Research & Education

Awesome-Visual-Transformer is a comprehensive, open-source repository dedicated to collecting and organizing academic papers focused on the application of transformers in computer vision (CV). This tool serves as an invaluable resource for researchers, academics, and practitioners looking to stay updated on the latest advancements in this rapidly evolving field. The collection includes original transformer papers, surveys, and numerous arXiv preprints covering diverse topics such as 3D semantic segmentation, object detection, image generation, medical image synthesis, and video processing. Users can easily browse papers, often with links to associated code, making it a practical resource for both theoretical understanding and implementation. The repository encourages community contributions through issues and pull requests, fostering a collaborative environment for knowledge sharing.

Awesome-LLM-Inference is a comprehensive, curated list of research papers and associated code focused on Large Language Model (LLM) and Vision-Language Model (VLM) inference. This resource is designed for researchers and engineers looking to explore advanced techniques for optimizing LLM performance, including Flash-Attention, Paged-Attention, WINT8/4 quantization, and various parallelism strategies. The repository provides an organized collection of topics such as multi-GPU/multi-node parallelism, disaggregating prefill and decoding, KV cache scheduling, and long-context attention optimization. It also features sections on LLM algorithmic/evaluation surveys, inference frameworks, and specific topics like Mixture-of-Experts (MoE) LLM inference. Users can download all listed PDFs via a Python script, making it a valuable hub for staying updated on the latest advancements in efficient LLM inference.

Artemis AI is an innovative application designed to transform bedtime routines by creating personalized stories for children. Leveraging advanced AI technology, it allows users to select heroes, settings, and morals, generating unique narratives tailored to their child's interests and developmental needs. The platform emphasizes fostering empathy, emotional intelligence, and inclusive representation through its diverse characters and story themes. It offers a free trial with up to three stories, encouraging a stronger bond between parents and children through shared storytelling experiences. Artemis AI also provides a library of articles on nurturing emotional intelligence and effective strategies for teaching empathy, making it a comprehensive tool for enriching children's growth.

ChatMusician is an AI chatbot specifically developed for musical applications, enabling users to understand and generate music. This tool facilitates the exploration of musical ideas and the automation of various music-related tasks. It is provided with comprehensive resources including code, models, data, and benchmarks, making it suitable for a wide range of users interested in music creation. The platform aims to assist musicians, students, and anyone with an interest in leveraging AI for musical endeavors, offering a foundation for both learning and practical application in music technology.

Xinquiry is an AI education platform designed to help individuals, particularly students from tier 2 and tier 3 cities, master English for IT jobs. The platform focuses on improving communication skills, teaching industry-specific terminology, and preparing users to ace job interviews. By combining AI-powered learning with human mentorship, Xinquiry aims to provide an affordable, fair, and high-quality educational experience. It's ideal for job seekers looking to upskill their English proficiency specifically for the demands of the IT sector, ensuring they are well-prepared for career advancement.

Bloom Book is an AI tool available on Hugging Face Spaces, designed for text generation and related tasks. It leverages the Streamlit framework to create interactive data applications, providing a platform for users to explore and utilize AI models. While the live website currently shows a runtime error, indicating it may not be fully operational at this moment, its intended purpose is to facilitate engagement with AI-powered text generation. The tool is part of the bigscience initiative, aiming to make advanced machine learning applications accessible to the community.

Awesome-LLM-3D is a comprehensive, curated list of academic papers focusing on the intersection of Large Language Models (LLMs) and 3D-related tasks. This repository serves as a valuable resource for researchers and academics interested in 3D understanding, reasoning, generation, and embodied agents, as well as other foundational models like CLIP and SAM. It is actively maintained, with regular updates to include the latest advancements in the field. The list is organized by various sub-topics, including 3D Unified Understanding and Generation, 3D Understanding (LLM and other Foundation Models), 3D Reasoning, 3D Generation, 3D Embodied Agents, and 3D Benchmarks, making it easy to navigate and find relevant literature. The project also highlights recent news and survey papers in the domain.

Chinese Instruments is an AI-powered tool designed to identify traditional Chinese musical instruments from short audio clips. Users can upload an audio snippet, typically around 3 seconds in length, and optionally select a pre-trained model for analysis. The tool then processes the audio and returns the name of the Chinese instrument detected. This application is hosted on Hugging Face Spaces, making it accessible for anyone interested in identifying traditional Chinese instrument sounds, whether for research, education, or personal curiosity. It leverages machine learning to provide insights into the rich soundscape of Chinese traditional music.

CL EVA02 LoRA ONNX Tagger is an AI tool designed for image tagging, specifically for anime images and illustrations. Users can upload an image or provide an image URL to receive predicted tags that describe its content. The tags are categorized into types such as rating, general, and character. The tool also offers a visualization of the generated tags, providing a comprehensive overview of the image's characteristics. It utilizes ONNX models for efficient image classification, making it suitable for tasks like organizing image datasets and supporting computer vision research.

Awesome-LLM-based-Text2SQL is a comprehensive repository dedicated to large language model-based text-to-SQL (LLM-based Text-to-SQL). It serves as a curated list of research papers, benchmarks, and open-source projects in this rapidly evolving field. The repository includes content from the survey paper "Next-Generation Database Interfaces: A Survey of LLM-based Text-to-SQL" and is regularly updated to reflect the latest advancements and notable contributions from the text-to-SQL community. It categorizes resources into trends, surveys, benchmarks (including BIRD, Spider 1.0, Spider 2.0, BIRD-CRITIC, BIRD-INTERACT), datasets (original and post-annotated), and a taxonomy covering in-context learning and fine-tuning methods. This makes it an invaluable resource for researchers and developers working on natural language interfaces for databases.

awesome-llm-json is a comprehensive, open-source resource list dedicated to leveraging Large Language Models (LLMs) for generating structured outputs, primarily JSON. This curated collection encompasses various categories such as hosted and local LLM models, Python libraries, blog articles, videos, Jupyter notebooks, and leaderboards. It clarifies terminology like Function Calling, JSON Mode, Tool Usage, and Guided Generation, which are often used interchangeably. The list details specific models from providers like Anthropic, OpenAI, and Mistral that support structured output, as well as local models like Hermes 2 Pro. Python libraries such as DSPy, Instructor, LangChain, and Pydantic are highlighted for their roles in constrained generation and data validation. This resource is invaluable for developers and researchers looking to implement or understand structured output generation with LLMs.

Clevered is an online e-learning platform specializing in Artificial Intelligence and coding courses for students aged 6 to 18. The curriculum, developed by professors from institutions like MIT, Oxford, and IIT, focuses on practical skills in coding, data science, and AI. Programs include Junior Data Scientist, Young Coder, and AI Internship, offering live interactive sessions with expert mentors. Students gain hands-on experience through projects, receive certifications from Google, IBM, and Microsoft, and can participate in internships with opportunities for letters of recommendation from Oxford researchers or former NASA Chief Scientists. Clevered emphasizes a unique pedagogy that prepares students for real-world applications and future careers.

Danbooru Tags Transformer V2 with WD Tagger & Florence 2 Flux Captioner is an AI tool designed to assist users in creating detailed prompts for AI art generation. By uploading an image, users can leverage the power of WD Tagger and Florence 2 Flux Captioner models to automatically generate relevant tags and captions. The tool offers customization options for these generated prompts, allowing users to fine-tune them to their specific needs. Once satisfied, the prompts can be easily copied to the clipboard for use in various AI art generation platforms. This tool is hosted on Hugging Face Spaces, making it accessible for those looking to enhance their AI art creation workflow.

Dimple 7B is a discrete diffusion multimodal large language model designed for image-text-to-text tasks. This application enables users to upload images and type questions or prompts, receiving informative answers and detailed responses. Built upon Dream-org/Dream-v0-Instruct-7B, Dimple 7B has been trained on extensive datasets such as LLaVA-CC3M-Pretrain-595K and Lmms-lab/LLaVA-NeXT-Data, ensuring robust performance in multimodal understanding and generation. It provides a platform for advanced AI interactions, bridging the gap between visual and textual information to deliver comprehensive outputs.

Doc To Dialogue is an innovative AI tool designed to convert PDF documents into dynamic interview audio. Users can upload any PDF report or document, and the application will generate an engaging audio interview that summarizes the key insights. This tool offers the flexibility to choose the language for the interview, making it versatile for various users and content types. The output is a convenient audio file, perfect for quick consumption of document content. It's an ideal solution for anyone looking to transform static text into an interactive and easily digestible audio format, enhancing accessibility and engagement with information.

ReasonFlux is an advanced open-source LLM post-training suite developed by a collaboration of Princeton University, PKU, UIUC, University of Chicago, and ByteDance Seed. Its core mission is to build next-generation reasoning capabilities by focusing on innovative algorithms for data selection, reinforcement learning, and inference scaling. The suite includes ReasonFlux-PRM, which offers trajectory-aware process reward models for long Chain-of-Thought (CoT) reasoning, providing dense supervision for data selection and policy optimization. ReasonFlux-Coder introduces a co-evolutionary reinforcement learning approach for LLM coders and unit testers, leading to more robust coding capabilities. Additionally, the suite incorporates preliminary work on thought templates, such as Buffer of Thoughts and SuperCorrect, to guide complex problem-solving and achieve state-of-the-art performance with higher efficiency.

text-classification-surveys is an open-source GitHub repository dedicated to compiling extensive resources for text classification within Natural Language Processing (NLP). It offers a detailed overview of various models, ranging from deep learning approaches like SpanBERT, ALBERT, and BERT, to shallow learning techniques such as LightGBM, SVM, and Random Forest. The repository also covers a wide array of text classification datasets, including MR, SST, IMDB, and Yelp, alongside common evaluation metrics like accuracy, Precision, Recall, and F1. Furthermore, it addresses technical challenges, including multi-label text classification. The content is primarily derived from the paper "A Survey on Text Classification: From Shallow to Deep Learning," making it a valuable resource for researchers and students in the field.

CausalNLP_Papers is a curated, open-source reading list hosted on GitHub, dedicated to papers exploring causality within natural language processing (NLP). This repository serves as an invaluable resource for researchers, academics, and practitioners seeking to understand and apply causal inference methods to NLP problems. It categorizes papers into sections such as Causality Basics, Causality Applied to General NLP, and Causality for Various Applications, including persuasion, psychology, economics, and healthcare. The list also features toolboxes for causal discovery and effect estimation, as well as resources from prominent labs like Schoelkopf's and Bengio's, making it a comprehensive guide for systematic learning and exploration in causal NLP.

Talk To Qwen Webrtc is an AI tool designed for real-time voice interaction with the Qwen2Audio model, leveraging Gradio and WebRTC technologies. Users can speak into a microphone, and the application will transcribe their speech into text. Following transcription, the tool processes the audio input and generates a text-based response, enabling dynamic communication with an AI. This platform is hosted on Hugging Face Spaces, making it accessible for experimentation with AI-driven audio processing and voice agents. It offers a straightforward interface for those looking to explore speech-to-text and AI response generation capabilities.

Reinforcement-Learning-Papers is an open-source GitHub repository that serves as a curated collection of research papers in the field of reinforcement learning. It encompasses both foundational classic papers and the latest research presented at top conferences such as ICLR, ICML, and NeurIPS. The collection primarily focuses on single-agent reinforcement learning, offering a structured overview of various topics including Model-Free (Online) RL, Model-Based (Online) RL, Offline RL, Meta RL, Adversarial RL, and RL with Transformer/LLM. This resource is invaluable for researchers, academics, and students who need to stay updated on significant advancements and foundational concepts in reinforcement learning.

scGPT is an open-source codebase designed to build a foundation model for single-cell multi-omics data using generative AI. It offers pre-trained models for various human cell types and organs, including whole-human, brain, blood, heart, lung, kidney, and pan-cancer. The tool supports zero-shot applications for cell embedding tasks and features efficient reference mapping for millions of cells. Researchers can install scGPT via pip and utilize online apps for reference mapping, cell annotation, and Gene Regulatory Network inference. It is ideal for bioinformaticians and computational biologists working with complex single-cell datasets.

Category_Theory_Machine_Learning is a GitHub repository that curates a list of academic papers exploring the intersection of machine learning and category theory. The repository organizes papers by various fields such as Deep Learning, Equivariance, Graph Neural Networks, Differentiable Programming, Probability Theory, Bayesian/Causal Inference, Topological Data Analysis, and more. It also includes theses and blog posts related to the subject. This resource is designed for researchers, academics, and enthusiasts interested in the theoretical foundations of machine learning, offering a structured overview of relevant publications and encouraging community contributions through pull requests or issue submissions.

Emu2 is a generative multimodal model developed by BAAI, designed for in-context learning and capable of processing both image and text inputs. This application, hosted on Hugging Face Spaces, enables users to generate various forms of content and engage in interactive chat experiences. By providing a combination of text and images, users can receive generated responses or participate in conversations, making it a versatile tool for multimodal AI research and experimentation. The model aims to push the boundaries of AI's ability to understand and create content across different modalities.

rwa is an open-source recurrent neural network (RNN) model designed for machine learning on sequential data. It introduces a novel approach by computing a recurrent weighted average (RWA) over every previous processing step, allowing for direct connections anywhere along a sequence. This method contrasts with traditional RNN architectures that typically rely only on the immediate previous step. The RWA model can be computed as a running average, meaning it doesn't need complete recomputation at each step, leading to scalability comparable to LSTM models. Notably, rwa demonstrates considerably faster training on most tasks, often by a factor of five or more, with performance improvements scaling further for longer sequences. While effective for many tasks, it has not yielded competitive results for Natural Language Problems.