Open-source deep-learning software
2785 open-source libraries, tools and applications tagged deep-learning, ranked by how closely deep-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.
2785 products, primary matches first, then adoption-weighted; health v2 shown.
| Product | Health v2 | Stars | Maturity |
|---|---|---|---|
| lucidrains/imagen-pytorch A PyTorch implementation of Imagen, Google's text-to-image neural network based on cascading DDPMs conditioned on T5 text embeddings. It pr… | 23 | 8428 | active |
| XPixelGroup/BasicSR BasicSR is an open-source PyTorch toolbox for image and video restoration tasks such as super-resolution, denoising, deblurring, and JPEG a… | 23 | 8391 | stable |
| QwenLM/Qwen-Image Qwen-Image is a 20B MMDiT image generation foundation model from the Qwen team, with strong complex text rendering (especially Chinese) and… | 45 | 8377 | active |
| meituan-longcat/LongCat-Video LongCat-Video is a 13.6B-parameter foundational video generation model from Meituan that unifies text-to-video, image-to-video, and video-c… | 51 | 8358 | active |
| mikel-brostrom/boxmot BoxMOT is a pluggable Python and C++ library providing state-of-the-art multi-object tracking (MOT) algorithms such as ByteTrack, BoT-SORT,… | 91 | 8301 | active |
| lltcggie/waifu2x-caffe A Windows GUI/CLI application that reimplements the waifu2x image upscaling and noise-reduction tool using the Caffe deep learning framewor… | 52 | 8212 | stable |
| jessevig/bertviz BertViz is an interactive Python library for visualizing attention mechanisms in Transformer language models such as BERT, GPT-2, and RoBER… | 47 | 8200 | stable |
| brycedrennan/imaginAIry A Python library and CLI tool (imaginairy/aimg) for generating images and videos with Stable Diffusion and Stable Video Diffusion models. I… | 60 | 8186 | active |
| InternLM/lmdeploy LMDeploy is a toolkit for compressing, quantizing, deploying, and serving large language models, built around its high-performance TurboMin… | 96 | 8107 | active |
| shenweichen/DeepCTR DeepCTR is a Python library of easy-to-use, modular, and extendible deep-learning based CTR (click-through rate) prediction models built on… | 71 | 8051 | stable |
| suno-ai/bark Bark is Suno's open-source transformer-based text-to-audio model that generates highly realistic multilingual speech, music, background noi… | 31 | 39276 | maintenance |
| lanpa/tensorboardX A Python library for writing TensorBoard event files from PyTorch, Chainer, MXNet, NumPy, and other frameworks without needing TensorFlow. … | 71 | 8002 | active |
| open-mmlab/mmpose MMPose is an open-source pose estimation toolbox and benchmark built on PyTorch as part of the OpenMMLab ecosystem. It provides implementat… | 35 | 7921 | active |
| MeiGen-AI/InfiniteTalk InfiniteTalk is an open-source model and framework for unlimited-length audio-driven talking video generation, supporting both image-to-vid… | 52 | 7888 | active |
| TensorRTx A C++ collection of popular deep learning networks (YOLO variants, ResNet, MobileNet, Swin Transformer, OCR models, and more) implemented f… | 73 | 7838 | active |
| TencentARC/GFPGAN GFPGAN is a Python library built on PyTorch that restores and enhances real-world degraded face photos using GAN-based priors. It provides … | 23 | 37683 | maintenance |
| ERNIE Official repository for Baidu's ERNIE 4.5 family of large multimodal models and ERNIEKit, an industrial-grade training toolkit built on Pad… | 62 | 7737 | active |
| meta-llama/llama-models Meta's official repository of Llama large language models with Python utilities for working with them, including model cards, licenses, and… | 53 | 7693 | active |
| babysor/MockingBird MockingBird is a PyTorch-based AI voice cloning toolbox that can clone a voice from a 5-second sample and generate arbitrary speech in real… | 51 | 36895 | maintenance |
| SkyReels SkyReels-V2 is an open-source infinite-length film/video generative model using an AutoRegressive Diffusion-Forcing architecture, released … | 45 | 7589 | active |
| face-alignment A Python library built on PyTorch that detects 2D and 3D facial landmarks in images using the FAN deep learning face alignment network. It … | 64 | 7543 | active |
| SAM 3D SAM 3D Objects is a foundation model from Meta that reconstructs full 3D shape geometry, texture, and layout from a single image, with code… | 52 | 7516 | active |
| EleutherAI/gpt-neox GPT-NeoX is EleutherAI's library for training large-scale autoregressive transformer language models on GPUs, built on NVIDIA's Megatron an… | 67 | 7471 | active |
| facebookresearch/SlowFast PySlowFast is a PyTorch-based open-source video understanding codebase from Facebook AI Research (FAIR). It provides implementations of sta… | 61 | 7432 | active |
| LargeWorldModel/LWM Large World Model (LWM) is a family of open-source 7B-parameter multimodal autoregressive transformer models trained on long videos and boo… | 26 | 7425 | active |
| apple/ml-fastvlm Official implementation of FastVLM, a vision language model with an efficient hybrid vision encoder (FastViTHD) that reduces token count an… | 62 | 7423 | active |
| arcee-ai/mergekit mergekit is a Python toolkit for merging pre-trained large language models directly in weight space, supporting many merge methods (SLERP, … | 64 | 7389 | active |
| vladmandic/sdnext SD.Next is an open-source, self-hosted WebUI server application for AI generative image and video creation built on Stable Diffusion and Di… | 77 | 7353 | active |
| google/flax Flax is a neural network library and ecosystem for JAX designed for flexibility, featuring the newer NNX API with first-class Python refere… | 99 | 7348 | active |
| Paddle Lite Paddle Lite is a high-performance, lightweight deep learning inference engine from Baidu's PaddlePaddle ecosystem, designed for mobile, emb… | 55 | 7277 | active |
| mit-han-lab/streaming-llm StreamingLLM is a research framework from MIT Han Lab implementing the Attention Sinks method (ICLR 2024) for efficient streaming language … | 28 | 7267 | stable |
| kohya-ss/sd-scripts A collection of Python training, generation, and utility scripts for Stable Diffusion and other image generation models, most widely used f… | 96 | 7242 | active |
| Zyphra/Zonos Zonos-v0.1 is an open-weight text-to-speech model trained on over 200k hours of multilingual speech, with a Python library for inference. I… | 22 | 7236 | active |
| TensorBoard TensorBoard is TensorFlow's visualization toolkit, a suite of web applications for inspecting and understanding machine learning training r… | 82 | 7231 | active |
| deepseek-ai/DeepSpec DeepSpec is a full-stack Python codebase from DeepSeek for training and evaluating draft models used in speculative decoding of large langu… | 51 | 7207 | active |
| BVLC/caffe Caffe is a fast, modular deep learning framework developed by Berkeley AI Research, focused on convolutional neural networks for vision, sp… | 23 | 34549 | maintenance |
| OpenPose OpenPose is a real-time multi-person keypoint detection library that jointly estimates human body, face, hand, and foot keypoints (135 tota… | 23 | 34493 | maintenance |
| ControlNet ControlNet is a neural network architecture that adds conditional control (edges, poses, depth, etc.) to pretrained text-to-image diffusion… | 32 | 34128 | maintenance |
| yangchris11/samurai SAMURAI is the official implementation of a zero-shot visual object tracker built on top of Segment Anything Model 2 (SAM 2), using a motio… | 24 | 7116 | active |
| threestudio-project/threestudio threestudio is a unified open-source framework for 3D content generation from text prompts, single images, and few-shot images by lifting 2… | 21 | 7069 | active |
| TheTom/turboquant_plus TurboQuant+ is a Python reference implementation of the TurboQuant KV cache compression method (ICLR 2026), using PolarQuant codebooks and … | 54 | 7036 | active |
| apple/corenet CoreNet is Apple's deep neural network training toolkit for training standard and novel small and large-scale models, including foundation … | 68 | 7008 | active |
| deeppavlov/DeepPavlov DeepPavlov is an open-source Python NLP library built on PyTorch and Hugging Face transformers for developing, training, and deploying stat… | 35 | 6995 | active |
| PaddlePaddle/models PaddlePaddle's officially maintained industry-grade model repository containing 600+ models across computer vision, NLP, speech, recommenda… | 23 | 6931 | active |
| eriklindernoren/ML-From-Scratch A Python library providing bare-bones NumPy implementations of fundamental machine learning models and algorithms, from linear regression t… | 32 | 32980 | maintenance |
| ClearML ClearML is an open-source Python SDK and platform for MLOps/LLMOps that provides auto-logged experiment tracking, data versioning, pipeline… | 96 | 6903 | stable |
| SenseNova SenseNova-U is a series of open-weight unified multimodal models (e.g., SenseNova-U1.5-8B-MoT) built on the NEO-unify architecture that com… | 58 | 6884 | active |
| facebookresearch/fairseq Fairseq is a PyTorch-based sequence modeling toolkit from Facebook AI Research for training custom models for translation, summarization, l… | 10 | 32211 | maintenance |
| FoundationVision/ByteTrack ByteTrack is a PyTorch-based multi-object tracking (MOT) library implementing the ECCV 2022 paper 'Multi-Object Tracking by Associating Eve… | 32 | 6740 | stable |
| kijai/ComfyUI-WanVideoWrapper ComfyUI custom nodes wrapping WanVideo (Wan2.1) and related video generation models. It provides a standalone sandbox for quickly implement… | 55 | 6737 | active |
| flashinfer-ai/flashinfer FlashInfer is a GPU kernel library and kernel generator for LLM inference, providing unified APIs for attention, GEMM, and MoE operations w… | 92 | 6728 | active |
| tencent-ailab/IP-Adapter IP-Adapter is a lightweight (22M parameter) adapter that adds image prompt capability to pretrained text-to-image diffusion models like Sta… | 29 | 6694 | stable |
| yangjianxin1/Firefly Firefly is an open-source one-stop training tool for large language models, supporting pretraining, instruction fine-tuning (SFT), and DPO … | 21 | 6653 | active |
| TMElyralab/MuseTalk MuseTalk is a real-time, high-fidelity lip-sync model that modifies a face region in video according to input audio via latent space inpain… | 42 | 6651 | active |
| AILab-CVC/YOLO-World YOLO-World is a real-time open-vocabulary object detection model and Python toolkit from Tencent AI Lab and HUST, published at CVPR 2024. I… | 26 | 6579 | active |
| open-mmlab/mmdetection3d MMDetection3D is OpenMMLab's next-generation platform for general 3D object detection, built on PyTorch. It provides a modular toolbox with… | 23 | 6535 | active |
| rtqichen/torchdiffeq torchdiffeq is a PyTorch library of differentiable ordinary differential equation (ODE) solvers, best known as the canonical implementation… | 35 | 6482 | stable |
| open-mmlab/mmcv MMCV is the foundational computer vision library for the OpenMMLab ecosystem, providing image/video I/O, data transformations, and CUDA ope… | 48 | 6475 | stable |
| HVision-NKU/StoryDiffusion StoryDiffusion is the official implementation of a NeurIPS 2024 Spotlight paper introducing Consistent Self-Attention for character-consist… | 25 | 6475 | active |
| ByteDance-Seed/Depth-Anything-3 Depth Anything 3 (DA3) is a transformer-based model that predicts spatially consistent depth and geometry from any number of visual inputs,… | 56 | 6445 | active |
| haifengl/smile SMILE is a comprehensive, high-performance machine learning framework for the JVM with idiomatic APIs for Java, Scala, and Kotlin. It cover… | 100 | 6423 | active |
| tensorflow/serving TensorFlow Serving is a flexible, high-performance serving system for machine learning models designed for production environments. It mana… | 89 | 6364 | stable |
| StyleTTS StyleTTS 2 is a PyTorch text-to-speech model that uses style diffusion and adversarial training with large speech language models (e.g., Wa… | 30 | 6362 | active |
| docTR docTR is a Python OCR library that extracts text from documents and images using a two-stage deep learning approach: text detection followe… | 86 | 6359 | active |
| canopyai/Orpheus-TTS Orpheus TTS is an open-source text-to-speech system built on a Llama-3b backbone that produces human-sounding speech with emotion control a… | 42 | 6347 | active |
| KevinMusgrave/pytorch-metric-learning A PyTorch library providing modular losses, miners, samplers, and testers for deep metric learning. It supports building complete train/tes… | 43 | 6342 | active |
| RangiLyu/nanodet NanoDet-Plus is a super fast, lightweight anchor-free object detection model implemented in PyTorch, with model sizes as small as 980KB (IN… | 23 | 6269 | stable |
| PaddlePaddle/PaddleX PaddleX is a low-code, all-in-one AI development tool built on the PaddlePaddle framework, bundling 200+ pretrained models into 33 producti… | 86 | 6268 | active |
| meta-pytorch/gpt-fast A minimal (<1000 lines) PyTorch-native implementation of fast transformer text generation, demonstrating low-latency LLM inference with int… | 41 | 6254 | active |
| meta-llama/llama3 The official Meta repository for Llama 3, providing model weights download scripts, tokenizer, and minimal example code for running inferen… | 10 | 29219 | maintenance |
| open-edge-platform/anomalib Anomalib is a Python deep learning library for anomaly detection, offering state-of-the-art unsupervised algorithms for detecting and local… | 96 | 6196 | active |
| skorch-dev/skorch skorch is a Python library that wraps PyTorch neural networks in a scikit-learn compatible API, providing estimators like NeuralNetClassifi… | 82 | 6179 | active |
| Akegarasu/lora-scripts SD-Trainer is a GUI application and set of scripts for training LoRA and Dreambooth fine-tunes of Stable Diffusion diffusion models, wrappi… | 66 | 6166 | active |
| FlashDrive DFlash is a lightweight block diffusion model used as a draft model for speculative decoding of large language models, drafting entire toke… | 65 | 6134 | active |
| timeseriesAI/tsai tsai is an open-source deep learning library built on PyTorch and fastai for time series and sequential data tasks such as classification, … | 78 | 6124 | active |
| bytedance/LatentSync LatentSync is an end-to-end lip-sync framework from ByteDance based on audio-conditioned latent diffusion models, using Stable Diffusion to… | 30 | 6120 | active |
| bytedance/MegaTTS3 MegaTTS 3 is ByteDance's open-source PyTorch text-to-speech model with a lightweight 0.45B-parameter Diffusion Transformer backbone. It pro… | 56 | 6096 | active |
| om-ai-lab/VLM-R1 VLM-R1 is a framework for training R1-style large vision-language models using reinforcement learning (GRPO) on top of Qwen2.5-VL. It provi… | 60 | 6029 | active |
| OFA-Sys/Chinese-CLIP Chinese-CLIP is a Chinese version of the CLIP model trained on ~200 million Chinese image-text pairs, built on open_clip. It provides APIs,… | 62 | 6014 | active |
| lucidrains/x-transformers A concise PyTorch library implementing full-attention transformer architectures (encoder, decoder, encoder-decoder, and vision transformers… | 83 | 5960 | active |
| fla-org/flash-linear-attention A PyTorch library providing hardware-efficient implementations of emerging sequence model architectures, including linear attention, sparse… | 86 | 5940 | active |
| pytorch/torchtitan torchtitan is a PyTorch-native platform for large-scale training of generative AI models, offering a clean-room implementation of PyTorch's… | 81 | 5830 | active |
| google-deepmind/gemma The official JAX-based Python library from Google DeepMind for running, sampling from, and fine-tuning the Gemma family of open-weight larg… | 85 | 5796 | active |
| DeepLabCut DeepLabCut is an open-source Python toolbox for markerless 2D and 3D pose estimation of user-defined body parts using deep neural networks … | 94 | 5784 | stable |
| NVIDIA/DALI NVIDIA DALI is a GPU-accelerated data loading and preprocessing library with optimized building blocks and an execution engine for deep lea… | 91 | 5761 | active |
| pjreddie/darknet Darknet is an open-source neural network framework written in C and CUDA, best known as the original home of the YOLO real-time object dete… | 32 | 26503 | maintenance |
| apple/ml-depth-pro Depth Pro is Apple's reference implementation of a foundation model for zero-shot metric monocular depth estimation, producing sharp high-r… | 66 | 5737 | active |
| open-mmlab/OpenPCDet OpenPCDet is a PyTorch-based open-source toolbox for LiDAR-based 3D object detection. It provides official implementations of models like P… | 49 | 5718 | active |
| Gemma The official PyTorch implementation of Google's Gemma family of open large language models, including text-only and multimodal variants. It… | 10 | 5710 | active |
| meta-pytorch/captum Captum is a model interpretability and understanding library for PyTorch, providing implementations of algorithms like Integrated Gradients… | 77 | 5703 | active |
| huggingface/alignment-handbook A collection of robust training recipes and scripts from Hugging Face for aligning large language models with human and AI preferences, cov… | 72 | 5696 | active |
| alexlenail/NN-SVG A web-based tool for creating publication-ready neural network architecture diagrams parametrically, supporting FCNN, LeNet-style CNN, and … | 57 | 5691 | stable |
| Fanghua-Yu/SUPIR SUPIR is a Python-based photo-realistic image restoration system built on SDXL diffusion priors and LLaVA captioning, presented at CVPR 202… | 33 | 5685 | active |
| lyuwenyu/RT-DETR Official implementation of RT-DETR and RT-DETRv2, real-time object detection transformers that outperform YOLO models, in PyTorch and Paddl… | 75 | 5609 | active |
| Parler‑TTS Parler-TTS is a lightweight text-to-speech library from Hugging Face that generates high-quality, natural-sounding speech controllable via … | 26 | 5587 | active |
| matterport/Mask_RCNN A Python implementation of the Mask R-CNN model for object detection and instance segmentation, built on Keras and TensorFlow with a ResNet… | 23 | 25573 | maintenance |
| obss/sahi SAHI (Slicing Aided Hyper Inference) is a Python vision library for detecting small objects in large images via sliced/tiled inference, wor… | 94 | 5513 | active |
| LaurentMazare/tch-rs tch-rs is a Rust crate providing thin bindings to the C++ API of PyTorch (libtorch), staying close to the original API. It enables tensor o… | 64 | 5503 | active |
| mosaicml/composer Composer is an open-source PyTorch-based deep learning training library by MosaicML (now Databricks) for training neural networks faster an… | 66 | 5499 | active |
| HarisIqbal88/PlotNeuralNet A LaTeX/TikZ-based library for drawing neural network architecture diagrams, with a Python interface for generating the TikZ code. It is co… | 23 | 25035 | maintenance |