whatisgithub

Popular Python repositories on GitHub, each answered in plain English for non-coders: what it is, its tech stack, and how to run it.

• meta-pytorch/torchtuneA Python library from Meta for fine-tuning large language models using methods like LoRA, QLoRA, and DPO. Acti
• microsoft/promptbaseResearch code from Microsoft for improving GPT-4 accuracy using Medprompt — a method combining dynamic example
• tmelyralab/musetalkAI tool from Tencent Music Entertainment that replaces lip movements in a face video to match new audio, enabl
• andrewyng/translation-agentA Python tool by Andrew Ng that translates text using a three-step LLM loop: initial translation, self-critiqu
• martinblech/xmltodictxmltodict is a tiny Python library that converts XML documents into Python dictionaries and back, making XML d
• klavis-ai/klavisKlavis connects AI agents to 100+ external tools like Gmail, Slack, and GitHub using the Model Context Protoco
• open-edge-platform/anomalibPython library for detecting visual defects and anomalies in images without needing labeled defect examples. B
• 649453932/chinese-text-classification-pytorchA PyTorch toolkit with 7 ready-to-train neural network models for classifying Chinese text into categories lik
• commixproject/commixCommix automates the discovery and exploitation of command injection vulnerabilities in web applications, help
• layout-parser/layout-parserLayoutParser is a Python library that reads scanned documents and PDFs by detecting where paragraphs, tables,
• oarriaga/face_classificationA Python project that detects faces in images or live webcam video and classifies each face by emotion and gen
• i-tu/haskligA monospace coding font based on Adobe Source Code Pro that adds ligatures for multi-character symbols like ->
• redsiege/eyewitnessA Python security tool that automatically screenshots a large list of websites in one pass, records HTTP heade
• vi3k6i5/flashtextFlashText is a Python library that finds and replaces keywords in text much faster than regular expressions wh
• aiogram/aiogramA Python library for building Telegram bots that handles many users at once using async code, with built-in to
• xhongc/music-tag-webA browser-based music tag editor that lets you view and fix song metadata — title, artist, cover art, lyrics —
• frostming/legitA command-line tool that wraps Git with simpler everyday commands, automatically stashing uncommitted work whe
• tvytlx/ai-agent-deep-diveA collection of research PDFs (written in Chinese) that analyze the internal architecture of production AI age
• wepe/machinelearningA personal Chinese-language study collection of classic machine learning and deep learning algorithm implement
• dennybritz/cnn-text-classification-tfA Python implementation of a convolutional neural network that sorts text into categories, based on a 2014 res
• pylint-dev/pylintPylint is a Python code checker that reads your code without running it, finding real errors, style problems,
• karpathy/arxiv-sanity-preserverArxiv Sanity Preserver is a web app by AI researcher Andrej Karpathy that helps you search, filter, and get pe
• python-websockets/websocketswebsockets is a Python library that handles the low-level plumbing for real-time two-way connections between a
• agiresearch/aiosAIOS is a Rutgers University research project that acts as an operating system kernel for AI agents — scheduli
• holoviz/panelPanel is a Python library that turns data analysis scripts into interactive web apps with sliders, charts, and
• bytedance/latentsyncLatentSync is a ByteDance research tool that automatically replaces lip movements in a video to match a new au
• google/gemma_pytorchGoogle's official PyTorch code for running Gemma open-weight AI language models locally, from 1B to 27B parame
• gboeing/osmnxA Python package for downloading and analyzing real-world street networks and geographic data from OpenStreetM
• thudm/slimeA Python framework from Tsinghua University for post-training large language models with reinforcement learnin
• elceef/dnstwistA Python tool that generates lookalike domain names to help security teams detect phishing, typo squatting, an
• edoverflow/can-i-take-over-xyzA community-maintained reference list that tells you whether hundreds of third-party services are vulnerable t
• erikbern/ann-benchmarksA benchmarking suite that measures speed and accuracy of 30+ approximate nearest neighbor search libraries — i
• owid/covid-19-dataArchived dataset of COVID-19 cases, deaths, hospitalizations, testing, and vaccination figures for every count
• baichuan-inc/baichuan-7bBaichuan-7B is an open-source 7-billion-parameter language model trained on Chinese and English text, outperfo
• jpadilla/pyjwtPyJWT is a Python library that creates and verifies JSON Web Tokens — small signed data packets used to prove
• techxuexi/techxuexiTechXueXi is a Python automation tool that earns daily study points on the Chinese government platform Xuexi Q
• nekoparapa/ainieeDesktop app for translating games, novels, subtitles, and documents in bulk using AI. Configure an API key or
• pyinfra-dev/pyinfraA Python tool that automates server setup by running shell commands over SSH. Write Python scripts describing
• ramon-victor/freegpt-webuiAn archived Python web app that provides a ChatGPT-like chat interface using free third-party AI access — no O
• openbmb/toolbenchToolBench is an AI research dataset and open-source model that teaches language models to use real software AP
• rmax/scrapy-redisScrapy-Redis connects the Scrapy web scraping library to Redis so multiple crawlers can share the same URL que
• dmunozv04/isponsorblocktviSponsorBlockTV is a self-hosted program that automatically skips sponsor segments, intros, and ads on YouTube
• idealo/imagededupimagededup is a Python library that finds duplicate and near-duplicate images in a folder using fast hashing o
• oomol-lab/pdf-craftpdf-craft converts scanned PDF books into Markdown or EPUB locally on your machine, using AI to read text from
• facebookresearch/mmfMMF is a Facebook AI research framework for building AI models that understand both images and text together —
• google/seq2seqA general-purpose TensorFlow encoder-decoder framework from Google, originally built for machine translation r
• lucidrains/dalle-pytorchA community PyTorch implementation of OpenAI's original DALL-E model that generates images from text descripti
• sezanzeb/input-remapperA Linux desktop app that lets you remap keys, gamepad buttons, mouse buttons, and joystick axes to anything yo
• deeplabcut/deeplabcutA tool for scientists to track animal body parts in video without physical markers, using AI trained on a smal
• meta-pytorch/captumA Python library that explains why a PyTorch model made a prediction by scoring which inputs or neurons contri
• mouredev/retos-programacion-2023A full year of 52 weekly programming challenges from the Spanish-language MoureDev community, each with diffic
• flashinfer-ai/flashinferFlashInfer is a Python library of highly optimized GPU kernels that makes AI language models run faster at inf
• biolab/orange3Orange is a drag-and-drop data analysis and visualization tool where you connect boxes on a canvas to load dat
• modelscope/funclipA locally-run Python tool that transcribes speech in a video file and lets you cut out specific clips by selec
• huggingface/alignment-handbookThe Alignment Handbook provides ready-to-run training scripts and YAML config files for turning a base languag
• vibora-io/viboraVibora is an experimental high-speed Python web framework for building async servers and making HTTP client re
• aws-samples/aws-cdk-examplesA collection of official AWS CDK example projects in TypeScript, Python, Java, C#, and Go — copy-paste infrast
• samuelschmidgall/agentlaboratoryPython tool that automates academic research using AI agents that review papers, run experiments, and write La
• adamchainz/django-cors-headersDjango add-on that adds CORS headers to server responses so web browsers allow JavaScript on one domain to mak
• coleam00/ottomator-agentsA growing library of open-source AI agents for the oTTomator Live Agent Studio, where you can run ready-made a
• lexiforest/curl_cffiA Python HTTP library that makes your scraper or automation script look like a real Chrome, Firefox, or Safari
• fluentpython/example-codeCode examples from the first edition of Fluent Python, an O'Reilly book on idiomatic Python for developers who
• lanmaster53/recon-ngA modular Python framework for gathering publicly available information about people, organizations, and websi
• huggingface/parler-ttsParler-TTS is an open-source text-to-speech system by Hugging Face that generates spoken audio from text using
• python-openxml/python-docxpython-docx is a Python library for creating, reading, and editing Microsoft Word .docx files entirely from co
• open-mmlab/openpcdetOpenPCDet is a Python research toolkit for training and testing AI models that detect 3D objects in LiDAR poin
• beancount/beancountBeancount is a plain-text, double-entry bookkeeping system where you keep financial records in a text file and
• tortoise/tortoise-ormTortoise ORM is an async-native Python database library with a Django-style API, letting you define tables as
• raspberrypi/documentationThe source files and build tools for the official Raspberry Pi documentation website, written in AsciiDoc — co
• cchen156/learning-to-see-in-the-darkA research project that trains a neural network to brighten extremely dark photos — turning a quick, noisy nig
• google/clusterfuzzGoogle's large-scale automated testing system that finds crashes and security bugs in software by feeding it m
• lonepatient/awesome-pretrained-chinese-nlp-modelsA curated index of pretrained Chinese NLP models — from large language models to BERT-family variants — with d
• keras-rl/keras-rlkeras-rl is a Python library of deep reinforcement learning algorithms — DQN, DDPG, and others — that plug dir
• openbmb/ultraragA Python framework for building AI document question-answering systems using RAG, with a visual drag-and-drop
• cursortouch/windows-mcpA Python server that lets AI assistants like Claude control a Windows PC — clicking, typing, navigating files,
• cryptosignal/crypto-signalCrypto Signal is a command-line tool that monitors 500+ cryptocurrencies across major exchanges, runs technica
• encode/apistarAPI Star is a Python command-line toolkit for OpenAPI and Swagger schemas: validate your API description for e
• dnhkng/gladosA Python project that builds a real-world voice AI modeled on the sardonic Portal game character — it listens,
• agentops-ai/agentopsA Python SDK that adds monitoring and debugging to AI agent workflows in two lines of code — it records every
• anishathalye/neural-styleA Python program that repaints any photo in the visual style of another image, like making a selfie look like
• azure/azure-sdk-for-pythonMicrosoft's official Python SDK for Azure cloud services, letting you write Python code to upload files, query
• spiderclub/haipproxyA self-hosted proxy pool manager that automatically collects, validates, and rotates working proxy IP addresse
• airbnb/knowledge-repoA knowledge-sharing platform that Airbnb built so data scientists can publish Jupyter notebooks and R Markdown
• cog-creators/red-discordbotAn older, unmaintained version of Red, a self-hosted multi-purpose Discord bot written in Python covering musi
• fanghua-yu/supirSUPIR is an academic AI tool that restores damaged, blurry, or low-resolution photos using large image and lan
• shawn-shan/fawkesFawkes adds invisible modifications to your face photos so that AI facial recognition systems cannot identify
• dropsdevopsorg/ecommercecrawlersA Python collection of web scrapers targeting major Chinese websites including Taobao, Weibo, Dianping, and mo
• qd-today/qdQD is a self-hosted Python tool that records HTTP requests from your browser and schedules them to run automat
• apeworx/web3.pyweb3.py is a Python library for connecting to the Ethereum blockchain — check wallet balances, send transactio
• trustedsec/ptfPTF is a Python tool that automates installing and updating penetration testing tools on Linux, organized by s
• pennyw0rth/netexecNetExec is a Python tool for authorized penetration testers to assess Windows networks and Active Directory en
• miguelgrinberg/flask-socketioA Python package that adds real-time, two-way messaging between a Flask web app and browsers using Socket.IO,
• apple/ml-depth-proAn Apple research model that takes a single photo and estimates the real-world distance of every pixel in mete
• rllm-org/rllmAn open-source Python framework that improves AI agents over time using reinforcement learning — record your a
• father-bot/chatgpt_telegram_botA Telegram bot that lets you chat with OpenAI's GPT-4 models directly inside Telegram, with streaming replies,
• karpathy/neuraltalkAn early AI research project that looks at a photo and generates a text description of it. Now deprecated and
• facebookresearch/sapiensA set of AI models trained on 300 million human images that can detect body pose, segment body parts, and esti
• karpathy/build-nanogptA step-by-step coding tutorial that teaches you how to build and train your own GPT-2 language model from scra
• apple/coremltoolsApple's official Python library for converting AI models trained in PyTorch, TensorFlow, or scikit-learn into
• zedr/clean-code-pythonA practical guide to writing cleaner, more readable Python 3 code based on Robert C. Martin's Clean Code book,

All repos · Explain any other repo →