Open-source machine-learning software
4007 open-source libraries, tools and applications tagged machine-learning, ranked by how closely machine-learning is their primary purpose, then by adoption and maintenance health. Each entry links to a page with sources, use cases and when to choose or avoid it.
4007 products, primary matches first, then adoption-weighted; health v2 shown.
| Product | Health v2 | Stars | Maturity |
|---|---|---|---|
| SysCV/sam-hq HQ-SAM (Segment Anything in High Quality) upgrades Meta's Segment Anything Model with a learnable High-Quality Output Token for accurate ze… | 45 | 4262 | active |
| ThilinaRajapakse/simpletransformers Simple Transformers is a Python library built on Hugging Face Transformers that lets users train, fine-tune, and evaluate Transformer model… | 57 | 4256 | active |
| kubeflow/pipelines Kubeflow Pipelines is a Kubernetes-native platform for building, deploying, and orchestrating reusable end-to-end machine learning workflow… | 98 | 4248 | stable |
| ZhengPeng7/BiRefNet BiRefNet is a PyTorch implementation of the CAAI AIR 2024 paper 'Bilateral Reference for High-Resolution Dichotomous Image Segmentation'. I… | 64 | 4247 | active |
| ArcInstitute/evo2 Evo 2 is a DNA foundation language model (1B-40B parameters) that models genomes at single-nucleotide resolution with up to 1 million base … | 56 | 4247 | active |
| danveloper/flash-moe An application for running large language models locally on small laptops, written in Objective-C. It enables on-device inference of big mo… | 44 | 4238 | active |
| SwanHubX/SwanLab SwanLab is an open-source AI training tracking and visualization platform with a Python SDK, CLI, and cloud or self-hosted dashboard. It in… | 88 | 4237 | active |
| cvg/Hierarchical-Localization hloc is a modular Python toolbox for state-of-the-art 6-DoF visual localization, combining image retrieval and feature matching (SuperPoint… | 44 | 4232 | active |
| stemdeckapp/stemdeck StemDeck is a free, local stem separation application that splits songs into up to six stems (vocals, drums, bass, guitar, piano, other) us… | 78 | 4230 | active |
| MaartenGr/KeyBERT KeyBERT is a minimal Python library that extracts keywords and keyphrases from documents using BERT embeddings and cosine similarity. It re… | 64 | 4223 | stable |
| SylphAI-Inc/AdalFlow AdalFlow is a PyTorch-like Python library for building and auto-optimizing LLM application workflows such as chatbots, RAG pipelines, and a… | 62 | 4220 | active |
| facebookresearch/nevergrad Nevergrad is a Python toolbox for gradient-free (derivative-free) optimization, providing many optimization algorithms such as NGOpt and di… | 79 | 4215 | active |
| thinking-machines-lab/tinker-cookbook Tinker Cookbook is a Python library of realistic examples and abstractions for post-training (fine-tuning) language models via the Tinker A… | 84 | 4199 | active |
| torchgeo/torchgeo TorchGeo is a PyTorch domain library, similar to torchvision, providing datasets, samplers, transforms, and pre-trained models specific to … | 93 | 4192 | active |
| iterative/cml CML (Continuous Machine Learning) is an open-source CLI tool that brings CI/CD practices to machine learning projects on GitHub, GitLab, an… | 41 | 4191 | active |
| JEPA Official PyTorch implementation of V-JEPA, a self-supervised method for learning visual representations from video using a joint-embedding … | 26 | 4179 | active |
| Spark NLP Spark NLP is an open-source natural language processing library built natively on Apache Spark, providing scalable NLP annotations and tran… | 98 | 4160 | stable |
| OpenManus/OpenManus-RL OpenManus-RL is an open-source project for reinforcement-learning-based tuning of LLM agents, built on the verl training framework. It prov… | 44 | 4159 | active |
| SamGeo SamGeo (segment-geospatial) is a Python package that applies Meta AI's Segment Anything Model (SAM, SAM2, SAM3, HQ-SAM) to geospatial data … | 93 | 4147 | active |
| city96/ComfyUI-GGUF A ComfyUI custom node pack that adds GGUF quantization support for loading diffusion models like FLUX and Stable Diffusion 3.5 in low-bitra… | 48 | 4138 | active |
| justadudewhohacks/face-api.js A JavaScript face detection and face recognition library built on top of tensorflow.js, usable in the browser and Node.js. It provides mode… | 23 | 17981 | maintenance |
| zml/zml ZML is a production LLM inference stack written in Zig, built on MLIR and OpenXLA, that compiles models to run at peak performance across N… | 69 | 4136 | active |
| marcelroed/gigatoken Gigatoken is a Rust-based tokenizer library for language modeling that tokenizes text at GB/s throughput, claiming ~1000x speedups over Hug… | 58 | 4134 | active |
| IDEA-CCNL/Fengshenbang-LM Fengshenbang-LM is an open-source suite of Chinese large language models and a PyTorch training framework from IDEA Research's CCNL lab, ai… | 67 | 4123 | active |
| openai/transformer-debugger Transformer Debugger (TDB) is an OpenAI tool for investigating specific behaviors of small language models, combining automated interpretab… | 57 | 4123 | active |
| huggingface/distil-whisper Distil-Whisper is a distilled version of OpenAI's Whisper model for English speech recognition, offering 6x faster inference, 49% fewer par… | 28 | 4120 | active |
| GuyTevet/motion-diffusion-model Official PyTorch implementation of the Human Motion Diffusion Model (MDM) paper, generating 3D human motion sequences from text prompts usi… | 48 | 4119 | active |
| CSGHub CSGHub is an open-source, self-hostable platform for managing the full lifecycle of LLM assets including models, datasets, spaces, and agen… | 90 | 4115 | active |
| princeton-vl/RAFT Official PyTorch implementation of RAFT (Recurrent All Pairs Field Transforms for Optical Flow), an ECCV 2020 model for estimating dense op… | 46 | 4103 | stable |
| PaddlePaddle/PaddleRec PaddleRec is a large-scale recommendation algorithm library built on PaddlePaddle, containing classic and state-of-the-art recommendation m… | 26 | 4086 | active |
| google-deepmind/acme Acme is a Python library of reinforcement learning building blocks and agent implementations from DeepMind. It provides readable reference … | 67 | 4080 | active |
| open-spaced-repetition/fsrs4anki FSRS4Anki is a modern spaced-repetition scheduler for Anki based on the Free Spaced Repetition Scheduler (FSRS) algorithm. It replaces Anki… | 74 | 4074 | active |
| DLRM A PyTorch-based implementation of the Deep Learning Recommendation Model (DLRM) from Facebook Research, which processes dense and sparse fe… | 56 | 4074 | stable |
| uxlfoundation/oneDNN oneDNN is an open-source cross-platform performance library providing optimized building blocks (primitives) for deep learning applications… | 99 | 4071 | stable |
| FedML FedML (TensorOpera) is a unified Python library for large-scale distributed training, model serving, and federated learning across GPU clou… | 41 | 4070 | active |
| lllyasviel/Paints-UNDO Paints-UNDO is a family of deep learning models that take an image as input and generate the step-by-step drawing sequence (sketching, inki… | 37 | 4063 | active |
| StarsfieldAI/R1-V R1-V is an open-source research codebase for training vision-language models with reinforcement learning (RLVR/GRPO), demonstrating strong … | 18 | 4063 | active |
| deepchecks/deepchecks Deepchecks is an open-source Python library for continuous validation of machine learning models and data, covering testing, CI integration… | 57 | 4060 | active |
| tensorflow/tensor2tensor Tensor2Tensor (T2T) is a Python library of deep learning models and datasets built on TensorFlow, developed by the Google Brain team to mak… | 10 | 17473 | maintenance |
| Dora Dora is a Rust-based dataflow middleware framework for building real-time robotics and AI applications, where applications are modeled as d… | 95 | 4030 | active |
| OSU-NLP-Group/HippoRAG HippoRAG is a Python RAG framework inspired by human long-term memory that combines LLMs, knowledge graphs, and Personalized PageRank to in… | 57 | 4021 | active |
| lucidrains/vector-quantize-pytorch A PyTorch library implementing vector and scalar quantization, including Residual VQ and techniques like DiVeQ codebook updates. It origina… | 81 | 4019 | active |
| urchade/GLiNER GLiNER is a Python framework for training and deploying lightweight, zero-shot Named Entity Recognition models that can extract arbitrary e… | 87 | 3997 | active |
| Marin Marin is an open-source Python framework and research program for training foundation models, covering the full pipeline from data curation… | 79 | 3995 | active |
| huggingface/huggingface_hub The official Python client and CLI for the Hugging Face Hub, a platform for sharing machine learning models, datasets, and apps. It lets us… | 95 | 3988 | stable |
| IREE IREE is an MLIR-based end-to-end machine learning compiler and runtime that lowers models from frameworks like PyTorch, TensorFlow, JAX, an… | 95 | 3980 | active |
| AnswerDotAI/RAGatouille RAGatouille is a Python library that makes it easy to use and train state-of-the-art late-interaction retrieval models like ColBERT in any … | 23 | 3964 | active |
| Lightning-AI/LitServe LitServe is a Python framework for building custom AI inference servers with full control over batching, routing, streaming, and scaling lo… | 92 | 3956 | active |
| thu-ml/SageAttention SageAttention is a family of quantized attention kernels (INT8/FP8/FP4) that accelerate transformer inference 2-5x over FlashAttention with… | 38 | 3956 | active |
| NeuralOperator A PyTorch library for learning neural operators, which map between function spaces rather than finite-dimensional vectors. It provides the … | 80 | 3955 | active |
| Nunchaku Nunchaku is a high-performance inference engine for 4-bit quantized diffusion models (and LLMs) based on the SVDQuant technique from an ICL… | 73 | 3953 | active |
| TransformerLensOrg/TransformerLens TransformerLens is a Python library for mechanistic interpretability of GPT-style transformer language models. It lets researchers load tho… | 98 | 3949 | active |
| stanford-futuredata/ColBERT ColBERT is a fast and accurate neural retrieval model that encodes passages and queries into token-level embedding matrices and scores them… | 40 | 3941 | active |
| thuml/Transfer-Learning-Library TLlib is a PyTorch-based open-source library for transfer learning, covering domain adaptation, task adaptation (finetuning), and domain ge… | 23 | 3940 | active |
| huggingface/smollm Hugging Face's repository for the SmolLM and SmolVLM families of compact, fully open language and vision-language models, including trainin… | 67 | 3926 | active |
| GPyTorch GPyTorch is a Gaussian process library built on PyTorch, designed for scalable, flexible, and modular GP models. It uses linear algebra tec… | 82 | 3925 | stable |
| hustvl/Vim Vision Mamba (Vim) is a PyTorch implementation of a generic vision backbone built on bidirectional Mamba state space models, published at I… | 27 | 3913 | active |
| NVIDIA/k8s-device-plugin NVIDIA's official Kubernetes device plugin that exposes NVIDIA GPUs as schedulable resources in Kubernetes clusters. It supports GPU sharin… | 94 | 3901 | stable |
| safetensors/safetensors Safetensors is a simple, secure file format and library for storing and distributing tensors, designed as a fast zero-copy alternative to p… | 87 | 3901 | stable |
| vllm-project/llm-compressor LLM Compressor is a Python library for applying quantization and pruning algorithms to large language models, producing compressed-tensors … | 91 | 3897 | active |
| openai/baselines OpenAI Baselines is a collection of high-quality reference implementations of reinforcement learning algorithms such as DQN, PPO, and ACKTR… | 32 | 16764 | maintenance |
| PySR PySR is a high-performance symbolic regression library that searches for symbolic mathematical expressions optimizing a given objective, im… | 99 | 3893 | stable |
| Plachtaa/seed-vc Seed-VC is a Python tool and model for zero-shot voice conversion, real-time voice conversion, and singing voice conversion, cloning a voic… | 10 | 3892 | active |
| LazyAGI/LazyLLM LazyLLM is a low-code Python framework for building multi-agent LLM applications, covering prototype assembly, data feedback, and iterative… | 92 | 3888 | active |
| hkust-nlp/simpleRL-reason A research codebase from HKUST-NLP implementing a simple reinforcement learning recipe (rule-based rewards on GSM8K/Math data) to train LLM… | 44 | 3877 | active |
| ComfyUI-3D-Pack An extensive ComfyUI custom node suite for processing 3D inputs like meshes and UV textures using algorithms such as 3D Gaussian Splatting … | 47 | 3871 | active |
| Stability-AI/stable-audio-tools Stability AI's training and inference toolkit for conditional audio generation models, including Stable Audio Open. It supports training cu… | 73 | 3868 | active |
| danijar/dreamerv3 A Python/JAX reimplementation of DreamerV3, a model-based reinforcement learning algorithm that learns a world model and trains an actor-cr… | 76 | 3859 | active |
| open-mmlab/mmpretrain MMPretrain is OpenMMLab's PyTorch-based toolbox and benchmark for image classification model pre-training, covering supervised, self-superv… | 23 | 3855 | active |
| roboflow/trackers A Python library of clean-room, Apache 2.0 implementations of multi-object tracking algorithms including SORT, ByteTrack, OC-SORT, BoT-SORT… | 85 | 3850 | active |
| MrNeRF/LichtFeld-Studio LichtFeld Studio is a native open-source desktop application for 3D Gaussian Splatting that combines training, real-time inspection, splat … | 96 | 3849 | active |
| mseitzer/pytorch-fid A PyTorch port of the official TensorFlow implementation of the Fréchet Inception Distance (FID), a metric for measuring similarity between… | 23 | 3847 | stable |
| piskvorky/gensim Gensim is a Python library for topic modelling, document indexing, and similarity retrieval over large text corpora. It provides memory-ind… | 43 | 16492 | maintenance |
| Netflix/maestro Netflix Maestro is a general-purpose workflow orchestrator providing workflow-as-a-service for data, ML, and software pipelines. It schedul… | 68 | 3843 | active |
| google-research/scenic Scenic is a JAX-based library from Google Research focused on attention-based models for computer vision, providing shared lightweight libr… | 76 | 3842 | active |
| shenweichen/GraphEmbedding A Python library providing implementations of classic graph embedding algorithms including DeepWalk, LINE, Node2Vec, SDNE, and Struc2Vec. I… | 64 | 3841 | active |
| predibase/lorax LoRAX is a multi-LoRA inference server that serves thousands of fine-tuned LLM adapters on a single GPU by sharing a common base model. It … | 60 | 3835 | active |
| benfred/implicit A fast Python library implementing collaborative filtering recommendation algorithms for implicit feedback datasets, including ALS, BPR, lo… | 68 | 3829 | active |
| lightly-ai/lightly LightlySSL is a Python library built on PyTorch for self-supervised learning on images, offering modular implementations of methods like Si… | 91 | 3821 | active |
| google/deepvariant DeepVariant is a deep learning-based genomic variant caller that converts aligned DNA sequencing reads (BAM/CRAM) into pileup image tensors… | 78 | 3818 | stable |
| facebookresearch/map-anything MapAnything is an open-source research framework from Meta and CMU for universal feed-forward metric 3D reconstruction using an end-to-end … | 71 | 3813 | active |
| ddbourgin/numpy-ml numpy-ml is a collection of machine learning models and algorithms implemented exclusively in NumPy and the Python standard library, coveri… | 32 | 16323 | maintenance |
| SciSharp/LLamaSharp LLamaSharp is a cross-platform C#/.NET library that wraps llama.cpp to run LLaMA-family and other LLMs (including multimodal LLaVA) efficie… | 89 | 3803 | active |
| AI4Finance-Foundation/FinRL-Trading FinRL-X is an open-source, AI-native modular infrastructure for quantitative trading that unifies data processing, strategy composition, ba… | 76 | 3782 | active |
| aubio aubio is a C library (with Python bindings and CLI tools) for labeling music and sounds, detecting events such as note onsets, pitch, and t… | 53 | 3765 | stable |
| HazyResearch/ThunderKittens ThunderKittens is a C++/CUDA framework of tile-based primitives for writing fast deep learning GPU kernels. It embeds natively into CUDA so… | 68 | 3754 | active |
| genmoai/mochi Mochi 1 is Genmo's open-source, state-of-the-art text-to-video generation model released under Apache 2.0, with a Python API, CLI, and Grad… | 67 | 3750 | active |
| starVLA/starVLA StarVLA is an open-source, Lego-like modular codebase for developing Vision-Language-Action (VLA) models for generalist robots. It unifies … | 67 | 3745 | active |
| ferdous-alam/GenCAD GenCAD is a research codebase for image-conditioned CAD model generation using transformer-based contrastive representations (CCIP) and dif… | 33 | 3744 | active |
| DLR-RM/BlenderProc BlenderProc is a procedural Python pipeline built on Blender for generating photorealistic synthetic training images with ground-truth anno… | 59 | 3740 | active |
| Polyaxon Polyaxon is an open-source AI engineering control plane that teams self-host on their own Kubernetes clusters to schedule, track, and gover… | 75 | 3735 | active |
| MTG/essentia Essentia is an open-source C++ library for audio analysis and music information retrieval, offering a large collection of reusable DSP algo… | 66 | 3733 | stable |
| Belval/TextRecognitionDataGenerator A Python library and CLI tool (trdg) that generates synthetic text images for training OCR and text recognition models. It supports multipl… | 23 | 3696 | stable |
| 0xShug0/audio.cpp audio.cpp is a pure C++ inference engine for audio models built on ggml, supporting TTS, STT, VAD, voice conversion, music generation, and … | 82 | 3690 | active |
| PAIR-code/lit The Learning Interpretability Tool (LIT) is a visual, interactive web-based tool for understanding and debugging ML model behavior across t… | 72 | 3667 | active |
| NVlabs/GR00T-WholeBodyControl NVIDIA's unified platform for developing, training, and deploying whole-body controllers for humanoid robots, including the decoupled WBC m… | 60 | 3654 | active |
| facebookresearch/sam-audio SAM-Audio is Meta's foundation model for isolating any sound in audio using text, visual, or temporal prompts. This repository provides inf… | 52 | 3654 | active |
| DI-engine DI-engine is an open-source reinforcement learning framework from OpenDILab that provides comprehensive implementations of deep RL algorith… | 44 | 3647 | active |
| kaldi-asr/kaldi Kaldi is a C++ toolkit for speech recognition research and development, including acoustic modeling, feature extraction, decoding, and spea… | 48 | 15502 | maintenance |
| mit-han-lab/llm-awq AWQ is a Python library implementing Activation-aware Weight Quantization for compressing large language models to low-bit (INT3/4) weights… | 41 | 3638 | active |