and the McGill AI Podcast, which connects ML principles to research disciplines. The society aims to connect McGill students with the broader Montreal AI ecosystem through industry events and networking opportunities.

MathModelAgent is an AI agent specifically designed for mathematical modeling, capable of automating the entire process from problem analysis to paper generation. It can produce a complete, submission-ready paper, significantly reducing the time required for modeling competitions. Key features include automatic problem analysis, mathematical modeling, code writing, error correction, and paper drafting. The tool offers both local and cloud-based code interpreters, supports multi-agent collaboration (e.g., modeler, coder, writer), and allows for the use of multiple large language models. It is designed to be cost-effective and offers customizable templates for prompt injection. Future plans include a web UI, CLI, English support for American Mathematical Contest in Modeling, LaTeX integration, and visual model integration.

Sirch is an AI-powered search engine designed to enhance the web search experience. It allows users to search the web and receive AI-powered summaries of every page, providing quick insights without needing to visit each site individually. A key feature is the inclusion of real site previews, enabling users to quickly assess the relevance and content of search results before clicking. This tool aims to streamline the research process by combining traditional web search with advanced AI summarization and visual previews, all within a single interface. It focuses on delivering a dynamic and efficient search experience.

stable-diffusion-tutorial provides a complete set of tutorials for Stable Diffusion, meticulously crafted over three months. This resource guides users from initial setup and configuration to advanced techniques like ControlNet and Lora model training. It covers essential topics such as installing Stable Diffusion, understanding model types, performing text-to-image and image-to-image generation, and installing extensions. Additionally, the tutorial includes sections on AI painting websites, high-definition image upscaling, and integrating AI painting plugins with Photoshop, making it a valuable resource for anyone looking to master Stable Diffusion.

OS Ninja provides an intelligent way to explore and learn open-source projects by generating AI-powered learning paths for any repository. It decodes complex codebases and creates structured tutorials, diagrams, and documentation that evolve with the code. Users can search for open-source projects or request new ones to be added. The platform performs deep research on entire codebases, which can take up to 24 hours, to generate high-fidelity learning paths. It caters to various learning styles, including Socratic questioning, Feynman technique, and traditional book format. OS Ninja also offers curated collections of open-source repositories across categories like Generative AI, Data, Robotics, Game Engines, Crypto & Web3, and Machine Learning, making it a comprehensive resource for developers looking to master new codebases.

Leeroo is an AI platform designed to continuously learn and adapt to an organization's knowledge base and expert playbooks. It facilitates the deployment of data and AI programs, along with their operational user interfaces, directly onto existing infrastructure. The platform emphasizes experimentation and human approval in its process, ensuring that AI initiatives are aligned with business objectives and validated by human oversight. Leeroo aims to automate and streamline complex data and AI workflows, making them more efficient and integrated within an organization's operations. This approach helps businesses leverage AI for continuous improvement and operational excellence.

OccamzRazor is at the forefront of innovative drug discovery, leveraging advanced digital science and machine learning methods to accelerate the understanding and treatment of brain aging. The platform's primary focus is on Parkinson's disease, aiming to map out the complexities of the condition to develop effective cures. By applying sophisticated AI techniques, OccamzRazor seeks to enhance the efficiency and success rates of pharmaceutical research and development, moving beyond traditional approaches to unlock new therapeutic possibilities. This tool is designed for researchers and scientists in the pharmaceutical and biotech sectors who are dedicated to tackling neurodegenerative diseases.

Dhitva Technologies offers a comprehensive learning ecosystem that integrates Virtual Reality with AI analytics to significantly enhance educational outcomes. The platform is designed to improve retention rates by 80% and accelerate learning speed by 4x, as supported by global research. Dhitva stands out by combining high-fidelity VR simulations with a proprietary AI Chat Bot that provides real-time guidance and feedback. This AI analyzes student gaze and decisions, offering instant corrections and generating adaptive scenarios so no two learning sessions are the same. The platform is compatible with Meta Quest, HTC Vive, and standard web browsers, making it accessible for modern educational institutions.

The National Centre of Artificial Intelligence (NCAI) at UET Lahore, established as the Al-Khwarizmi Institute of Computer Science (KICS) in August 2002, is dedicated to advancing research and development in Computer Science and Information Technology. KICS engages in various research activities, including the development of AI and machine learning solutions. The institute also emphasizes technology transfer through research labs, technology centers, and incubated startups. It actively participates in research collaborations and hosts conferences and workshops, such as the IEEE International Conference on Open Source System and Technologies (ICOSST), to disseminate knowledge and foster innovation in the AI domain.

Studiolo is an innovative AI tool designed to facilitate rapid and personalized learning experiences. Users can select any topic, from 'Korean Cooking Fundamentals' to 'Intro to Machine Learning,' and the platform aims for them to remember it within 45 minutes. It allows for the augmentation of learning inquiries with local assets, enabling a truly customized educational path. Studiolo also offers options to define expertise levels, current depth of understanding, and session lengths, ranging from 5 to 45 minutes. For those unsure what to learn, a chat feature helps guide the user, making it accessible for a wide range of learners seeking efficient knowledge acquisition.

AIGC_Interview is a GitHub repository designed as a comprehensive guide for individuals seeking jobs in the AIGC (AI-Generated Content) field. It compiles essential resources such as interview experiences, fundamental knowledge, and prompt engineering techniques. The repository covers critical topics like ChatGPT, Stable Diffusion, Prompt, Embedding, and Fine-tuning, offering insights into what job seekers need to know for AIGC-related positions. It aims to assist users in preparing for interviews, understanding industry trends, and navigating the job market, particularly for roles like prompt engineers and AI algorithm specialists. The project also encourages community contributions, including sharing job opportunities and interview experiences.

BELLE, which stands for "Be Everyone's Large Language model Engine," is an open-source initiative by LianjiaTech focused on advancing Chinese dialogue large language models. Unlike projects primarily concerned with pre-training, BELLE emphasizes enabling individuals to create their own high-performing, instruction-following language models based on existing open-source pre-trained models. The project continuously releases instruction training data, relevant models, training code, and application scenarios. It also evaluates the impact of different training data and algorithms on model performance, with a specific optimization for Chinese language using ChatGPT-generated data. Recent updates include enhanced Chinese speech recognition models, multimodal large language models, and research reports on fine-tuning strategies and RLHF training.

Awesome-ChatGPT-prompts-ZH_CN is a GitHub repository dedicated to providing a diverse collection of Chinese prompts for ChatGPT and other large language models like Claude. It offers creative methods to customize AI behavior, such as transforming ChatGPT into a 'cat-girl' persona for role-playing scenarios. The repository also includes techniques for bypassing certain AI content restrictions and limitations, with specific instructions for ChatGPT and NewBing. It features tools for exporting conversations, bypassing WAF errors, and enhancing the AI's mathematical capabilities. The project is open-source, encouraging community contributions and providing updates on new bypass methods and prompt engineering techniques.

awesome-chatgpt-zh is a comprehensive, open-source Chinese guide designed to empower users with the knowledge and resources to effectively leverage ChatGPT. Hosted on GitHub, this project offers detailed instructions, prompt engineering guidelines, and application development insights. It curates a wide array of free and paid ChatGPT resources, along with a list of top open-source projects and productivity tools built on ChatGPT's capabilities. The guide covers various aspects, from understanding what ChatGPT is to advanced topics like LLM development, RAG guidance, and AGI concepts, aiming to significantly boost user productivity.

This GitHub repository, ChatGPT-Prompt-Engineering-for-Developers-in-Chinese, offers unofficial Chinese and English subtitles for the popular "ChatGPT Prompt Engineering for Developers" course. It aims to make the technical content accessible to a broader audience, particularly Chinese-speaking developers. The repository includes core bilingual subtitles, as well as separate English and Chinese subtitle files, along with course notebooks. This resource is invaluable for those looking to master prompt engineering for ChatGPT, covering best practices, sentiment classification, text summarization, email writing, translation, and building chatbots. It also provides insights into GPT API development, allowing users to extend their learning into building impressive applications.

Voaige is developing a Test Time Cognition layer for Large Language Models (LLMs), aiming to enhance their reasoning capabilities beyond traditional reinforcement learning and fixed next-token predictions. This innovative approach involves dynamically allocating computational resources during inference, allowing LLMs to perform efficient search and adaptation at test time, similar to how biological cognition navigates complex problems. By understanding and implementing principles from neuroscience, Voaige seeks to enable LLMs to assess difficulty, allocate compute where uncertainty is high, and scale back where it's not, leading to better generalization and the ability to handle novel planning and open-ended complexity without extensive retraining. Their research focuses on architecturally grounded inference systems inspired by the brain's adaptive search mechanisms.

Deep Learning IndabaX South Africa is an annual, locally-organized conference dedicated to fostering knowledge and capacity in machine learning across the African continent. The multi-day event features over 400 attendees, 50+ speakers showcasing cutting-edge research and applied AI, and 4+ tutorials providing practical introductions to machine learning for beginners. Participants can also present their work through 50+ posters, engage in hackathon problems, and connect with a diverse community of researchers and practitioners. The conference aims to make machine learning accessible, offering free attendance options for students and travel grants, supported by various partners and paid registration for academics and industry professionals.

dive-into-llms is an open-source programming tutorial series designed to help users dive into large language models (LLMs). Originating from courses at Shanghai Jiao Tong University, this free resource offers practical programming exercises covering a wide range of LLM-related topics. Users can learn about fine-tuning and deploying pre-trained models, prompt engineering and chain-of-thought, knowledge editing, mathematical reasoning, model watermarking, and even jailbreak attacks. The tutorial also delves into advanced concepts like LLM steganography, multimodal models, GUI agents, and AI agent security. Additionally, it features a newly launched series on full-process LLM development in collaboration with Huawei Ascend, providing comprehensive guidance with PPTs, experimental manuals, and videos.