Home>Programming related>AI Source Code
Download

AI Resource Central

Gathering excellent artificial intelligence open source projects around the world!

Simplified Chinese English


Welcome to AI Resource Central ! This warehouse focuses on bringing together excellent artificial intelligence (AI) open source projects from around the world. Whether you are looking for inspiration to start your own project or want to learn how to use the latest AI technologies, this is a great starting point. We are committed to providing AI developers, researchers and enthusiasts with a platform to explore, communicate and share the code and implementation of various AI projects.

AI project summary

Table of contents

  • AI Resource Central

    • Intelligent body

    • Prompt word engineering

    • AI development interface

    • Agile development framework

    • Model

    • AI List

    • Reasoning Optimization

    • Information aggregation

    • Code Assistant

    • AI Tutorial

    • Workflow automation

    • AI robot

    • Multimodal model

    • Multilingual Model

    • Data organization

    • AI Services

    • Vector database

    • AI Education

    • AI development tools

    • Model training

    • AI image generation

    • Dataset

    • Model evaluation

    • Computational Management

    • AI Writing

    • Intelligent monitoring

    • Video generation

    • Data Management

    • AI Applications

    • AI Engineering

    • Target

    • Project Catalog

    • How to contribute

    • license

Target

Our main goals are:

  • Establish a comprehensive AI project library : projects covering a wide range of fields such as machine learning, deep learning, and natural language processing.

  • Promote the spirit of open source : By demonstrating high-quality open source projects, more people are encouraged to participate in the open source community.

  • Promote technological innovation : provide developers with practical cases and solutions, and accelerate the application and development of AI technology.

  • Support Learning and Development : Provide practical opportunities for learners of all levels to help them master the latest AI tools and technologies.

Project Catalog

Here is a list of some of our selected AI projects by different fields:

Intelligent body

  1. Significant-Gravitas/AutoGPT - AutoGPT is designed to enable everyone to use and develop artificial intelligence. Its mission is to provide people with tools to focus on important matters.

  2. geekan/MetaGPT - The first AI software company's multi-agent framework is aimed at natural language programming.

  3. microsoft/autogen - A programming framework for autonomous artificial intelligence with related resources on PyPi, Discord and Office Hour.

  4. reworkd/AgentGPT - Assemble, configure and deploy autonomous AI agents in the browser.

  5. joaomdmoura/crewAI - Role-playing and autonomous AI body framework. It enables agents to collaborate and handle complex tasks.

  6. microsoft/JARVIS - JARVIS is a system used to connect large language models (LLMs) with the machine learning (ML) community. (Paper: https://arxiv.org/pdf/2303.17580.pdf)

  7. mem0ai/mem0 - The storage layer of artificial intelligence applications.

  8. microsoft/semantic-kernel - Quickly and easily integrate state-of-the-art large language model (LLM) technology into your application.

  9. yoheinakajima/babyagi -

  10. openai/swarm - an ergonomic, lightweight multi-agent orchestration educational framework managed by the OpenAI solution team.

  11. phidatahq/phidata - builds multimodal agents with memory, knowledge, tools and reasoning capabilities and chats through a beautiful agent user interface.

  12. TransformerOptimus/SuperAGI - SuperAGI is a developer-first open source autonomous artificial intelligence proxy framework that allows developers to quickly and reliably build, manage and run useful autonomous agents.

  13. composiohq/composio - Composio equips more than 100 high-quality integrations for artificial intelligence agents and large language models (LLMs) through function calls.

  14. cpacker/MemGPT - Letta (formerly known as MemGPT), a framework for creating large language model (LLM) services with memory capabilities.

  15. google-deepmind/deepmind-research - This repository contains implementation code and sample code for DeepMind publications.

  16. botpress/botpress - An open source center for building and deploying GPT/LLM agents.

  17. OpenMOSS/MOSS - An open source conversational language model developed by Fudan University and enhanced by tools.

  18. smol-ai/developer - The first library that allows you to embed developer agents in your own applications.

  19. OpenBMB/XAgent - An autonomous language model agent for solving complex tasks.

  20. langchain-ai/langgraph - Constructs elastic language agents in the form of graphs.

  21. e2b-dev/e2b - Secure open source cloud runtime for AI applications and proxy.

  22. modelscope/agentscope - Easier to build multiagent applications empowered by large language models (LLMs).

  23. homanp/superagent - Run an AI proxy through the API.

  24. aiwaves-cn/agents -

  25. frdel/agent-zero - Agent Zero artificial intelligence framework.

  26. microsoft/TinyTroupe - Multiagent role simulation powered by large language models (LLM) to enhance imagination and gain business insights.

  27. QwenLM/Qwen-Agent - A proxy framework and application based on Qwen≥2.0, with function calls, code interpreter, retrieval enhancement generation (RAG) and Chrome extension functions.

  28. OpenBMB/AgentVerse - AgentVerse is designed to deploy multiple large language model (LLM)-based agents in applications, mainly providing task resolution and simulation frameworks.

  29. Significant-Gravitas/Auto-GPT-Plugins - Auto-GPT plugin.

  30. huggingface/smolagents - Smolagents is a basic library for agents. Agents use it to write Python code for tool-calling and agent-orchestrating.

  31. Ironclad/rivet - An open source visual artificial intelligence programming environment and TypeScript library.

  32. gmpetrov/databerry - A codeless platform for creating custom large language model (LLM) agents.

  33. OpenBMB/BMTools - Open source solution for large models tool learning with ChatGPT plugin.

  34. langroid/langroid - Use multi-agent programming to control large language models.

  35. muellerberndt/mini-agi - MiniAGI is a simple universal autonomous agent that relies on the OpenAI API.

  36. Farama-Foundation/PettingZoo - A multi-agent reinforcement learning application program interface (API) standard, including commonly used reference environments and utilities.

  37. Josh-XT/AGiXT - AGiXT is a dynamic AI platform that uses adaptive memory, intelligent features and plug-in systems to manage instructions and perform tasks among multiple AI vendors to provide efficient AI solutions.

  38. Togethercomputer/moa - Hybrid Agent (MoA) achieved 65.1% in Alpaca Evaluation (AlpacaEval) using open source software models.

  39. AgentOps-AI/agentops - Python Software Development Kit (SDK) for AI Agent Monitoring, Large Language Model (LLM) Cost Tracking and Benchmarking. It integrates with a variety of large language models and proxy frameworks.

  40. noahshinn/reflexion - [NeurIPS 2023] "Reflexion: Language Agents Based on Speech Reinforcement Learning"

  41. SciSharp/BotSharp - Artificial Intelligence Multi-Agility Framework in .NET.

  42. dot-agent/nextpy -

  43. iterative/datachain - Extraction, transformation, loading (ETL), analysis and versioning of unstructured data.

  44. agiresearch/OpenAGI - OpenAGI: Encountering of Large Language Models (LLMs) with domain experts.

  45. InternLM/lagent - A lightweight framework for creating proxy based on large language models.

  46. MineDojo/MineDojo - Use Internet-scale knowledge to build open embodied intelligent bodies.

  47. Forethought-Technologies/AutoChain - AutoChain is used to create lightweight, scalable and testable large language model (LLM) agents.

  48. landing-ai/vision-agent - Visual agent.

  49. BCG-X-Official/agentkit - A starter kit for building restricted agents using Nextjs, FastAPI, and Langchain.

  50. jina-ai/thinkgpt - A proxy technology for enhancing large language models (LLMs) and breaking through its limitations.

  51. farizrahman4u/loopgpt - A modular framework for Auto - GPT.

  52. Farama-Foundation/chatarena - ChatArena is a multiagent language gaming environment for large language models (LLMs) designed to develop the communication and collaboration capabilities of artificial intelligence.

  53. THUDM/AgentTuning - Agent Tuning gives general agent capabilities to large language models.

  54. Yifan-Song793/RestGPT - Autonomous proxy based on large language models controls real-world applications through RESTful API (Denotative State Transfer Application Program Interface).

  55. Link-AGI/AutoAgents - At the 2024 International Joint Conference on Artificial Intelligence (IJCAI), different GPT roles were generated to form a collaborative entity to handle complex tasks.

  56. AI-Engineer-Foundation/agent-protocol - This is a common interface that interacts with artificial intelligence agents, independent of the technology stack and can be used for any agent building framework.

  57. kreneskyp/ix - A proxy platform for autonomous GPT-4.

Prompt word engineering

  1. f/awesome-chatgpt-prompts - This library organizes ChatGPT prompt words to better use ChatGPT and other large language model (LLM) tools.

  2. PlexPt/awesome-chatgpt-prompts-zh - ChatGPT Chinese training guide. Guide to use various scenarios. Learn how to make it follow your instructions.

  3. dair-ai/Prompt-Engineering-Guide - Guides, papers, lectures, notes and resources for prompting engineering.

  4. stanfordnlp/dspy - DSPy: A framework for programming programming language models rather than prompting.

  5. guidance-ai/guidance - A bootstrapping language for controlling large language models.

  6. outlines-dev/outlines - Structured text generation

  7. mshumer/gpt-prompt-engineer -

  8. jxnl/instructor - Structured output of large language models (LLMs).

  9. brexhq/prompt-engineering - Tips and tricks for using OpenAI's GPT - 4 and other large language models.

  10. LouisShark/chatgpt_system_prompt - A set of GPT system prompt words and knowledge about prompt injection/leakage.

  11. microsoft/TypeChat - TypeChat is a library for building natural language interfaces with types.

  12. sgl-project/sglang - SGLang is a fast service framework for large language models and vision-language models.

  13. mit-han-lab/streaming-llm - An efficient streaming language model with attention convergence (ATtention Sinks) proposed at the 2024 International Conference on Learning Representation (ICLR).

  14. spdustin/ChatGPT-AutoExpert - Enhanced custom instruction for ChatGPT (non-encoding) and ChatGPT advanced data analysis (encoding).

  15. civitai/civitai - A repository containing model and text inversion.

  16. Moonvy/OpenPromptStudio - AIGC Prompt Word Visual Editor | Operation and Maintenance | Open Prompt Studio

  17. rockbenben/ChatGPT-Shortcut - Maximize efficiency and productivity with AI shortcuts. Customize, save and share tips, and find tips for different scenarios in the shared community.

  18. microsoft/promptbase - Everything related to prompt project.

  19. PrefectHQ/marvin - Create a pleasant AI interface.

  20. promptfoo/promptfoo - Test prompt words, proxy and retrieval enhancement generation (RAG). At the same time, red team testing, penetration testing and vulnerability scanning are performed on large language models (LLMs), compare the performance of large language models, and simple configuration through command line and continuous integration/continuous deployment (CI/CD) integration.

  21. princeton-nlp/tree-of-thought-llm - Thought on the use of large language models for deliberate problem solving at the 2023 Neural Information Processing System Conference (NeurIPS 2023).

  22. pydantic/pydantic-ai - A proxy framework or populator for using Pydantic in large language models (LLMs).

  23. 1rgs/jsonformer - A reliable way to generate structured JSON from a language model.

  24. thunlp/OpenPrompt - An open source framework for prompt learning.

  25. guardrails-ai/guardrails - Add security restrictions or constraints to large language models.

  26. eth-sri/lmql - A language that efficiently program large language models (LLMs) under constraint guidance.

  27. prompttslab/Promptify - Prompt engineering and version control, using GPT or other prompt-based models to get structured output. Join Discord for related research.

  28. shreyashankar/gpt3-sandbox - This project is designed to allow users to create excellent web demonstrations with a few lines of Python code by using the new OpenAI GPT-3 API.

  29. hegelai/prompttools - an open source tool for prompting tests/experiments, supporting large language models (such as OpenAI, LLaMA) and vector databases (such as Chroma, Weaviate, LanceDB).

  30. bigscience-workshop/promptsource - A toolkit for natural language prompts, including creation, sharing, and use.

  31. YiVal/YiVal - Your general artificial intelligence application automatically prompts the engineering assistant.

  32. microsoft/prompt-engine - A library that helps developers create prompts for large language models.

  33. ianarawjo/ChainForge - An open source environment for visual programming for practical testing of tips for large language models (LLMs).

  34. spcl/graph-of-thoughts - The official implementation of "Thinking Map: Solving Complex Problems with Large Language Models".

  35. ysymyth/ReAct - [ICLR 2023] "ReAct: Combining Reasoning and Action in Language Models"

  36. Microsoft/genaiscript - Automatic generation of artificial intelligence (general artificial intelligence) scripts.

  37. jackmpcollins/magentic - Uninterrupted integration of large language models into Python functions.

  38. adieyal/sd-dynamic-prompts - A custom script written for AUTOMATIC1111/stable - diffusion - webui to create a small template language to randomly generate prompt words.

  39. zjunlp/EasyEdit - An easy-to-use knowledge editing framework for large language models (LLMs) at the 2024 American Association for Computational Linguistics (ACL) Conference.

  40. microsoft/aici - AICI: prompt as a WebAssembly program.

  41. zou-group/textgrad - TextGrad: uses large language models to backpropagate text gradients through automatic "derivative" of text.

  42. microsoft/PromptCraft-Robotics - A community of using large language models (LLMs) in the field of robotics and a robot simulator integrated with ChatGPT.

  43. greshake/llm-security - A new way to destroy large language models for integrated applications.

  44. noamgat/lm-format-enforcer - Strengthen the output format of the language model (such as JSON pattern, regular expressions, etc.).

  45. Ber666/llm-reasoners - A library for complex reasoning in large language models.

  46. jujumilk3/leaked-system-prompts - A collection of leaked system prompts.

  47. laiyer-ai/llm-guard - A security toolkit for large language model interactions.

  48. hiyouga/FastEdit - Quickly edit large language models in 10 seconds.

  49. timqian/openprompt.co - Create, use and share ChatGPT tips.

  50. exploration/spacy-llm - Integrate large language models (LLMs) into structured natural language processing (NLP) processes.

  51. protecttai/rebuff - Large Language Model (LLM) prompts the injection detector.

  52. getmetal/motorhead - Motorhead is a server for large language models (LLMs) for memory and information retrieval.

  53. Mirascope/mirascope - A large language model (LLM) abstract concept that is not obstructive.

  54. cocacola-lab/ChatIE -

AI development interface

  1. jmorganca/ollama - Get started quickly with Llama 3.3, Mistral, Gemma 2 and other large language models.

  2. ChatGPTNextWeb/ChatGPT-Next-Web - ChatGPT, Gemini and other cross-platform user interface (UI) enables you to have your own large language model (LLM) applications in one click.

  3. xtekky/gpt4free - The official gpt4free repository contains a variety of powerful language models.

  4. oobabooga/text-generation-webui - A Gradio web user interface for large language models that support multiple inference backends.

  5. RVC-Boss/GPT-SoVITS - One minute of speech data can be used to train a good speech synthesis (TTS) model (small sample speech clone).

  6. gradio-app/gradio - Build and share excellent Python machine learning applications. Like to support.

  7. mckaywrigley/chatbot-ui - All models offer AI chat capabilities.

  8. openai/openai-python - The official Python library for OpenAI API.

  9. danny-avila/LibreChat - An enhanced version of ChatGPT with various features such as different APIs, AI models and features, and it is an active self-hosted open source project.

  10. sunner/ChatALL - Chat with multiple chatbots (such as ChatGPT, Bing Chat, etc.) simultaneously to find the best answer.

  11. GaiZhenbiao/ChuanhuChatGPT - Graphical user interface (GUI) for ChatGPT API and many large language models (LLMs). It has various functions, such as agents and file-based QA, and has a beautiful user interface.

  12. CopilotKit/CopilotKit - React UI and elegant infrastructure for a variety of artificial intelligence applications such as copilot, in-app agent, chatbot and text areas.

  13. mlc-ai/web-llm - High-performance in-browser large language model (LLM) inference engine.

  14. jina-ai/clip-as-service - Use CLIP to perform scalable embedding, reasoning, and sorting images and sentences.

  15. chathub-dev/chathub - One-stop chatbot client.

  16. TheRamU/Fay - Fay is an open source digital human framework. It has different versions for various applications.

  17. sashabaranov/go-openai - Go wrapper for OpenAI ChatGPT, GPT - 3, GPT - 4, DALL·E and Whisper APIs.

  18. SillyTavern/SillyTavern - Front-end for advanced users in large language models (LLM).

  19. openai/openai-node - The official JavaScript/TypeScript library for OpenAI API.

  20. sebastianstarke/AI4Animation - Use computer intelligence to bring characters to life in Unity.

  21. xiangsx/gpt4free-ts - Provides a free OpenAI GPT-4 API in a replica project of a TypeScript version of xtekky/gpt4free.

  22. wzpan/wukong-robot - Wukong Robot is a simple, flexible and elegant Chinese voice conversation robot/smart speaker project. It supports ChatGPT multi-round dialogue and may be the first open source smart speaker project to support brain-computer interaction.

  23. yihong0618/xiaogpt - Play ChatGPT and other large language models (LLM) with Xiaomi smart speakers.

  24. nat/openplayground - A large language model (LLM) playground that can run on a laptop.

  25. postgresml/postgresml - Postgres (a database management system) with GPU for machine learning and artificial intelligence applications.

  26. Shaunwei/RealChar - Create, customize and have real-time conversations with AI characters/partners. Use various technologies to achieve seamless AI dialogue anytime, anywhere.

  27. ParisNeo/lollms-webui - The web user interface of the master of large language models.

  28. zhayujie/bot-on-anything - A chatbot builder based on large models can quickly integrate artificial intelligence models such as ChatGPT, Claude and Gemini into software applications and websites such as Telegram, Gmail, Slack.

  29. deanxv/coze-discord-proxy - Through Coze-Bot proxy Discord dialogue, request the GPT4 model through the API, providing functions such as dialogue, text-generated graphics, graphics-generated text and knowledge base search.

  30. vocodedev/vocode-python - Build a large language model (LLM) agent based on voice, modular and open source.

  31. alexrudall/ruby-openai - OpenAI API and Ruby.

  32. ahmadbilaldev/langui - Artificial Intelligence User Interface. Open source Tailwind components for GPT, Generative Artificial Intelligence, and Large Language Model (LLM) projects.

  33. ollama/ollama-js - Ollama JavaScript library.

  34. xusenlinzy/api-for-open-llm - OpenAI-style API for open large language models. Supports various models, such as LLaMA, ChatGLM, etc.

  35. Anse-app/anse - ChatGPT, DALL - E and Stable Diffusion models are super experience.

  36. mylxsw/aidea-server - AIdEA is a multi-functional APP that supports GPT, domestic large-scale language models (such as Tongyi Qianwen and Wenxin Yiyan), as well as used in literary and biographical pictures, biographical pictures, SDXL1.0, super Stable Diffusion for resolution and image coloring.

  37. aallam/openai-kotlin - Kotlin OpenAI API client that supports multi-platform and coroutines.

  38. guinmoon/LLMFarm - Use GGML library for Llama and other large language models offline on iOS and MacOS.

  39. uezo/ChatdollKit - ChatdollKit allows you to convert your own 3D models into chatbots.

Agile development framework

  1. langchain-ai/langchain - Build an application with situational awareness reasoning capabilities.

  2. nomic-ai/gpt4all - GPT4All is able to run local large language models (LLMs) on any device. It is open source and can be used for commercial purposes.

  3. comfyanonymous/ComfyUI - The most powerful and modular diffusion model has graph/node interfaces for graphical user interfaces (GUI), application program interfaces (APIs), and backends.

  4. langgenius/dify - Dify is an open source large language model (LLM) application development platform with an intuitive interface and a variety of functions to realize the process from rapid prototype production to production.

  5. lobehub/lobe-chat - Lobe Chat is an open source artificial intelligence chat framework with modern design. It supports multiple AI vendors, knowledge bases and multimodals, and can deploy private chat applications for free with one click.

  6. logspace-ai/langflow - Langflow is a Python-based, model-independent low-code application builder for RAG (retrieval enhancement generation) and multi-agent AI applications. It can work in conjunction with any API or database.

  7. run-llama/llama_index - LlamaIndex is a data framework for large language model (LLM) applications.

  8. FlowiseAI/Flowise - Create your personalized large language model (LLM) process using a drag-and-drop user interface.

  9. chatchat-space/Langchain-Chatchat - Langchain - Chatchat (originally named Langchain - ChatGLM) is a search-enhanced generation (RAG) and agent (Agent) application based on Langchain, ChatGLM, Qwen, Llama, etc. for local knowledge-based Large Language Model (LLM).

  10. go-skynet/LocalAI - An open source project that can replace services such as OpenAI and Claude. It can run on consumer hardware and perform various tasks such as generating different media types.

  11. infiniflow/ragflow - RAGFlow is an open source RAG (retrieval enhancement generation) engine for deep document understanding.

  12. mindsdb/mindsdb - AGI's query engine is a platform for building artificial intelligence that can learn and answer questions on federated data.

  13. embedchain/embedchain - The storage layer of your artificial intelligence application.

  14. songquanpeng/one-api - This is an OpenAI key management and redistribution system. It supports multiple large language models (LLMs), has an English user interface (UI), can be executed in a single file, and has Docker images for easy deployment.

  15. Cinnamon/kotaemon - An open source tool based on retrieval enhancement generation (RAG) technology that can be used to chat with documents.

  16. labring/FastGPT - FastGPT is a knowledge platform based on large language models (LLMs) that provides multiple functions to easily develop and deploy Q&A systems.

  17. deepset-ai/haystack - An artificial intelligence orchestration framework for building large language model (LLM) applications for tasks like search augmented generation (RAG) and chatbots with advanced search methods.

  18. BerriAI/litellm - Python SDK and proxy server (LLM gateway) are able to call more than 100 large language model (LLM) APIs in OpenAI format, including APIs such as Bedrock, Azure, etc.

  19. flairNLP/flair - A very basic framework for high-level natural language processing.

  20. langchain-ai/langchainjs - Build a reasoning application with context awareness.

  21. xenova/transformers.js - State-of-the-art machine learning technology for networking allows running in browsers without servers? Transformers model.

  22. netease-youdao/QAnything - Q&A based on anything.

  23. h2oai/h2ogpt - Private chat with local GPT, supporting documents, images, videos and other content. It is 100% private, based on the Apache 2.0 protocol, supports oLLaMa, Mixtral, llama.cpp, etc., and there are examples in the given link.

  24. pathwaycom/llm-app - Ready-to-use cloud templates for retrieval enhanced generation (RAG), artificial intelligence pipelines and enterprise search, can process real-time data, docker-friendly and synchronize with various data sources.

  25. ludwig-ai/ludwig - A low-code framework for creating customized artificial intelligence models like large language models (LLMs) and neural networks.

  26. vercel/ai - Build AI-powered applications using React, Svelte, Vue and Solid.

  27. microsoft/promptflow - Build high-quality large language model (LLM) applications through prototyping, testing, production deployment and monitoring.

  28. Unstructured-IO/unstructured - Open source libraries and APIs for creating custom preprocessing pipelines such as tagging, training, or production tasks in machine learning.

  29. dataelement/bisheng - BISHENG is an open large language model (LLM) operation and maintenance platform for enterprise artificial intelligence applications. It has functions such as Generative Artificial Intelligence (GenAI) workflow, Retrieval Enhanced Generation (RAG) and other functions.

  30. Togethercomputer/OpenChatKit -

  31. llmware-ai/llmware - A unified framework for creating enterprise-level search-enhanced generation (RAG) pipelines using small dedicated models.

  32. leptonai/search_with_lepton - Quickly create a conversation-based search demo using Lepton AI.

  33. Deeptrain-Community/chatnio - Next-generation artificial intelligence one-stop B/C-end solution, supporting a variety of models and various functions.

  34. Chainlit/chainlit - Quickly build conversational AI in minutes.

  35. modelscope/modelscope - ModelScope turns the concept of model as a service into reality.

  36. deeppavlov/DeepPavlov - An open source library for deep learning end-to-end conversation systems and chatbots.

  37. langchain-ai/opengpts -

  38. TaskingAI/TaskingAI - An open source platform for developing native artificial intelligence applications.

  39. wenda-LLM/wenda - Wenda is a large language model (LLM) call platform designed to efficiently generate content in a specific environment, taking into account the limitations of personal and small and medium-sized enterprise computing resources as well as knowledge security and privacy issues.

  40. rustformers/llm - An unmaintained Rust library ecosystem for handling large language models. See the readme file for details.

  41. josStorer/RWKV-Runner - An 8MB fully automatic RWKV management and startup tool with an interface compatible with the OpenAI API. RWKV is a large language model that is fully open source and commercially available.

  42. langchain4j/langchain4j - Java version of LangChain.

  43. OpenBMB/ToolBench - An open platform for training, service and evaluation of large language models for tool learning (ICLR'24 Focus Paper).

  44. microsoft/FLAML - A fast library for automatic machine learning (AutoML) and parameter adjustment. There is also a link to join Discord (a chat software).

  45. microsoft/lmops - A general technology for realizing artificial intelligence capabilities through large language models (LLMs) and multimodal large language models (MLLMs).

  46. llm-workflow-engine/llm-workflow-engine - Power CLI and workflow manager for large language models (core packages).

  47. timescale/pgai - A set of tools for easier development of search-enhanced generation (RAG), semantic search, and other AI applications using PostgreSQL.

  48. FreedomIntelligence/LLMZoo - LLM Zoo is a project that provides data, models and evaluation benchmarks for large language models.

  49. casibase/casibase - AI Cloud is an open source retrieval enhanced generation (RAG) knowledge base similar to LangChain. It supports multiple models and has chatbots and administrative user interface (UI) demos.

  50. getzep/zep - Zep: The memory basis of your AI stack.

  51. leptonai/leptonai - A Python framework for simplifying the construction of artificial intelligence services.

  52. pezzolabs/pezzo - an open source, developer-first LLMOps platform for simplifying work in all aspects such as prompt design and version management.

  53. cheshire-cat-ai/core - Artificial intelligence proxy microservices.

  54. aurelio-labs/semantic-router - Ultra-high-speed artificial intelligence for multimodal data decision-making and intelligent processing.

  55. install-ai/vdp - Instill Core is a full-stack AI infrastructure tool for data, models, and pipeline orchestration that simplifies the process of building multiple AI-first applications.

  56. intelligence/intel-extension-for-transformers - Use state-of-the-art compression technology to quickly build your chatbots on your device and run large language models efficiently on Intel platforms.

  57. griptape-ai/griptape - A modular Python framework for artificial intelligence agents and workflows with thinking chain reasoning, tools and memory capabilities.

  58. run-llama/LlamaIndexTS - Data framework for large language model (LLM) applications with a focus on server-side solutions.

  59. Agenta-AI/agenta - An open source LLMOps platform that integrates prompt word playground, prompt word management, large language model (LLM) evaluation and large language model observability.

  60. marella/ctransformers - Python binding provided for Transformer models in C/C++ through the GGML library.

  61. devflowinc/trieve - An API-based infrastructure that integrates search, recommendation, search augmented generation (RAG) and analysis.

  62. YangLing0818/RPG-DiffusionMaster - [ICML 2024] Master text-to-image diffusion role-playing games (RPGs) through resubtitles, planning, and generation of multimodal large language models (LLMs).

  63. trypromptly/LLMStack - A codeless multi-proxy framework for building large language model (LLM) agents, workflows, and applications using your data.

  64. getzep/graphiti - Build and query dynamic knowledge graphs with time-awareness.

  65. KimMeen/Time-LLM - Official implementation of "Time-LLM: Time Series Prediction by Reprogramming Large Language Models" in ICLR 2024.

  66. floneum/floneum - Instant, controllable, and artificial intelligence model in the locally pre-trained Rust language.

  67. jina-ai/langchain-serve - Langchain application produced using Jina and FastAPI.

  68. SqueezeAILab/LLMCompiler - LLM Compiler (LLMCompiler) proposed at the 2024 International Conference on Machine Learning (ICML) is a large language model (LLM) compiler for parallel function calls.

  69. andreibondarev/langchainrb - Build an application powered by large language model (LLM) using Ruby.

  70. psychic-api/rag-stack - Deploy a private version of ChatGPT alternative in a virtual private cloud (VPC), connect to your organization's knowledge base, and support open source large language models (LLMs).

  71. DAGWorks-Inc/burr - Build applications for decision-making, such as chatbots, and manage them on your own infrastructure.

  72. IntelLabs/fastRAG - Efficient search enhancement and generation framework.

  73. sobelio/llm-chain - "llm-chain" is a powerful Rust crate (crate, which can be understood as a code base) used to build chains in large language models, implement text summary and completion of complex tasks.

  74. microsoft/windows-ai-studio -

  75. vercel/modelfusion - A TypeScript library for creating artificial intelligence applications.

  76. axflow/axflow - A TypeScript framework for artificial intelligence development.

  77. gabrielchua/RAGxplorer - An open source tool for visualizing your search enhancement generation (RAG).

  78. parthsarthi03/raptor - Official implementation of RAPTOR (Raptor) for tree-like organization retrieval through recursive abstraction processing.

  79. google/generative-ai-swift - The official Swift library for Google Gemini API.

  80. pinecone-io/canopy - Retrieval enhanced generation (RAG) framework and context engine powered by Pinecone.

  81. safevideo/autollm - Launches large language model (LLM) network applications based on search augmented generation (RAG) in seconds.

Model

  1. openai/whisper - Enable robust voice recognition through large-scale weak supervision.

  2. CompVis/stable-diffusion - A potential text-to-image diffusion model.

  3. facebookresearch/llama - Inference code for Llama models.

  4. xai-org/grok-1 - Public version of Grok is released.

  5. Stability-AI/stablediffusion - High resolution image synthesis using potential diffusion models.

  6. karpathy/nanoGPT - The easiest and fastest library for training/fine-tuning medium-sized GPT.

  7. TencentARC/GFPGAN - GFPGAN focuses on creating practical algorithms for face repair in real-life scenarios.

  8. lllyasviel/ControlNet -

  9. tatsu-lab/stanford_alpaca - Code and documentation for training Stanford Alpaca models and generating data.

  10. meta-llama/llama3 - Official Meta Llama 3 GitHub website.

  11. Stability-AI/generative-models - Generative Models of Stability AI

  12. lucidrains/vit-pytorch - Implementing Vision Transformer in PyTorch, using only one transformer encoder to reach the most advanced level (SOTA) in visual classification tasks.

  13. apple/ml-stable-diffusion - Stable Diffusion using Core ML on Apple Silicon.

  14. facebookresearch/codellama - Inference code for CodeLlama model.

  15. QwenLM/Qwen - The official code base of Qwen, Tongyi Qianwen is a large-scale language model proposed by Alibaba Cloud.

  16. AI4Finance-Foundation/FinGPT - FinGPT - Open source finance large-scale language model. Training model released in HuggingFace.

  17. state-spaces/mamba - Mamba SSM architecture

  18. BlinkDL/RWKV-LM - RWKV is a recurrent neural network (RNN) that performs well in large language models (LLM) and can be trained like a GPT transformer. It has the characteristics of excellent performance and linear time.

  19. CompVis/latent-diffusion - High resolution image synthesis using potential diffusion models.

  20. QwenLM/Qwen1.5 - Tongyi Qianwen 2.5 is a large-scale language model series developed by Alibaba Cloud Tongyi Qianwen team.

  21. lucidrains/DALLE2-pytorch - A neural network DALL for text and graphics synthesis that implements OpenAI update in PyTorch - E 2.

  22. NVIDIA/Megatron-LM - Continues the research on large-scale training of Transformer models.

  23. guoyww/AnimateDiff - The official implementation of AnimateDiff.

  24. databrickslabs/dolly - Databricks' Dolly is a large language model trained on the Databricks machine learning platform.

  25. mlfoundations/open_clip - an open source implementation of CLIP (contrast language-image pre-training).

  26. THUDM/CogVideo - Text and Image to Video Generation: CogVideoX (2024) and CogVideo (2023 International Conference on Characterization Learning)

  27. AIGC-Audio/AudioGPT - AudioGPT is related to understanding and generating voice, music, sound, and speaking avatars.

  28. nlpxucan/WizardLM - Large language models (LLMs) are built on Evol Insturct (Evol instructions): WizardLM (wizard language model), WizardCoder (wizard encoder), WizardMath (wizard mathematics).

  29. lucidrains/denoising-diffusion-pytorch - Implementing a denoising diffusion probability model in Pytorch.

  30. THUDM/CodeGeeX - CodeGeeX is an open source multilingual code generation model (KDD 2023).

  31. Vaibhavs10/insanely-fast-whisper -

  32. 01-ai/Yi - A series of large language models developed from the ground up by 01-ai developers.

  33. lucidrains/PaLM-rlhf-pytorch - Implementing human feedback reinforcement learning (RLHF) similar to ChatGPT on PaLM.

  34. HumanAIGC/EMO - Use the Audio2Video Diffusion Model to generate expressive portrait videos: vivid emoticon portraits.

  35. alembics/disco-diffusion - No translateable description is provided.

  36. openlm-research/open_llama - OpenLLaMA is an open source replica of LLaMA 7B of Meta AI with a loose license agreement and is trained on RedPajama dataset.

  37. OpenBMB/MiniCPM - MiniCPM3 - 4B, a large language model (LLM) at the edge, with better performance than GPT - 3.5 - Turbo.

  38. LargeWorldModel/LWM - Large world model for text and video modeling with millions of large contexts.

  39. LiheYoung/Depth-Anything - "Deep Everything: Unleashing the Power of Large-scale Unmarked Data", a basic model for depth estimation based on monocular images in the 2024 Conference on Computer Vision and Pattern Recognition (CVPR).

  40. openai/point-e - Point cloud diffusion is used for 3D model synthesis.

  41. google-research/text-to-text-transfer-transformer - Code for the paper "Exploring the limits of transfer learning with a unified text-to-text converter"

  42. Lightning-AI/lit-llama - The nanoGPT-based LLaMA language model implementation supports a variety of features such as fast attention mechanism, quantization, fine-tuning, and pre-training, and is licensed under the Apache 2.0 license.

  43. OpenGVLab/LLaMA-Adapter - At the 2024 International Conference on Learning Characterization (ICLR), LLaMA is fine-tuned in 1 hour to follow the instructions using 1.2 million parameters.

  44. NVIDIA/DALI - A GPU (Graphics Processor) acceleration library has highly optimized building blocks and execution engines for data processing to accelerate deep learning training and inference applications.

  45. allenai/OLMo - OLMo code for modeling, training, evaluation and reasoning.

  46. salesforce/CodeGen - CodeGen is an open source model series for program synthesis, trained on TPU-v4, comparable to OpenAI Codex.

  47. lucidrains/x-transformers - A highly focused, concise and complete transformer with experimental characteristics from multiple papers.

  48. SCIR-HI/Huatuo-Llama-Med-Chinese - Ben Cao (formerly known as Hua Tuo), a model library that uses Chinese medical knowledge to fine-tune instructions on large language models.

  49. luosiallen/latent-consistency-model - latent consistency model: synthesis of high-resolution images by small step inference.

  50. microsoft/BioGPT -

  51. google-research/simclr - SimCLRv2: Large self-supervised models are powerful semi-supervised learners.

  52. llSourcell/Doctor-Dignity - Doctor Dignity is a large language model (LLM) that can pass the US Medical License Examination (USMLE), be offline, cross-platform, and protect the privacy of healthy data.

  53. google-research/multinerf - Code release for Mip - NeRF 360, Ref - NeRF and RawNeRF.

  54. jaymody/picoGPT - A very small GPT implemented with NumPy - 2 version.

  55. google-research/albert - ALBERT is a streamlined version of BERT (Bidirectional Encoder Representations from Transformers) for self-supervised learning of language representation.

  56. project-baize/baize-chatbot - Use ChatGPT for hours with just one GPU.

  57. salesforce/CodeT5 - CodeT5 is designed to provide support for code understanding and generation for large language models (LLMs) of open code.

  58. facebookresearch/jepa - PyTorch code and model for self-supervised learning from videos through vision - based on joint embedding prediction architecture (V-JEPA).

  59. paperswithcode/galai - GALACTICA's model application program interface.

  60. dvlab-research/LongLoRA - Code and documentation for LongLoRA and LongAlpaca (ICLR 2024 verbal report).

  61. baaivision/Painter - Painter and SegGPT series: Visual basic model from Beijing Zhiyuan Artificial Intelligence Research Institute (BAAI).

  62. databricks/dbrx - Code examples and resources for large language model DBRX for Databricks.

  63. state-spaces/s4 - Structured state space sequence model.

  64. google-research/electra - ELECTRA pre-trains text encoder as a discriminator rather than a generator.

  65. EleutherAI/pythia - Center for research on interpretability and learning dynamics.

  66. ise-uiuc/magicoder - Magicoder (ICML'24) implements code generation through open source directives (OSS - Instruct).

  67. epfLLM/meditron - Meditron is an open source medical large language model.

  68. MetaGLM/FinGLM - FinGLM aims to build an open, public welfare and lasting financial model project to promote the development of "AI+finance" through open source.

  69. deepseek-ai/DeepSeek-LLM - DeepSeek Large Language Model (LLM): There will be an answer.

  70. allenai/scispacy - Complete spaCy pipelines and models for scientific/biomedical documentation.

  71. apple/ml-4m - 4M: Large-scale multimodal mask modeling.

  72. google-research/language - a shared library for Google's artificial intelligence language team's open source project.

  73. google/maxtext - A simple, high-performance, and scalable Jax large language model (LLM).

  74. netease-youdao/BCEmbedding - Open source model for retrieving enhanced generation (RAG) products (embedding and resorting).

  75. SHI-Labs/OneFormer - OneFormer in CVPR 2023 is a Transformer for universal image segmentation.

  76. google-research/FLAN -

  77. lxtGH/OMG-Seg - OMG-LLaVA and OMG-Seg code bases are related to CVPR-24 (Computer Vision and Pattern Recognition Conference - 2024) and NeurIPS-24 (Neural Information Processing System Conference - 2024).

  78. SHI-Labs/Versatile-Diffusion - Versatile Diffusion: A diffusion model that combines text, images and variants, published on preprint platform arXiv in 2022 and presented at the International Conference on Computer Vision (ICCV) in 2023.

  79. time-series-foundation-models/lag-llama - Lag - Llama: Methods for predicting basic models of probabilistic time series

  80. openai/lm-human-preferences - Code for papers on fine-tuning language model.

  81. IBM/Dromedary - Dromedary is designed to be a useful, ethical and reliable large-scale language model.

  82. dauparas/ProteinMPNN - The paper code called ProteinMPNN.

  83. SHI-Labs/Neighborhood-Attention-Transformer - Neighborhood Attention Transformer published in arxiv in 2022 and CVPR in 2023. In addition, the hollow neighborhood attention transformer published in arxiv in 2022.

  84. THUDM/SwissArmyTransformer - SwissArmyTransformer is a flexible and powerful library for developing Transformer variants.

  85. ctllllll/LLM-ToolMaker -

  86. Xwin-LM/Xwin-LM - Xwin-LM: A powerful, stable and reproducible large language model alignment.

  87. microsoft/ToRA - ToRA (for ICLR'24) is a large language model agent with integrated tools to solve difficult mathematical reasoning problems.

  88. SalesforceAIResearch/uni2ts - The universal time series prediction transformer is trained uniformly.

  89. replit/ReplitLM - ReplitLM model series inference code and configuration.

  90. HazyResearch/safari - Convolution in the context of sequence modeling.

AI List

  1. fighting41love/funNLP -

  2. linexjlin/GPTs - GPT related leak tips.

  3. e2b-dev/awesome-ai-agents - a series of autonomous agents for artificial intelligence.

  4. eugeneyan/open-llms - A list of open large language models (LLMs) available for commercial use.

  5. Shubhamsaboo/awesome-llm-apps - A great set of large language model (LLM) applications with search enhanced generation (RAG) capabilities that use OpenAI, Anthropic, Gemini and open source models.

  6. RUCAIBox/LLMSurvey - A Survey of Large Language Models The official GitHub page for this survey paper.

  7. WooooDyy/LLM-Agent-Paper-List - A list of papers on the rise and potential of agents based on large language models: A review, written by Xi Zhiheng (transliteration) and others

  8. steven2358/awesome-generative-ai - List of contemporary generative artificial intelligence projects and services.

  9. wgwang/LLMs-In-China - Large-scale Chinese model.

  10. lonePatient/awesome-pretrained-chinese-nlp-models - A series of high-quality Chinese pretrained models, large models, multimodal models and large language models.

  11. tensorchord/Awesome-LLMOps - A large list of excellent LLMOps tools selected for developers.

  12. opendilab/awesome-RLHF - A continuously updated list of reinforcement learning resources based on human feedback.

  13. DSXiangLi/DecryptPrompt - Summary of the application of Prompt and Large Language Models (LLM), open source data and models, and artificial intelligence generated content (AIGC).

  14. FreedomIntelligence/Medical_NLP - Medical Natural Language Processing Competition, Datasets, Large Models and Papers.

  15. archinetai/audio-ai-timeline - A timeline for the latest audio-generated AI models starting in 2023.

  16. chiphuyen/aie-book - Resources of AI engineers and auxiliary materials for Artificial Intelligence Engineering (Chip Hughn, 2025).

  17. EgoAlpha/prompt-in-context-learning - A high-quality resource for context learning and prompt engineering. Master large language models with the latest updates such as ChatGPT, GPT-3 and FlanT5.

  18. taranjeet/awesome-gpts - A collection of all GPTs (generated pretrained converters) produced by the community.

  19. cfahlgren1/natural-sql - A series of high-performance text-to-SQL large language models.

  20. yokoffing/ChatGPT-Prompts - Prompt word management for ChatGPT and Bing AI.

Reasoning Optimization

  1. ggerganov/llama.cpp - Large Language Model (LLM) Inference in C/C++.

  2. ggerganov/whisper.cpp - A ported version of the Whisper model of OpenAI written in C/C++.

  3. karpathy/llm.c - Train a large language model (LLM) using simple raw C/CUDA.

  4. Mozilla-Ocho/llamafile - Use only one file to distribute and run large language models (LLMs).

  5. unslothai/unsloth - Increase fine-tuning speeds of Llama 3.3, Mistral, Phi, Qwen 2.5 and Gemma large language models (LLMs) by 2-5 times while using 70% less memory.

  6. mlc-ai/mlc-llm - A universal large language model (LLM) deployment engine with machine learning (ML) compilation capabilities.

  7. karpathy/llama2.c - Infer Llama 2 in a single pure C file.

  8. Dao-AILab/flash-attention - Fast and efficient precise attention mechanism, both fast and memory-saving.

  9. openai/triton - a development library for the Triton language and compiler.

  10. microsoft/BitNet - Official inference framework for 1-bit large language models (LLMs).

  11. ggerganov/ggml - A tensor library for machine learning.

  12. NVIDIA/TensorRT - NVIDIA TensorRT is a software development toolkit (SDK) for high-performance deep learning reasoning on NVIDIA GPUs. This repository has its open source components.

  13. bigscience-workshop/petals - Run large language models (LLMs) at home in a BitTorrent-like manner, fine-tuning and inference speeds up to 10 times faster than offloading.

  14. NVIDIA/TensorRT-LLM - TensorRT-LLM provides an easy-to-use Python API for defining large language models (LLMs) and building an optimized TensorRT engine for efficient GPU inference, and has a Python and C++ runtime for creating Python and C++ runtimes to execute components of these engines.

  15. intelligence-analytics/BigDL - Accelerate local large language model (LLM) inference and fine-tuning on Intel XPUs and integrate with various related frameworks.

  16. intelligence-analytics/ipex-llm - Accelerate local large language model (LLM) inference and fine-tuning on Intel XPU (accelerated processor for Intel architecture) and integrate with a variety of tools.

  17. TimDettmers/bitsandbytes - A large language model accessible through K-bit quantization of PyTorch.

  18. google/gemma.cpp - A lightweight standalone C++ inference engine for Google Gemma models.

  19. NVIDIA/cutlass - CUDA template for linear algebra subroutines.

  20. pytorch-labs/gpt-fast - Use less than 1000 lines of Python code to implement a simple and efficient native PyTorch transformer for text generation.

  21. PanQiWei/AutoGPTQ - A large language model quantization package based on GPTQ algorithm, easy to use and user-friendly interface.

  22. turboderp/exllamav2 - A fast inference library for running large language models (LLMs) locally on common consumer-grade GPUs.

  23. OpenNMT/CTranslate2 - Fast transformer model inference engine.

  24. ztxz16/fastllm - A pure C++ full-platform large language model (LLM) acceleration library that supports Python calls. It can make the single-card ChatGLM-6B-level model reach more than 10,000 tokens per second, support GLM, Llama, Moss basic models, and run smoothly on mobile devices.

  25. qwopqwop200/GPTQ-for-LLaMa - Quantize LLaMA into 4 bits using GPTQ.

  26. VainF/Torch-Pruning - [CVPR 2023] DepGraph: Pruning for arbitrary structures.

  27. turboderp/exllama - A rewrite version of the HF Transformers version of Llama for quantizing weights, with more memory efficiency.

  28. lucidrains/vector-quantize-pytorch - Vector (and scalar) quantization in PyTorch

  29. mit-han-lab/llm-awq - AWQ: Activation-aware weight quantification (method) for compression and acceleration of large language models won the 2024 MLSys Best Paper Award.

  30. Jittor/JittorLLMs - The Jittor model inference library has the characteristics of high performance, low configuration requirements, good Chinese support and portability.

  31. FasterDecoding/Medusa - Medusa: A simple framework that accelerates the generation of large language models (LLMs) through multiple decoding terminals.

  32. intel/neural-compressor - SOTA low-bit LLM quantization (including INT8/FP8/INT4/FP4/NF4) and sparseness are leading model compression techniques for TensorFlow, PyTorch and ONNX runtimes.

  33. neuralmagic/sparseml - A library that facilitates easy application of sparseness to neural networks, resulting in faster and smaller models.

  34. IST-DASLab/gptq - The paper code on the precise post-training quantization of Generative Pre-training Transformers (GPT) at the 2023 International Conference on Learning Characterization (ICLR), called "GPTQ".

  35. HazyResearch/ThunderKittens - Tile primitives for fast kernels.

  36. uTensor/uTensor - A micro machine learning artificial intelligence inference library.

  37. pytorch-labs/ao - PyTorch native quantization and sparseness for training and reasoning.

  38. saharNooby/rwkv.cpp - INT4/INT5/INT8 and FP16 inference on CPUs for RWKV language model.

  39. mit-han-lab/smoothquant - SmoothQuant: Accurate and efficient post-training quantization of large language models.

  40. Lightning-AI/lightning-thunder - Thunder is a PyTorch source-to-source compiler that can increase model speed by up to 40% and use different hardware actuators on multiple GPUs.

  41. pytorch-labs/segment-anything-fast - A Segment for batch offline reasoning - Anything version.

  42. Vahe1994/AQLM - Official PyTorch library, containing two papers on extreme compression of large language models: one is quantized by addition (https://arxiv.org/pdf/2401.06118.pdf), and the other is PV - Fine Tuning ( https://arxiv.org/abs/2405.14852).

  43. hao-ai-lab/LookaheadDecoding - Use lookahead decoding to break the sequential dependencies in large language model (LLM) inference (ICML 2024).

  44. horseee/LLM-Pruner - [NeurIPS 2023] Large language models (such as Llama - 3/3.1, Llama - 2, LLaMA, etc.) structure pruned LLM - Pruner.

  45. kuleshov/minillm - MiniLLM is a system that runs modern large language models (LLM) on consumer-grade GPUs in a minimal way.

Information aggregation

  1. binary-husky/gpt_academic - Provides practical interactive interface for large language models (LLMs) such as GPT/GLM, especially optimizes paper reading, polishing and writing experience. Supports multiple functions and integrates multiple models.

  2. imartinez/privateGPT - Use GPT to interact with documents privately, no data breach.

  3. Mintplex-Labs/anything-llm - Desktop and Docker AI applications are integrated, with built-in retrieval enhancement generation (RAG) and AI proxy.

  4. khoj-ai/khoj - It is a self-hosted second brain of artificial intelligence that can get answers from various sources, build custom agents, arrange automated tasks, and conduct research, and can transform large language models into personal artificiality for free intelligent.

  5. PromtEngineer/localGPT - Chat privately with local documents through GPT model, data does not leave the device.

  6. kaixindelele/ChatPaper - Use ChatGPT to accelerate scientific research, including summary of arXiv papers, professional translation, polishing, peer review and response to peer review.

  7. assafelovic/gpt-researcher - Autonomous agent based on large language model (LLM) conducts local and web research on any topic and creates a comprehensive report with citations.

  8. arc53/DocsGPT - Document chatbots are able to chat with data, deploy privately, and integrate knowledge into AI workflows for sharing.

  9. mayooear/gpt4-pdf-chatbot-langchain - GPT4 and LangChain chatbot for large PDF documents.

  10. danswer-ai/danswer - Gen - AI Chat for Teams is like ChatGPT, but can gain special knowledge of the team.

  11. josStorer/chatGPTBox - Deeply integrate ChatGPT into your browser. Everything you need is here.

  12. facebookresearch/nougat - Application of Nougat in neurooptical understanding of academic documents.

  13. bhaskatripathi/pdfGPT - PDF GPT is able to interact with PDF content through GPT capabilities and is an effective open source solution for turning PDF into a chatbot.

  14. whitead/paper-qa - High-precision search-enhanced generation (RAG) technology for answering cited questions based on scientific literature.

  15. weaviate/Verba - A search-enhanced generation (RAG) chatbot powered by Weaviate.

  16. run-llama/rags - Use your data to build ChatGPT, all in natural language.

  17. MuiseDestiny/zotero-gpt - GPT meets Zotero.

  18. madawei2699/myGPTReader - A community-driven method of interacting with artificial intelligence robots using ChatGPT.

  19. swirlai/swirl-search - AI Search & RAG can obtain instant answers from company knowledge in numerous applications while ensuring data is secure and fast deployment.

  20. dvorka/mindforger - a thought note and a Markdown editor.

  21. kha-white/manga-ocr - mainly used for Japanese character optical character recognition in Japanese comics.

  22. nlmatics/llmsherpa - Developer application program interface (API) for accelerating large language model (LLM) projects.

  23. ucbepic/docetl - A data processing and ETL (decision, transformation, loading) system driven by autonomous language model (LLM).

  24. KnowledgeCanvas/knowledge - Knowledge is a tool used for various operations (such as saving, search, etc.) of websites, documents, and files.

  25. rotatemweiss57/gpt-newspaper - An autonomous GPT-based agent that can create personalized newspapers based on user preferences.

  26. nlmatics/nlm-ingestor - This repository provides server-side code for llmsherpa API connections and parsers for different file formats.

  27. kha-white/mokuro - Read Japanese comics with selectable text in your browser.

  28. BruceMacD/chatd - Chat with your documents via local AI.

  29. akshata29/entaoai - Use your own data to chat and ask questions. Quickly upload enterprise data to use OpenAI services to chat and ask questions about uploaded data.

Code Assistant

  1. abi/screenshot-to-code - Insert a screenshot and convert it into concise code (HTML/Tailwind/React/Vue).

  2. gpt-engineer-org/gpt-engineer - A terminal-based platform for experiencing artificial intelligence software engineers, unlike https://gptengineer.app.

  3. OpenDevin/OpenDevin - OpenHands: implement more functions with less code.

  4. Pythongora-io/gpt-pilot - The first person to truly become a developer in the field of artificial intelligence.

  5. getcursor/cursor - Artificial Intelligence Code Editor.

  6. OpenBMB/ChatDev - Create customized software based on natural language ideas through multi-agent collaboration with large language models (LLM).

  7. paul-gauthier/aider - Aider is artificial intelligence pairing programming in terminals.

  8. TabbyML/tabby - A self-hosted AI coding assistant.

  9. continuedev/continue - Continue is an open source artificial intelligence code assistant. It can be connected to models and contexts for custom autocomplete and chat in VS Code and JetBrains.

  10. stageai/devika - Devika is an agent artificial intelligence software engineer who can understand human instructions, break down instructions, conduct research and write code. Its goal is to be an open source alternative to Devin developed by Cognition AI, and there is no official website.

  11. emilwallner/Screenshot-to-code - A neural network for converting design models into static websites.

  12. fauxpilot/fauxpilot - FauxPilot is an open source alternative to GitHub Copilot servers.

  13. eosphoros-ai/DB-GPT - Native artificial intelligence data application development framework with AWEL (Agistrator Workflow Expression Language) and agents.

  14. princeton-nlp/SWE-agent - SWE - Agent automatically fixes GitHub issues using GPT-4 or other language models, and can also be used for offensive cybersecurity or competitive programming challenges. [NeurIPS 2024]

  15. Sinaptik-AI/pandas-ai - Interact with various databases (SQL, CSV, etc.) and use PandasAI to perform conversational data analysis through large language models (LLMs) and search augmented generation (RAG).

  16. vanna-ai/vanna - Interact with your SQL database. Generate accurate text-to-SQL statements through large language models (LLM) using search-based generation (RAG) technology.

  17. ShishirPatil/gorilla - Gorilla: Training and evaluation of large language models for function calls (tool calls).

  18. codota/TabNine - AI code completion refers to the function of an AI system to provide programmers with suggestions or complete code segments.

  19. TheR1D/shell_gpt - A command line productivity tool powered by AI large language models such as GPT-4, helps to complete tasks faster and more efficiently.

  20. Nutlope/aicommits - A command line interface (CLI) that uses artificial intelligence to write git submission information for you.

  21. GreyDGL/PentestGPT - A penetration testing tool empowered by GPT.

  22. joshpxyne/gpt-migrate - Easily migrate your codebase between frameworks or languages.

  23. kuafuai/DevOpsGPT - An artificial intelligence-powered software development multi-agent system combines large language models (LLMs) with DevOps tools to transform natural language requirements into runnable software, supporting any development language and extending existing code.

  24. di-sukharev/opencommit - The GPT wrapper for Git is able to generate commit messages in 1 second using a large language model (LLM), works well with Claude 3.5 and supports local models.

  25. sqlchat/sqlchat - A SQL client and editor based on chat functions for the next decade.

  26. Exafunction/codeium.vim - A free and super fast Copilot alternative for Vim and Neovim.

  27. varunshenoy/GraphGPT - Inferring knowledge graphs from unstructured text using GPT - 3.

  28. Nutlope/llamacoder - Open source product made by Claude, built using Llama 3.1 405B.

  29. mckaywrigley/ai-code-translator - leverages artificial intelligence to translate code between different languages.

  30. shobrook/adrenaline - Interact with and visualize the code base.

  31. QwenLM/Qwen2.5-Coder - Tongyi Qianwen 2.5 - Coder is the code version of Tongyi Qianwen 2.5. Tongyi Qianwen 2.5 is a large language model series developed by Alibaba Cloud Tongyi Qianwen team.

  32. ricklamers/gpt-code-ui - An open source implementation of OpenAI's ChatGPT code interpreter.

  33. gofireflyio/aiac - Artificial Intelligence Infrastructure Code Generator.

  34. defog-ai/sqlcoder - The most advanced language model for converting natural language problems into SQL queries.

  35. gptscript-ai/gptscript - Build an artificial intelligence assistant for interacting with your system.

  36. RootbeerComputer/backend-GPT -

  37. mpoon/gpt-repository-loader - Converts the codebase to a large language model (LLM) prompt-friendly format created primarily by GPT-4.

  38. Canner/WrenAI - An open source artificial intelligence agent enables data and product teams to interact and chat with data through text-to-SQL (Text - to - SQL) to create charts, spreadsheets, reports and business intelligence (BI).

  39. nus-apr/auto-code-rover - An autonomous software engineer who understands the structure of the project is dedicated to autonomous program improvement. It completed a percentage of tasks in both benchmarks and costs less than $0.7 per task.

  40. fern-api/fern - Enter OpenAPI and output the SDK (Software Development Kit) and documentation.

  41. georgia-tech-db/evadb - An application database system powered by artificial intelligence.

  42. AbanteAI/mentat - Mentat - Artificial Intelligence Coding Assistant

  43. emcf/engshell - An English language shell (shell) powered by large language models (LLMs) that can be used on any operating system (OS).

  44. AI-Citizen/SolidGPT - A developer artificial intelligence for searching for roles.

  45. context-labs/autodoc - An experimental toolkit that automatically generates codebase documents using large language models (LLMs).

  46. knuckleswtf/scribe - Generate API documentation for humans from the Laravel code base.

  47. jina-ai/dev-gpt - Your virtual development team can be a group of remote developers working to develop software or other projects through virtual collaboration.

  48. Pythongora-io/pythagora - Generate automated tests for Node.js applications using large language models (LLMs) without the need for developers to write any code.

  49. eli64s/readme-ai - A readme generator powered by artificial intelligence.

  50. mattzcarey/code-review-gpt - Code review using large language models (GPT4, Sonnet 3.5) and embeddings improve code quality and detect errors in the pre-production phase, and continuous integration with Github/GitLab/Azure DevOps (CI )integrated.

  51. smallcloudai/refact - WebUI for fine-tuning and self-hosting open source large coding language models.

  52. eyelonmiz/react-agent - Open source React.js Autonomous Large Language Model (LLM) proxy.

  53. gorilla-llm/gorilla-cli - Large language models (LLMs) for command line interfaces (CLIs).

  54. huggingface/llm-vscode - Development powered by large language model (LLM) in VSCode.

  55. peterw/Chat-with-Github-Repo - This repository has two Python scripts for creating chatbots through Streamlit, OpenAI GPT-3.5-turbo and Activeloop's Deep Lake.

  56. paralleldrive/sudolang-llm-support - SudoLang Large Language Model (LLM) support in Visual Studio Code.

  57. ricklamers/shell-ai - A command line interface (CLI) powered by LangChain for generating and running shell commands.

  58. google/oss-fuzz-gen - via OSS - Fuzz driven by large language model (LLM).

  59. kantord/SeaGOAT - Semantic code search engine that adopts a locally preferred approach.

  60. OpenAutoCoder/Agentless - An agentless method for automatically solving software development problems.

  61. ferrislucas/promptr - Promptr is a command line interface (CLI) tool that can modify the code base in easy-to-understand English.

AI Tutorial

  1. microsoft/generative-ai-for-beginners - 21 courses to start building using generative artificial intelligence. Link: https://microsoft.github.io/generative-ai-for-beginners/

  2. openai/openai-cookbook - OpenAI API usage examples and guides.

  3. mlabonne/llm-course - A course in-depth with large language models (LLMs) containing roadmaps and Colab notebooks.

  4. rasbt/LLMs-from-scratch - Gradually implement a ChatGPT-like large language model (LLM) in PyTorch from scratch.

  5. lutzroeder/netron - A visualization tool for neural networks, deep learning and machine learning models.

  6. datawhalechina/prompt-engineering-for-developers - Introduction to large language model (LLM) for developers, Chinese version of the Chinese version of the Ng model series course.

  7. liguodongiot/llm-action - This project aims to share technical principles and practical experience related to large models (large model engineering and large-scale model application implementation).

  8. stas00/ml-engineering - Open book on machine learning engineering.

  9. mikeroyal/Self-Hosting-Guide - Self-hosting Guide: Host and manage software applications locally by individuals or organizations, covering cloud, large language models (LLMs) and more.

  10. hua1995116/awesome-ai-painting - Collection of artificial intelligence painting materials, including domestic and foreign platforms, tutorials and news, such as Stable diffusion, AnimateDiff, Stable Cascade, Stable SDXL Turbo SDXL Turbocharged).

  11. Mooler0410/LLMsPracticalGuide - A list of practical guide resources for large language models (LLMs) including LLMs trees, examples, and papers.

  12. GoogleCloudPlatform/generative-ai - Sample code and notebook for Generative AI using Gemini in Vertex AI on Google Cloud.

  13. kyrolabs/awesome-langchain - Excellent list of tools and projects using LangChain framework.

  14. microsoft/DeepSpeedExamples - DeepSpeed ​​sample model.

  15. huggingface/alignment-handbook - A robust approach to bringing language models into line with human and artificial intelligence preferences.

  16. trigaten/Learn_Prompting - A guide to prompt engineering, generative artificial intelligence and large language models (LLM), provided by Learn Prompting. Join its Discord (a chat software) to get the biggest tips for engineering learning community.

  17. bbycroft/llm-viz - 3D visualization of GPT-style large language models.

  18. ray-project/llm-numbers - Numbers that every large language model (LLM) developer should know.

  19. luban-agi/Awesome-AIGC-Tutorials - Selected tutorials and resources for large language models, artificial intelligence painting, etc.

  20. georgezouq/awesome-ai-in-finance - A range of carefully selected excellent large language models (LLMs), deep learning strategies and tools in the financial market.

  21. howl-anderson/unlocking-the-power-of-llms - Make ChatGPT a powerful productivity tool using Prompts and Chains. Unlocking the potential of large language models (LLMs).

  22. ashishpatel26/LLM-Finetuning - Finetuning large language models (LLM) using PEFT (parameter efficient fine-tuning).

  23. ray-project/llm-applications - A comprehensive guide to the development of large language model (LLM) applications based on search augmented generation (RAG).

  24. premAI-io/state-of-open-source-ai - In the chaotic and fast-paced world of open source innovation, you need to have clear ideas.

  25. pionxzh/chatgpt-exporter - Export and share your ChatGPT chat history.

  26. ianand/spreadsheets-are-all-you-need-

  27. majacinka/crewai-experiments - Experiment using native models and models accessible through application program interfaces (APIs).

  28. thu-vu92/local-llms-analyse-finance -

Workflow automation

  1. KillianLucas/open-interpreter - The computer's natural language interface.

  2. StanGirard/quivr - An opinionated search-enhanced generation (RAG) technology used to integrate generative artificial intelligence (GenAI) into applications, focusing on products. Easily integrated and customized in existing products, and is versatile in large language models (LLM), vector storage and files.

  3. danielmiessler/fabric - Fabric is an open source AI-enhanced human framework with a modular structure that solves problems with crowdsourcing AI tips.

  4. openai-translator/openai-translator - Browser and desktop application that uses the ChatGPT API for translation.

  5. Skyvern-AI/skyvern - Automating browser-based tasks using large language models (LLMs) and computer vision technologies.

  6. activepieces/activepieces - Your most friendly open source artificial intelligence automation tool. It is a workflow automation tool with more than 200 integrated worksflow automation tools and is an alternative to Zapier in enterprise automation.

  7. OthersideAI/self-operating-computer - A framework for multimodal models to operate computers.

  8. microsoft/UFO - A Windows operating system interactive proxy focused on the user interface.

  9. yihong0618/bilingual_book_maker - Use artificial intelligence translation methods to create bilingual epub books.

  10. lavague-ai/LaVague - Large action model framework for developing artificial intelligence network agents.

  11. aisingapore/TagUI - A free robotic process automation (RPA) tool developed by Singapore’s artificial intelligence.

  12. openchatai/OpenCopilot - Language to Behavior Engine

  13. KillianLucas/01 - Top open source voice interface for desktop, mobile and ESP32 chips.

  14. katanaml/sparrow - Data processing using machine learning, large language models (LLM), and large language models based on vision.

  15. xlang-ai/OpenAgents - OpenAgents in 2024: an open platform for wild language agents.

  16. BAAI-Agents/Cradle -

  17. Cormanz/smartgpt - A program that enables large language models (LLMs) to complete complex tasks with the help of plug-ins.

  18. fiatrete/OpenDAN-Personal-AI-OS - OpenDAN is an open source personal artificial intelligence operating system that integrates various artificial intelligence modules for personal use.

  19. n4ze3m/page-assist - Use locally run AI models to assist in web browsing.

  20. OS-Copilot/FRIDAY -

  21. andrewnguonly/Lumos - A search-enhanced generation (RAG) large language model (LLM) co-pilot powered by the local large language model (LLM) for web browsing.

  22. Dicklesworthstone/swiss_army_llama - A FastAPI service that performs semantic text search through pre-computed embedding, similarity metrics, and text support file types through texttract.

AI robot

  1. lencx/ChatGPT - ChatGPT desktop application for Mac, Windows and Linux systems.

  2. LAION-AI/Open-Assistant - OpenAssistant is a chat-based assistant that can understand tasks, interact with third-party systems and retrieve information dynamically.

  3. zhayujie/chatgpt-on-wechat - A chatbot based on large models, supporting multiple platforms (WeChat official account, enterprise WeChat application, Feishu, DingTalk, etc.), multiple models (GPT3.5/GPT - 4o/GPT - o1/Claude/Wenxin Yiyan/iFlytek/Tongyi Qianwen/GLM-4/Claude/Qimi/Link Love), able to process text, voice and pictures, access the operating system and The Internet and supports customized enterprise intelligent customer service based on its own knowledge base.

  4. Chanzhaoyu/chatgpt-web - A ChatGPT demo webpage built using Express and Vue3.

  5. janhq/jan - Jan is an open source ChatGPT alternative that runs completely offline on your computer.

  6. Bin-Huang/chatbox - A friendly desktop client application for AI models/large language models (such as GPT, Claude, Gemini, Ollama, etc.).

  7. joonspk-research/generative_agents - Generative agent: interactive simulation of human behavior.

  8. Unity-Technologies/ml-agents - Unity ML - Agents Toolkit is an open source project for training agents in games and simulations through deep reinforcement learning and imitation learning.

  9. transitive-bullshit/chatgpt-api - A standard library of AI agents compatible with any large language model (LLM) and TypeScript AI Software Development Kit (SDK).

  10. leon-ai/leon - Leon is your open source personal assistant.

  11. xcanwin/KeepChatGPT - This is a plugin that enhances the security and efficiency of ChatGPT data. It offers many free innovative features to provide a better AI experience.

  12. lss233/chatgpt-mirai-qq-bot - One-click deployment! True AI chatbot, supporting multiple platforms and multiple functions.

  13. getumbrel/llama-gpt - A self-hosted offline chatbot like ChatGPT, powered by Llama 2, is private, has no data left the device, and now supports Code Llama.

  14. sfyc23/EverydayWechat - WeChat Assistant: 1. Send customized messages to friends (girlfriends) regularly every day. 2. The robot automatically responds to friends. 3. Group assistant functions (such as garbage classification query, weather, calendar, real-time movie box office, express logistics, PM2.5, etc.).

  15. BlinkDL/ChatRWKV - ChatRWKV is an open source language model like ChatGPT, but powered by RWKV, a 100% recurrent neural network.

  16. ztjhz/BetterChatGPT - A great user interface for ChatGPT, available on websites and on a variety of operating systems including Windows, MacOS and Linux.

  17. a16z-infra/ai-town - A starter kit that follows the MIT license for building AI towns where AI characters live, chat and socialize.

  18. memochou1993/gpt-ai-assistant - OpenAI, LINE and Vercel combine to form the GPT AI assistant.

  19. miurla/morphic - A search engine powered by artificial intelligence and with a generative user interface.

  20. interstellard/chatgpt-advanced - WebChatGPT is a browser extension that enhances ChatGPT prompts through network results.

  21. linyiLYi/street-fighter-ai - This is an artificial intelligence agent for Street Fighter II Champion Edition.

  22. vincelwt/chatgpt-mac - ChatGPT for Mac resides in your menu bar.

  23. camel-ai/camel - CAMEL: The first and best multi-agent framework for discovering the expansion law of agents. (https://www.camel - ai.org)

  24. MineDojo/Voyager - Open embossed agents associated with large language models.

  25. a16z-infra/companion-app - AI partner with memory: a lightweight stack for creating and hosting your own AI partner.

  26. ConnectAI-E/Feishu-OpenAI - Feishu (combined with GPT-4, GPT-4V, DALL·E-3 and Whisper) provides a great work experience, including voice conversations, role-playing, multi-topic discussions, image creation, Table analysis and document export.

  27. simonw/llm - Access large language models via command line.

  28. sigoden/aichat - A large language model (LLM) command line interface (CLI) tool integrating Shell Assistant, Chat - REPL, Retrieval Enhanced Generation (RAG), Artificial Intelligence Tools and Agents, which can access OpenAI , Claude and other platforms.

  29. lencx/nofwl - No FireWall (NoFWL) desktop application.

  30. Kent0n-Li/ChatDoctor -

  31. xtekky/chatgpt-clone - ChatGPT interface with improved user interface.

  32. deep-diver/LLM-As-Chatbot - Large Language Model (LLM) as a chatbot service.

  33. gragland/chatgpt-chrome-extension - A ChatGPT Chrome extension that integrates ChatGPT into every text box on the internet.

  34. ohmplatform/FreedomGPT - This code base is for a React with a chat interface - Electron (a technology that uses the React framework to build desktop applications) application that runs FreedomGPT large language models locally on Mac and Windows systems ( LLM).

  35. SoraWebui/SoraWebui - SoraWebui is an open source Sora network client that easily creates videos from text using OpenAI's Sora model.

  36. karthink/gptel - A simple Emacs client using large language models.

  37. a16z-infra/llama2-chatbot - LLaMA v2 chatbot.

  38. ItsPi3141/alpaca-electron - The easiest way to run alpaca and other native large language models (LLMs) based on LLaMA on your own PC.

  39. opendilab/DI-star - An artificial intelligence platform for StarCraft II, with large-scale distributed training and master-level agents.

  40. jncraton/languagemodels - Explore large language models with 512MB of memory.

  41. SamurAIGPT/Camel-AutoGPT - Introduction to CAMEL, a role-playing method for large language models (LLMs) and automatic agents (auto-agents). It enables agents to collaborate and have potential in multiple fields.

  42. Syan-Lin/CyberWaifu - A realistic chatbot composed of large language model (LLM) + voice synthesis (TTS), a QQ robot that supports emojis, QQ emoticons and Internet search.

Multimodal model

  1. PaddlePaddle/PaddleOCR - A great multilingual optical character recognition (OCR) toolkit based on PaddlePaddle. They are practical, ultra-lightweight, support over 80 languages ​​and can be used on a variety of devices.

  2. suno-ai/bark - A generative audio model prompted by text.

  3. openai/CLIP - CLIP (Contrast Language - Image Pretraining): Predict the most relevant text fragments for images.

  4. hpcaitech/Open-Sora - Open-Sora: Enable everyone to perform efficient video production in a democratic way.

  5. Visual instruction adjustment (LLaVA) in haotian-liu/LLaVA - NeurIPS'23 Oral (Personal Report of the Conference on Neural Information Processing Systems 2023) is designed to gain even stronger capabilities at the GPT-4V level.

  6. fishaudio/fish-speech - state-of-the-art open source speech synthesis (TTS) technology.

  7. borisdayma/dalle-mini - DALL·E Mini generates images based on text prompts.

  8. google-deepmind/alphafold - Open source code for AlphaFold 2.

  9. OpenBMB/OmniLMM - MiniCPM - V 2.6 is a multimodal large language model (MLLM) that reaches the GPT-4V level for single images, multiple images and videos on mobile phones.

  10. PKU-YuanGroup/Open-Sora-Plan - This project aims to reproduce Sora (OpenAI's text-to-video model) and hopes that the open source community will contribute.

  11. openai/shap-e - Generate 3D objects based on text or image.

  12. facebookresearch/seamless_communication - The basic model for state-of-the-art speech and text translation.

  13. openai/DALL-E - The PyTorch software package is used for discrete variational autoencoder (VAE) in DALL·E.

  14. google-research/vision_transformer -

  15. magic-research/magic-animate - MagicAnimate in CVPR 2024 uses a diffusion model to achieve time-consistent human image animation.

  16. ashawkey/stable-dreamfusion - Use Neural Radiation Field (NeRF)+Diffusion technology for text-to-3D, image-to-3D and mesh export.

  17. lucidrains/imagen-pytorch - Implement Google's text-to-image neural network Imagen in Pytorch.

  18. openai/jukebox - Code for the paper "Jukebox: A Music Generation Model".

  19. deep-floyd/IF -

  20. netease-youdao/EmotiVoice - EmotiVoice?: A voice synthesis (TTS) engine with multiple tones and prompt control.

  21. IDEA-Research/GroundingDINO - The official implementation of the paper "Grounding DINO: Combining DINO with Basic Pre-training for Open Set Object Detection" in ECCV 2024.

  22. FoundationVision/VAR - Official implementation of the NeurIPS 2024 oral report "Visual Autoregressive Modeling: Scalable Image Generation through Next-Scale Prediction". This is an extremely simple, user-friendly state-of-the-art code base for autoregressive image generation.

  23. threestudio-project/threestudio - A framework for uniformly generating 3D content.

  24. openai/guided-diffusion -

  25. THUDM/CogVLM - A state-of-the-art open visual language model, a multimodal pre-trained model.

  26. openai/consistency_models - Official library for consistency model.

  27. levihsu/OOTDiffusion - OOTDiffusion: Clothing fusion based on latent diffusion is used for controllable virtual trial-off-off-off.

  28. clovaai/donut - Official implementation of Donut (Document Understanding Transformer without Optical Character Recognition) and SynthDoG (Synthetic Document Generator) in ECCV 2022.

  29. google/gemma_pytorch - The official PyTorch implementation of Google Gemma model.

  30. QwenLM/Qwen-VL - Tongyi Qianwen-VL (an official warehouse of chat and pre-trained large visual language model of Alibaba Cloud).

  31. yl4579/StyleTTS2 - StyleTTS 2 aims to achieve human-level text-to-speech conversion through style diffusion and adversarial training using large speech language models.

  32. snakers4/silero-models - The Silero model is a pre-trained model for speech to text, text to speech and text enhancement, which are very simple to make.

  33. salesforce/BLIP - PyTorch code for BLIP (Bootstrapping Language - Image Pre - training), BLIP is used to achieve unified vision-language understanding and generation through bootstrapping language - image pre-training.

  34. google-deepmind/alphageometry -

  35. metavoiceio/metavoice-src - A human-like, expressive text-to-speech (TTS) basic model.

  36. Luodian/Otter - Otter is a multimodal model based on OpenFlamingo, trained on MIMIC - IT datasets with better instruction compliance and context learning capabilities.

  37. NExT-GPT/NExT-GPT - NExT - GPT (an arbitrary to any multimodal large language model).

  38. openai/improved-diffusion - Releases a model for improved denoising diffusion.

  39. X-PLUG/MobileAgent - Mobile Devices - Agent: A powerful family of mobile device operation assistants.

  40. dvlab-research/MiniGemini - "Mini - Gemini: Exploring the Potential of Multimodal Visual Language Models" Official Knowledge Base.

  41. lucidrains/musiclm-pytorch - Implement Google's state-of-the-art music generation model MusicLM in PyTorch using attention network.

  42. hustvl/Vim - Vision Mamba (proposed at the 2024 International Machine Learning Conference) realizes efficient visual representation learning through bidirectional state space models.

  43. OpenGVLab/Ask-Anything - CVPR2024 Highlights: VideoChatGPT enables ChatGPT to understand videos. It also supports other language models such as miniGPT4, StableLM and MOSS.

  44. microsoft/lida - Automatically generate visual content and infographics through large language models.

  45. google-research/frame-interpolation - Large motion frame interpolation (FILM) in ECCV 2022

  46. InternLM/InternLM-XComposer - InternLM - XComposer2.5 - OmniLive: A multimodal system for long-term video and audio interaction.

  47. yerfor/GeneFace - GeneFace: Broadly and high-fidelity 3D speaking face synthesis, ICLR 2023, including official code.

  48. OpenGVLab/InternImage - InternImage in CVPR 2023: Exploring large-scale visual basic models using deformable convolution.

  49. google-deepmind/gemma - Google DeepMind's open weight large language model.

  50. baaivision/EVA - EVA series: Fantasy of visual representation from Beijing Zhiyuan Artificial Intelligence Research Institute (BAAI).

  51. MzeroMiko/VMamba - VMamba: Visual state space model. Its code is based on Mamba.

  52. deepseek-ai/DeepSeek-VL - DeepSeek - VL aims to understand vision-language in the real world.

  53. openai/consistencydecoder - Consistent distillation differential variational autoencoder.

  54. gligen/GLIGEN - Open - Text-to-image generation based on grounding.

  55. dvlab-research/LISA - "LISA: Inference Segmentation Based on Large Language Models" project page.

  56. 3DTopia/LGM - LGM in ECCV 2024 Oral Report is used for high resolution 3D content creation.

  57. lyuchenyang/Macaw-LLM - Macaw - Large Language Model (LLM) integrates images, video, audio and text for multimodal language modeling.

  58. OpenMotionLab/MotionGPT - MotionGPT is a unified motion-language generation model using large language models (LLMs) at the 2023 Neural Information Processing Systems Conference (NeurIPS 2023).

  59. OpenGVLab/InternVideo - Video basic models and data for multimodal understanding in ECCV2024 (European International Conference on Computer Vision 2024).

  60. openai/Video-Pre-Training - Video Pre-training (VPT) involves learning actions by observing unlabeled online videos.

  61. THUDM/ImageReward - ImageReward in NeurIPS 2023: Learning and evaluating human preferences for literary and genomic images.

  62. evo-design/evo - Basic biological modeling from molecular to genome scale.

  63. google-research/tapas - An end-to-end neural model for understanding tables and texts.

  64. apple/ml-aim - This repository provides code and model checkpoints for AIMv1 and AIMv2 research projects.

  65. showlab/Show-o - Show-o code base, a single Transformer for unified multimodal understanding and generation.

  66. ELLA-Diffusion/ELLA - Equip diffusion models with large language models (LLMs) to enhance semantic alignment.

  67. declare-lab/tango - A family of diffusion models for literary and sound.

  68. OpenBMB/VisCPM - Chinese-English dual-mode model series based on CPM basic model (chat and painting)

  69. OpenGVLab/VisionLLM - Visual Large Language Model (VisionLLM) series.

  70. BAAI-DCAI/Bunny - Lightweight multimodal model family.

  71. Ligo-Biosciences/AlphaFold3 - Open source implementation of AlphaFold3.

  72. Vchitect/SEINE - SEINE: A short-to-long video diffusion model for generative transformation and prediction in the 2024 International Conference on Learning Representation (ICLR).

  73. google-deepmind/materials_discovery -

  74. OpenGVLab/SAM-Med2D - SAM - Official implementation of Med2D.

  75. OpenMOSS/AnyGPT - Code for "AnyGPT: Unified Multimodal Large Language Model (LLM) with Discrete Sequence Modeling".

Multilingual Model

  1. THUDM/ChatGLM-6B - ChatGLM-6B is an open bilingual dialogue language model.

  2. ymcui/Chinese-LLaMA-Alpaca - China's LLaMA (Alpaca) and Alpaca (Alpaca) Large Language Model + Local CPU (Central Processor)/GPU (Graphics Processor) Training and Deployment

  3. UKPLab/sentence-transformers - state-of-the-art text embedding technology.

  4. FlagAlpha/Llama2-Chinese - The Chinese community of Alpaca (Llama) has opened Llama3 for online experience and fine-tuning. It updates all code for Llama3, is completely open source and is available for commercial use, and also writes the latest Llama3 learning materials.

  5. THUDM/ChatGLM3 - ChatGLM3 series: large language model for open source bilingual chat.

  6. ymcui/Chinese-LLaMA-Alpaca-2 - Chinese version of LLaMA-2 and Alpaca-2 large-scale model phase II project and 64K long text model.

  7. InternLM/InternLM - The InternLM2.5 basic model and chat model are officially released, supporting a 1 million (1M) context.

  8. Facico/Chinese-Vicuna - China - Vicuna: A Chinese model based on LLaMA following instructions - a low-resource Chinese llama + lora solution that references alpaca structure.

  9. LC1332/Luotuo-Chinese-LLM - Luotuo is an open source Chinese language model developed by Chen Qiyuan of Central China Normal University, Li Lulu and Leng Zi'ang of SenseTime.

  10. wenge-research/YAYI2 - YAYI 2 is a new generation of open source large language model developed by Zhongke Wenge, which uses more than 2 trillion high-quality, multilingual corpus marks for pre-training.

  11. wenge-research/YaYi - Yayi Large Model is developed by the Zhongke Wenge Algorithm Team. It is a safe and reliable exclusive large model created for customers. It is trained based on large-scale Chinese and English multi-domain instruction data from the LlaMA2 and BLOOM series.

  12. TigerResearch/TigerBot - TigerBot is a large language model that supports multiple languages ​​and tasks.

  13. LinkSoul-AI/Chinese-Llama-2-7b - The first downloadable and operational Chinese LLaMA2 model in the open source community!

  14. MiuLab/Taiwan-LLM - Traditional Mandarin language model for Taiwan.

  15. zjunlp/KnowLM - An open source large language model framework with knowledge.

  16. google-research/multilingual-t5 -

  17. SkyworkAI/Skywork - The Skywork series model is pre-trained on 3.2TB of multilingual and code data, and related projects are open source.

Data organization

  1. photoprism/photoprism - Photo application powered by artificial intelligence on decentralized networks.

  2. freedmand/semantra - A multi-functional tool for semantic search.

  3. neo4j/NaLLM - Repository of NaLLM projects.

AI Services

  1. vllm-project/vllm - A high-throughput and memory-efficient engine for inference and services of large language models (LLMs).

  2. guillaumekln/faster-whisper - Faster-faster-whisper - Faster-faster Whisper transcription using CTranslate2.

  3. bentoml/OpenLLM - Run open source large language models (LLMs) like Llama and Mistral as OpenAI-compatible API endpoints in the cloud.

  4. huggingface/text-generation-inference - Large-scale language model text generation reasoning.

  5. FMInference/FlexGen - Run large language models on a single GPU in a throughput-focused scenario.

  6. triton-inference-server/server - Triton inference server provides optimized inference solutions for both the cloud and the edge.

  7. dusty-nv/jetson-inference - Guide to Deploying Deep Learning Inference Networks and Deep Vision Primitives with TensorRT and NVIDIA Jetson in Hello AI World.

  8. openvinotoolkit/openvino - OpenVINO™ is an open source AI inference optimization and deployment toolkit.

  9. zilliztech/GPTCache - Semantic cache for large language models (LLMs), fully integrated with LangChain and llama_index.

  10. Portkey-AI/gateway - An AI gateway that is very fast, has an integrated guardrail and can be routed to many large language models (LLMs) and AI guardrails via one API.

  11. tensorflow/serving - A flexible and high-performance machine learning model service system.

  12. xorbitsai/inference - Replace OpenAI GPT with another large language model (LLM) in your application by modifying a line of code using Xinference, Xinference supports reasoning for running various models anywhere.

  13. allegroai/clearml - ClearML is an MLOps/LLMOps solution for artificial intelligence workloads (including experiments and data management, etc.).

  14. InternLM/lmdeploy - LMDeploy is a toolkit for compressing, deploying, and serving large language models.

  15. argmaxinc/WhisperKit - Apple silicon chip device-side voice recognition.

  16. kserve/kserve - A standardized serverless machine learning inference platform based on Kubernetes.

  17. neuralmagic/deepsparse - CPU-based sparse-aware deep learning inference runtime.

  18. huggingface/text-embeddings-inference - A very fast inference solution for text embedding models.

  19. open-mmlab/mmdeploy - OpenMMLab's model deployment framework.

  20. ModelTC/lightllm - LightLLM is a large language model (LLM) inference and service framework based on Python. It is lightweight, easy to scale and fast.

  21. predibase/lorax - A multi-LoRA inference server that can be scaled to thousands of fine-tuned large language models (LLMs).

  22. langchain-ai/langserve - LangServe (LangServe) ?️?

  23. S-LoRA/S-LoRA - S-LoRA: LoRA adapter serving a large number of concurrent.

  24. michaelfeil/infinity - Infinity is a service engine for text embedding, reordering models, clip, clap and colpali, with high throughput and low latency.

  25. roboflow/inference - A command hub for transforming any computer or edge device into a computer vision project.

  26. ray-project/ray-llm - RayLLM - Large language model based on Ray.

  27. PygmalionAI/aphrodite-engine - Large-scale Language Model (LLM) inference engine.

  28. punica-ai/punica - serves multiple LoRA fine-tuned large language models as one.

  29. msoedov/langcorn - Automatically serves LangChain large language model (LLM) applications and agents for LLMops using FastApi.

  30. mosecorg/mosec - A high-performance machine learning model service framework with dynamic batch processing and CPU/GPU pipelines to maximize computer utilization.

Vector database

  1. facebookresearch/faiss - A library for efficient dense vector similarity search and clustering.

  2. milvus-io/milvus - Milvus is a high-performance, cloud-native vector database for scalable vector approximate nearest neighbor search.

  3. qdrant/qdrant - Qdrant is a high-performance, large-scale vector database and search engine for the next generation of artificial intelligence, and also provides cloud services.

  4. chroma-core/chroma - Open source embedded database native to artificial intelligence.

  5. spotify/annoy - C++/Python approximate nearest neighbor algorithm optimized for memory usage and disk I/O.

  6. weaviate/weaviate - Weaviate is an open source vector database. It stores objects and vectors, and can search vectors under structured filtering, which is fault-tolerant and scalable.

  7. neuml/txtai - An all-in-one open source embedded database for semantic search, large language model (LLM) orchestration and language model workflows.

  8. activeloopai/deeplake - A database for use by artificial intelligence that can store various types of data such as vectors, images, text and videos. It can be used with large language models/language chains (LLMs/LangChain) for operations such as storing, querying, versioning and visualizing AI data, and can transmit data in real time to PyTorch/TensorFlow.

  9. vespa-engine/vespa - Artificial Intelligence + data, available online at https://vespa.ai.

  10. lancedb/lancedb - A serverless vector database for artificial intelligence applications, developer-friendly and easily add long-term memory to large language model (LLM) applications.

  11. marqo-ai/marqo - A unified embedding generation and search engine, which can also be used in the cloud - cloud.marqo.ai.

  12. nmslib/hnswlib - A C++/python library with only header files for fast approximation of nearest neighbor searches.

  13. unum-cloud/usearch - A fast open source search and clustering engine for multiple programming languages.

  14. tensorchord/pgvecto.rs - Scalable, low-latency vector search in Postgres, a database management system, with the help of hybrid capabilities. It revolutionizes vector search rather than databases.

  15. spotify/voyager - A Python and Java library for approximate nearest neighbor searches, focusing on ease of use, simplicity, and deployability.

  16. rapidsai/raft - RAFT has basic algorithms and primitives for machine learning and information retrieval (IR) that are accelerated by CUDA for high-performance applications.

AI Education

  1. JushBJJ/Mr.-Ranedeer-AI-Tutor - A customizable GPT for personalized learning experiences - 4 artificial intelligence tutor tip words.

  2. Nutlope/llamatutor - An AI private tutor is built on Llama 3.1.

  3. codeacme17/examor - Large language models (LLMs) help learning for students, scholars, respondents and lifelong learners.

AI development tools

  1. jina-ai/jina - Use cloud native technology stack to build multimodal artificial intelligence applications.

  2. iterative/dvc - Data versioning and machine learning experiments.

  3. unifyai/ivy - Convert machine learning code between different frameworks.

  4. HigherOrderCO/HVM - Large-scale parallel optimal function runtime written in Rust.

  5. marimo-team/marimo - Python reactive notebooks can be used for reproducibility experiments, script execution, application deployment, and version control with Git.

  6. arogozhnikov/einops - Flexible and powerful tensor operations for creating code that has both readability and reliability, suitable for PyTorch, Jax, TensorFlow, etc.

  7. replicate/cog - Container used in machine learning.

  8. jessevig/bertviz - BertViz: Visualize attention in natural language processing models such as BERT, GPT2, BART, etc.

  9. AbdBarho/stable-diffusion-webuui-docker - Easily set up Docker for Stable Diffusion through a user-friendly interface.

  10. huggingface/safetensors - An easy and safe way to store and allocate tensors.

  11. wangzhaode/mnn-llm - Deploy a large language model (LLM) project based on MNN.

  12. ajndkr/lanarky - A network framework for building large language model (LLM) microservices.

Model training

  1. tensorflow/tensorflow - an open source machine learning framework that everyone can use.

  2. huggingface/transformers - Transformer: state-of-the-art machine learning technology for Pytorch, TensorFlow and JAX.

  3. pytorch/pytorch - Use powerful GPU acceleration capabilities, Python-based tensors and dynamic neural networks.

  4. hpcaitech/ColossalAI - Reduce the cost, increase its speed and enhance its accessibility of large AI models.

  5. hiyouga/LLaMA-Factory - Unified and efficient fine-tuning of more than 100 large language models (LLMs) (ACL 2024)

  6. lm-sys/FastChat - An open platform for training, service and evaluation of large language models and a publishing library for Vicuna and Chatbot Arena.

  7. coqui-ai/TTS - ? is a deep learning toolkit for text to speech that has been validated in research and production.

  8. microsoft/DeepSpeed ​​- DeepSpeed ​​is a deep learning library for easy, efficient and efficient distributed training and inference.

  9. ray-project/ray - Ray is an AI computing engine with core distributed runtime and AI libraries for accelerating machine learning workloads.

  10. google-research/google-research - Google Research

  11. google/jax - Python+NumPy programs can be combined and transformed in a variety of ways, such as derivative, vectorization, and real-time compilation (JIT) to GPU/TPU, etc.

  12. open-mmlab/mmdetection - OpenMMLab detection toolkit and benchmarking.

  13. tinygrad/tinygrad - If you like PyTorch and Micrograd, then you will also like Tinygrad.

  14. huggingface/diffusers - Diffusers: The most advanced diffusion model for PyTorch and FLAX for images, videos and audio.

  15. mozilla/DeepSpeech - DeepSpeech is an open source voice-to-text engine that can be used in real time on a variety of devices.

  16. modularml/mojo - Moho programming language (more details are not provided in the given description)

  17. microsoft/unilm - Self-supervised pre-training covering tasks, languages ​​and modalities at scale.

  18. ml-explore/mlx - MLX is an array framework for Apple chips.

  19. HigherOrderCO/Bend - A high-level programming language with large-scale parallelism.

  20. huggingface/peft - PEFT: Optimal high-efficiency fine-tuning of parameters.

  21. huggingface/candle - A minimalist machine learning framework for the Rust language.

  22. NVIDIA/NeMo - A framework for researchers and developers in the fields of AI, such as large language models (LLMs), multimodal and voice AI, is scalable and generative.

  23. PaddlePaddle/PaddleNLP - An easy-to-use and powerful natural language processing (NLP) and large language model (LLM) library with a large number of excellent models that support a wide range of natural language processing tasks from research to industrial applications.

  24. PaddlePaddle/PaddleSpeech - An easy-to-use voice toolkit contains multiple features and won the NAACL2022 Best Demo Award.

  25. Lightning-AI/litgpt - More than 20 high-performance large language models (LLMs) and related solutions for large-scale pre-training, fine-tuning and deployment.

  26. huggingface/trl - Use reinforcement learning to train Transformer language models.

  27. artidoro/qlora - QLoRA can achieve efficient fine-tuning of quantified large language models.

  28. salesforce/LAVIS - LAVIS is a one-stop language - visual smart library.

  29. nerfstudio-project/nerfstudio - A collaborative and friendly studio for neural radiation fields (NeRFs).

  30. mozilla/TTS - Deep learning for speech synthesis (discussion forum: https://discourse.mozilla.org/c/tts)

  31. trace-ai/burn - Burn is a new dynamic deep learning framework built by Rust to achieve flexibility, efficiency and portability.

  32. facebookresearch/pytorch3d - PyTorch3D is a library for FAIR (Facebook Artificial Intelligence Research) for deep learning of 3D data.

  33. facebookresearch/xformers - Transformer build modules are flexible and optimized to support combined builds.

  34. OptimalScale/LMFlow - An expandable toolkit for fine-tuning and inference of large basic models, allowing large models to be used by everyone.

  35. OpenAccess-AI-Collective/axolotl - Just ask the newt questions.

  36. FlagOpen/FlagEmbedding - Search and search for enhanced large language models (LLMs)

  37. huggingface/accelerate - A simple way to handle PyTorch models on a variety of devices and configurations, with features like automatic mixing accuracy and support for fully sharded data parallelism (FSDP) and DeepSpeed.

  38. LianjiaTech/BELLE - BELLE is an open source large-scale Chinese dialogue language model engine for everyone.

  39. cloneofsimo/lora - Quickly fine-tune diffusion model through low rank adaptation.

  40. EleutherAI/gpt-neox - Implement an autoregressive transformer with model parallelism on GPU using Megatron and DeepSpeed ​​libraries.

  41. open-mmlab/mmagic - OpenMMLab is a multimodal toolbox that can be used for various tasks such as artificial intelligence generated content (AIGC), with easy-to-use application program interfaces (APIs) and model libraries.

  42. facebookresearch/metaseq - External large-scale work database.

  43. Maartengr/BERTopic - Generate easy-to-understand topics using BERT and c-TF-IDF.

  44. Project-MONAI/MONAI - A toolkit for medical imaging (processing) in the field of artificial intelligence.

  45. yangjianxin1/Firefly - Firefly is a large model training tool that supports training multiple large models, such as Qwen2.5, Qwen2, etc.

  46. google-deepmind/graphcast -

  47. mosaicml/composer - Enhance your model training.

  48. cg123/mergekit - A tool for combining pre-trained large language models.

  49. CarperAI/trlx - A code base for distributed training of language models using human feedback reinforcement learning (RLHF).

  50. pytorch/torchtune - A PyTorch native post-training library.

  51. google-deepmind/open_spiel - OpenSpiel is a set of environments and algorithms for general reinforcement learning and game search/planning research.

  52. huggingface/autotrain-advanced - Automatic training advanced version.

  53. InternLM/xtuner - An efficient, flexible and fully functional toolkit for fine-tuning a wide range of large language models (LLMs).

  54. mosaicml/llm-foundry - Large Language Model (LLM) training code for Databricks basic models.

  55. baidu-research/warp-ctc - Fast parallel connectionist time classification (CTC).

  56. JohnSnowLabs/spark-nlp - The most advanced technology in natural language processing.

  57. FlagAI-Open/FlagAI - FlagAI is a toolkit for large-scale models that is fast, easy to use and scalable.

  58. mlfoundations/open_flamingo - An open source framework for training large multimodal models.

  59. OpenLLMAI/OpenRLHF - an easy-to-use, scalable and high-performance human feedback reinforcement learning (RLHF) framework, with nearly 70 billion parameters, full fine-tuning of the proximal strategy optimization (PPO) and iterative direct preference optimization (DPO), low-rank auto- Functions such as LoRA, RingAttention, and recursive fine-tuning (RFT).

  60. google-deepmind/acme - A library of components and proxy for reinforcement learning.

  61. open-mmlab/mmpretrain - Pre-training toolbox and benchmarks for OpenMMLab.

  62. shibing624/MedicalGPT - MedicalGPT uses ChatGPT training pipeline to train medical GPT models to achieve incremental pre-training, supervised fine-tuning, human feedback reinforcement learning (RLHF), direct preference optimization (DPO), and sort-based preference optimization (ORPO).

  63. iryna-kondr/scikit-llm - Integrate large language models (LLMs) into scikit-learning without any problem.

  64. google-research/scenic - Scenic: A Jax library for computer vision research and more.

  65. facebookresearch/fairscale - PyTorch extension for high performance and large-scale training.

  66. alpa-projects/alpa - Training and serving large-scale neural networks through automatic parallelization.

  67. microsoft/torchscale - The infrastructure for large and medium-sized language models.

  68. google-deepmind/dm-haiku - A neural network library based on JAX.

  69. eureka-research/Eureka - The official repository of the paper "Eureka: Reward Design at the Human Level by Coding Large Language Models" published at the ICLR 2024 conference.

  70. Alpha-VLLM/LLaMA2-Accessory - An open source toolkit for developing large language models.

  71. google-research/t5x -

  72. google-deepmind/alphatensor -

  73. PhoebusSi/Alpaca-CoT -

  74. huggingface/optimum - Use easy-to-use hardware optimization tools to accelerate reasoning and training of Transformer, Diffuser, TIMM and Sentence Transformer.

  75. stochasticai/xTuring - Easily build, customize and control your own large language models (LLMs) from data preprocessing to fine-tuning and join its Discord community.

  76. adapter-hub/adapters - A unified library for parameter efficient and modular transfer learning.

  77. openai/weak-to-strong -

  78. OpenPipe/OpenPipe - Convert expensive tips to affordable fine-tuning models.

  79. lamini-ai/lamini - The official Python client for the Lamini API.

  80. google-research/big_vision - Official code base for developing vision transformers, SigLIP, multi-layer perceptron mixers (MLP-Mixer), LiT, etc.

  81. young-geng/EasyLM - EasyLM (based on JAX/Flax) provides a one-stop solution for pre-training, fine-tuning, evaluation and service operations of various large language model (LLM) operations.

  82. pyro-ppl/numpyro - Probability programming is used with NumPy, and automatic derivatives are used with JAX and instant compilation (JIT compilation) to GPU/TPU/CPU.

  83. eric-mitchell/direct-preference-optimization - a reference implementation of direct preference optimization (DPO).

  84. huggingface/setfit - Use Sentence Transformers for efficient small sample learning.

  85. allenai/open-instruct -

  86. allenai/RL4LMs - A modular reinforcement learning (RL) library for fine-tuning language models according to human preferences.

  87. lxe/simple-llm-finenetuner - A simple user interface for fine-tuning large language models (LLMs).

  88. THUDM/P-tuning-v2 - An optimized depth prompt fine-tuning strategy works as well as fine-tuning in different sizes and tasks.

  89. tensorflow/privacy - A library for training machine learning models and privacy-protecting training data.

  90. xlang-ai/instructor-embedding - Single embedder for any task: Instructions - Fine-tune text embedding (ACL 2023)

  91. unslothai/hyperlearn - Machine learning algorithms are 2-2000 times faster, memory usage is reduced by 50%, and can run on all hardware.

  92. salesforce/ctrl - A conditional conversion language model for controllable generation.

  93. google-deepmind/optax - Optax is a JAX library for gradient processing and optimization.

  94. google-deepmind/penzai - A JAX toolkit for building, modifying and visualizing neural networks.

  95. microsoft/i-Code -

  96. kubeflow/training-operator - Distributed machine learning training and fine-tuning on Kubernetes.

  97. AetherCortex/Llama-X - Carry out open academic research to enable LLaMA to reach the most advanced large language model.

  98. salesforce/ALBEF - A new vision-code for language pre-training method ALBEF.

  99. kubeflow/katib - Automated machine learning in Kubernetes environment.

  100. facebookresearch/multimodal - TorchMultimodal is a PyTorch library for large-scale training of state-of-the-art multimodal multitasking models.

  101. jina-ai/finenetuner - Task-oriented embedding adjustment for BERT, CLIP, etc.

  102. salesforce/CodeTF - CodeTF: One-stop Transformer library for the latest code large language models (LLM).

  103. AnswerDotAI/fsdp_qlora - Train large language model (LLM) using QLoRA and full sharded data parallel (FSDP).

  104. nerfstudio-project/nerfacc - A universal neural radiation field (NeRF) acceleration toolbox based on PyTorch.

  105. jquesnelle/yarn - YaRN: Efficient context window extension for large language models.

  106. PKU-Alignment/safe-rlhf - Safe human feedback reinforcement learning (Safe RLHF) utilizes secure reinforcement learning based on human feedback to achieve limited value alignment.

  107. lucidrains/self-rewarding-lm-pytorch - Implementing the training framework proposed by MetaAI in the "Self-rewarding Language Model".

  108. OpenLMLab/MOSS-RLHF - This is the secret about human feedback reinforcement learning (RLHF) in large language models, especially the first part of the near-end strategy optimization algorithm (PPO).

  109. JonasGeiping/cramming - Training of compressed BERT type language models within a limited computing resource.

  110. AlibabaResearch/DAMO-ConvAI - Damo - Conversational Artificial Intelligence (ConvAI) is an official resource library with the code base of DAMO Conversational AI.

  111. databricks/megablocks -

  112. AGI-Edgerunners/LLM-Adapters - Code of EMNLP 2023 paper "LLM - Adapters: Adapter Family for Efficient Fine-Tuning of Parameters in Large Language Models".

  113. KhoomeiK/LlamaGym - Fine-tune large language model (LLM) agents through online reinforcement learning.

  114. thunlp/OpenDelta - A plug and play parameter efficient adjustment (Delta Tuning) library.

  115. Liuhong99/Sophia - Official implementation of the paper "Sophia: A scalable random second-order optimizer for pre-training of language models".

  116. yuchenlin/LLM-Blender - LLM-Blender is an integration framework in [ACL2023] that eliminates weaknesses through sorting and enhances the capabilities of large language models (LLMs) by generating fusion advantages.

  117. google-deepmind/xmanager - A platform for handling machine learning experiments.

  118. google-deepmind/chex -

AI image generation

  1. AUTOMATIC1111/stable-diffusion-webui - Stable diffusion Web user interface.

  2. lllyasviel/Fooocus - The description is about focusing on prompts and generation.

  3. upscayl/upscayl - Upscayl is the number one free open source artificial intelligence image amplifier for Linux, MacOS and Windows systems.

  4. s0md3v/roop - One-click face change means you can complete face change with just one click.

  5. invoke-ai/InvokeAI - Invoke is a creative engine for the Stable Diffusion model that provides a WebUI (user interface) and serves as the basis for commercial products.

  6. facefusion/facefusion - An industry-leading facial processing platform.

  7. Sanster/lama-cleaner - Image repair tools powered by state-of-the-art artificial intelligence models can remove unwanted elements in pictures or replace things in them.

  8. Mikubill/sd-webui-controlnet - WebUI extension for ControlNet.

  9. camera/stable-diffusion-webui-colab - Stable diffusion web user interface on Colab (Google Collaboration Platform).

  10. divamgupta/diffusionbee-stable-diffusion-ui - Diffusion Bee is the easiest way to run Stable Diffusion locally on an M1 Mac. It has a one-click installer that requires no dependencies or technical knowledge.

  11. Baiyuetribe/paper2gui - Convert AI papers to graphical user interface (GUI) so that everyone can easily use AI technology.

  12. easydiffusion/easydiffusion - The easiest way to create exquisite art works using artificial intelligence on the PC side is to operate it in one click. Just enter the text prompt word to generate images through the browser user interface without technical knowledge.

  13. Stability-AI/StableStudio - Community interface for generative artificial intelligence.

  14. carson-katri/dream-textures - Stable Diffusion integrated into Blender.

  15. TheLastBen/fast-stable-diffusion - Fast-stable-diffusion and DreamBooth.

  16. godly-devotion/MochiDiffusion - Run Stable Diffusion locally on your Mac.

  17. HumanAIGC/OutfitAnyone - Ultra-high quality virtual try-on for everyone and clothing.

  18. sensing-ai/dot - Deep forgery attack toolkit.

  19. leap-ai/headshots-starter -

  20. Nutlope/restorePhotos - Use artificial intelligence to fix blurred old face photos.

  21. jina-ai/discoart - A group creates disco diffusion artwork.

  22. mlc-ai/web-stable-diffusion - introduces a stable diffusion model into a web browser; everything runs in the browser without server support.

  23. all-in-aigc/aicover - A tool for generating AI covers.

Dataset

  1. huggingface/datasets - The largest machine learning model dataset center has fast, easy to use and efficient data manipulation tools.

  2. BuilderIO/gpt-crawler - Crawl website content from a URL to generate the knowledge file needed to create a custom GPT.

  3. joke2k/faker - Faker is a Python package for generating fake data.

  4. DS4SD/docling - Get your documentation ready for generative artificial intelligence.

  5. openai/tiktoken - Tiktoken is a fast BPE word segmenter suitable for OpenAI models.

  6. cleanlab/cleanlab - Standard AI packages focus on data for quality and machine learning, handling cluttered real-world data and labels.

  7. karpathy/minbpe - A concise code for byte-pair encoding (BPE) algorithms commonly used in large language model (LLM) tokenization.

  8. huggingface/tokenizers - Fast, advanced word segmentation optimized for research and production.

  9. arsenetar/dupeguru - Find duplicate files.

  10. QuivrHQ/MegaParse - Optimized file parser for lossless large language model (LLM) ingestion, able to parse PDF, Docx, PPTx files into an ideal format suitable for large language model (LLM).

  11. Togethercomputer/RedPajama-Data - RedPajama - There is code in the data repository for preparing large datasets for training large language models.

  12. lk-geimfari/mimesis - Mimesis is a Python data generator that can create various types of fake data in multiple languages.

  13. Instruction-Tuning-with-GPT-4/GPT-4-LLM - Use GPT-4 to fine-tune the instruction.

  14. yizhongw/self-instruct - Aligns the pretrained language model with its own generated instruction data.

  15. dedupeio/dedupe - A Python library for precise and scalable fuzzy matching, record deduplication and entity parsing.

  16. argilla-io/argilla - Argilla is a tool that allows artificial intelligence engineers and field experts to create high-quality data sets.

  17. mshumer/gpt-llm-trainer -

  18. life4/textdistance - Use over 30 pure Python algorithms, common interfaces, and optionally use external libraries to calculate distances between sequences.

  19. Docta-ai/docta - A doctor is here to take care of your data.

  20. alibaba/data-juicer - Provides high-quality, rich and easy-to-process data for the basic model.

  21. towhee-io/towhee - Towhee is a framework that makes neural data processing pipelines simple and fast.

  22. QData/TextAttack - TextAttack is a Python framework for natural language processing (NLP) tasks such as adversarial attacks, data augmentation, and model training.

  23. seatgeek/thefuzz - Fuzzy string matching in Python.

  24. ekzhu/datasketch - Minimum hash, locally sensitive hash, locally sensitive hash forest, weighted minimum hash, hyperlog, hyperlog++, locally sensitive hash integration and hierarchical navigational small world map.

  25. thunlp/UltraChat - Large, informative and diverse multi-round chat data and models.

  26. modAL-python/modAL - A modular active learning framework written in Python.

  27. chiphuyen/lazynlp - A library for crawling and cleaning web pages to generate large data sets.

  28. huggingface/datatrove - Provides a set of platform-independent customizable pipeline processing modules, so that data processing no longer depends on scripting.

  29. refuel-ai/autolabel - Use large language models to mark, clean and augment text datasets.

  30. google-deepmind/code_contests -

  31. Tencent/MedicalNet - Many studies have shown that the amount of training data can significantly affect the performance of deep learning. The MedicalNet project provides 3D-ResNet pre-trained models and code.

  32. argilla-io/distilabel - Distilabel is a synthetic data and artificial intelligence feedback framework for engineers who need to build fast, reliable and scalable pipelines based on proven research papers.

  33. google-deepmind/mathematics_dataset - This dataset code creates mathematical question and answer pairs at school difficulty levels from various question types.

  34. openai/prm800k - 800,000 step-level correctness labels for answers to large language models (LLMs) to solve mathematical problems.

  35. salesforce/WikiSQL - A large corpus for the development of natural language interfaces with semantic analysis annotations.

  36. anthropics/hh-rlhf - Human preference data for training assistants (making it useful and harmless) reinforcement learning through human feedback.

  37. moj-analytical-services/splink - Fast, scalable and accurate probabilistic data links, supporting multiple SQL backends.

  38. dleemiller/WordLlama - Tag embedding of large language models (LLMs) can accomplish things.

  39. AI4Finance-Foundation/FinRL-Meta - FinRL - Meta provides dynamic data sets and market environment for FinRL.

  40. tensorflow/text - Use text as the primary element in TensorFlow.

  41. google-research/deduplicate-text-datasets -

  42. allenai/dolma - Data and tools for creating and inspecting OLMo pretrained data.

  43. lilacai/lilac - Improve data management of large language models.

  44. 1e0ng/simhash - A Python implementation of Simhash algorithm.

  45. J535D165/recordlinkage - A Python toolkit for recording links and repeated data detection, powerful and modular.

  46. google-deepmind/tree - Tree is a library for processing nested data structures.

  47. xtreme1-io/xtreme1 - Xtreme1 is an integrated multimodal data training platform that supports 3D lidar point cloud, image and large language models (LLM) for data labeling and annotation.

  48. datadreamer-dev/DataDreamer - DataDreamer: Prompt, generate synthetic data, train and calibrate models.

  49. HazyResearch/meerkat - Creative and interactive views of all datasets.

Model evaluation

  1. openai/evals - Evals is a framework for evaluating large language models (LLMs) and their systems, and is also an open source benchmarking registry.

  2. exploitinggradients/ragas - Inject powerful momentum into your evaluation of large language model (LLM) applications.

  3. EleutherAI/lm-evaluation-harness - A framework for evaluating language models using a small number of samples.

  4. erikbern/ann-benchmarks - Benchmarks for approximate nearest neighbor libraries in Python.

  5. Trusted-AI/adversarial-robustness-toolbox - Adversarial Robustness Toolkit (ART) is a Python library for machine learning security, covering evasion, poisoning, extraction and reasoning for both red and blue teams (both offense and defense). (attack).

  6. open-compass/opencompass - OpenCompass is a large language model (LLM) evaluation platform that supports more than 100 data sets and a variety of models such as Llama3 and Mistral.

  7. Arize-ai/phoenix - Observability and evaluation of artificial intelligence.

  8. NVIDIA/NeMo-Guardrails - NeMo Guardrails is an open source toolkit that easily adds programmable guardrails to conversational systems based on large language models (LLM).

  9. confident-ai/deepeval - Large Language Model (LLM) evaluation framework.

  10. Giskard-AI/giskard - Open source evaluation and testing of artificial intelligence and large language model (LLM) systems.

  11. fchollet/ARC - The abstract reasoning corpus is a resource, but it is difficult to give a more specific description without more details. It may be used for tasks related to abstraction and reasoning in various fields.

  12. llm-attacks/llm-attacks - A general migration attack that aligns language models.

  13. leondz/garak - Large Language Model (LLM) Vulnerability Scanner.

  14. jeinlee1991/chinese-llm-benchmark - The list of Chinese big model capability assessments contains 134 models, including commercial and open source models, and provides capability scores and original output.

  15. google/BIG-bench - Beyond imitation game: a collaborative benchmark for measuring and inferring language model capabilities.

  16. meta-llama/PurpleLlama - A set of tools for evaluating and enhancing the security of large language models (LLMs).

  17. openai/human-eval - The paper "Evaluation of Code-Based Large Language Models" Code

  18. salesforce/decaNLP - Natural Language Decathlon (Contest) is a multitasking challenge in the field of natural language processing.

  19. THUDM/AgentBench - A comprehensive benchmark for evaluating large language models (LLMs) as agents (International Conference on Learning Representation ICLR'24).

  20. truera/trulens - Evaluation and tracking in large language model (LLM) experiments.

  21. princeton-nlp/SWE-bench - [International Conference on Learning Representation (ICLR) 2024] SWE - bench: Can language models solve the problem of GitHub (code hosting platform) in the real world.

  22. Lightning-AI/torchmetrics - Machine learning metrics for distributed and scalable PyTorch applications.

  23. openai/simple-evals -

  24. huggingface/evaluate - A library called Evaluate for easy evaluation of machine learning models and datasets.

  25. embeddings-benchmark/mteb - MTEB is a large-scale text embedding benchmark.

  26. Azure/PyRIT - PyRIT is an open source framework that helps security professionals and engineers proactively identify risks in generative artificial intelligence systems.

  27. stanford-crfm/helm - HELM is a framework for improving the transparency of language models, which is also used to evaluate other models such as text-to-image models in HEIM and vision-language models in VHELM.

  28. TransformerLensOrg/TransformerLens - A library for mechanical interpretability of GPT-like language models.

  29. beir-cellar/beir - A heterogeneous information retrieval (IR) benchmark for easily evaluating models with more than 15 different datasets.

  30. tatsu-lab/alpaca_eval - An automatic evaluator for directive-compliant language models. It has been manually proven, high quality, low cost and fast.

  31. microsoft/CodeXGLUE - CodeXGLUE is a project or entity, but without more context information, it is difficult to say more specific.

  32. google-deepmind/bsuite - Bsuite is a carefully designed set of experiments to explore the core capabilities of reinforcement learning agents.

  33. CalculatedContent/WeightWatcher - The WeightWatcher tool is used to predict the accuracy of deep neural networks.

  34. facebookresearch/LAMA - Language model analysis.

  35. evalplus/evalplus - Strict evaluation of code synthesized by large language model (LLM) in NeurIPS 2023 and COLM 2024.

  36. vectara/hallucination-leaderboard - A ranking for comparing large language models (LLMs) to produce hallucination performance when summarizing short essays.

  37. hendrycks/test - Measure large-scale multitasking language comprehension at the 2021 International Conference on Learning Representation (ICLR).

  38. mlcommons/inference - Reference implementation of MLPerf™ inference benchmarks.

  39. openai/grade-school-math -

  40. rlancemartin/auto-evaluator - An evaluation tool for large language model question and answer chains.

  41. openai/automated-interpretability -

  42. allenai/natural-instructions - This description is about extending natural instructions.

  43. WeOpenML/PandaLM -

  44. thu-coai/Safety-Prompts - Chinese security tips for evaluating and improving the security of large language models.

  45. salesforce/OmniXAI - OmniXAI is an interpretable artificial intelligence (XAI) library.

  46. bigcode-project/bigcode-evaluation-harness - A framework for evaluating autoregressive code generation language models.

  47. hsiehjackson/RULER - This repository contains the source code for RULER (the real context size regarding the long context language model).

Computational Management

  1. kubeflow/kubeflow - A machine learning toolkit designed specifically for Kubernetes.

  2. Netflix/metaflow - an open source platform for artificial intelligence and machine learning.

  3. skypilot-org/skypilot - SkyPilot is able to run AI and batch jobs on any infrastructure (Kubernetes or more than 12 clouds), providing unified execution, cost savings, and high GPU availability through a simple interface.

  4. gpuweb/gpuweb - This is where the GPU (Graphics Processor) works in the network.

  5. zenml-io/zenml - ZenML: The connection between machine learning and operations.

  6. higgsfield-ai/higgsfield - Fault-tolerant and highly scalable GPU orchestration, and a machine learning framework for training large-scale models.

  7. Haidra-Org/AI-Horde - A crowdsourcing distributed cluster for generating artificial intelligence art and text.

AI Writing

  1. steven-tey/novel - A Notion-like editor with an AI-powered autocompletion feature.

  2. reorproject/reor - A private and localized artificial intelligence personal knowledge management application for high entropy people.

  3. shibing624/pycorrector - Pycorrector is a text error correction toolkit that has multiple model applications for error correction and is easy to use.

  4. BlinkDL/AI-Writer - Artificial intelligence creates novels such as fantasy and romantic web novels. It is a Chinese pre-trained generative model similar to GPT-2. Using the RWKV model.

  5. mshumer/gpt-author -

  6. Nutlope/twitterbio - Create your Twitter profile using artificial intelligence.

  7. nhaouari/obsidian-textgenerator-plugin - Text Generator is an Obsidian plugin that can be used to generate text with various artificial intelligence vendors such as OpenAI, Anthropic, Google, and local models.

  8. google-deepmind/dramatron - Dramatron uses large language models to generate coherent scripts and scripts.

Intelligent monitoring

  1. nebuly-ai/nebuly - A set of libraries for optimizing the performance of AI models.

  2. langfuse/langfuse - An open source large language model engineering platform with multiple integrations (for large language model observability, metrics, etc.). It comes from YC W23.

  3. evidentlyai/evidently - Evidently is an open source framework for machine learning (ML) and large language model (LLM) observability, which can be used to evaluate, test and monitor artificial intelligence-related systems or data pipelines, with more than 100 items index.

  4. traceloop/openllmetry - Rely on OpenTelemetry to provide open source observability for your large language model (LLM) applications.

  5. Helicone/helicone - An open source large language model (LLM) observability platform. One line of code can be used for monitoring, evaluation and experimentation.

  6. whylabs/whylogs - An open source library for recording data in machine learning models and pipelines. It provides visibility into data quality and model performance, as well as privacy-protected data collection.

  7. uptrain-ai/uptrain - UpTrain is an open source platform for evaluating and improving generative artificial intelligence applications. It provides ratings for pre-configured inspections, analyzes failures and gives solutions.

  8. labmlai/labml - Monitor deep learning model training and hardware usage through mobile phones.

  9. lmnr-ai/lmnr - Laminar is an open source integrated platform for building artificial intelligence products. It creates a data flywheel for AI applications through tracking, evaluation, datasets and labeling (YC S24).

  10. llmonitor/llmonitor - The production toolkit for large language models (LLMs) involves observability, prompt management, and evaluation.

  11. lunary-ai/lunary - The production toolkit for large language models (LLMs) includes observability, prompt management, and evaluation.

  12. dillionverma/llm.report - llm.report is an open source platform for OpenAI for recording API requests, cost analysis, and prompt optimization.

  13. whylabs/langkit - LangKit is an open source toolkit for large language model (LLM) monitoring. It has features such as text quality and sentiment analysis for LLM observability.

Video generation

  1. RayVentura/ShortGPT - ShortGPT - An experimental AI framework for automated operation of YouTube Shorts/TikTok (TikTok International Edition) channels.

  2. all-in-aigc/sorafm - Sora Artificial Intelligence Video Generator provided by Sora.FM.

Data Management

  1. ibis-project/ibis - Portable Python dataframe library.

  2. SuperDuperDB/superduperdb - Superduper is able to build end-to-end AI applications and proxy workflows on existing data infrastructure and preferred tools without data migration.

  3. run-llama/llama-hub - A community-made data loader library for large language models (LLMs) for LlamaIndex and/or LangChain.

  4. webdataset/webdataset - A Python-based input and output (I/O) system suitable for deep learning problems of all sizes, strongly supports PyTorch.

  5. NVIDIA/aistore - AIStore: Scalable AI application storage.

  6. mosaicml/streaming - A database of data flows for efficient neural network training.

AI Applications

  1. code-kern-ai/refinery - An open source option for data scientists to scale, evaluate and maintain natural language data. Treat training data as software artifacts.

AI Engineering

  1. lastmile-ai/aiconfig - AIConfig is a framework for building generative AI applications based on configuration.

How to contribute

We highly welcome your contribution to this warehouse. You can:

  • Submit a new AI project or resource link.

  • Improve the description or classification of existing projects.

  • Report problems or make suggestions for improvement.

  • Participate in the discussion and help other members solve the problem.

Please read our contribution guide for more details.

license

This project is authorized using a MIT license - see the LICENSE file for details.

Expand
Additional Information
Related Applications
Recommended for You
Related Information All