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 |
|---|---|---|---|
| awentzonline/image-analogies A Python library implementing neural image analogies using VGG16 feature maps with PatchMatch-based matching and blending, based on the 'Im… | 23 | 3502 | maintenance |
| RIFE RIFE is a deep learning model for real-time video frame interpolation, estimating intermediate flow between frames to generate smooth slow-… | 73 | 1029 | active |
| microsoft/Tutel Tutel is Microsoft's optimized Mixture-of-Experts (MoE) library for efficient training and inference of large language models, featuring dy… | 79 | 1028 | active |
| kuprel/min-dalle min(DALL·E) is a fast, minimal PyTorch port of DALL·E Mini/Mega stripped down for text-to-image inference, with only numpy, requests, pillo… | 27 | 3490 | maintenance |
| waifu2x (nunif) waifu2x is an image super-resolution and noise-reduction tool for anime-style art (and photos) using deep convolutional neural networks, or… | 75 | 3486 | maintenance |
| bowang-lab/U-Mamba U-Mamba is a hybrid CNN-state-space-model (Mamba) network for biomedical image segmentation, built on top of the nnU-Net framework. It comb… | 27 | 1025 | active |
| aim-uofa/AdelaiDet AdelaiDet is an open-source Python toolbox built on Detectron2 that implements multiple instance-level detection and recognition algorithms… | 32 | 3479 | maintenance |
| microsoft/tensorwatch TensorWatch is a Python library from Microsoft Research for debugging, monitoring, and visualizing machine learning training in real time, … | 62 | 3478 | maintenance |
| DingXiaoH/RepVGG RepVGG is a PyTorch implementation of the VGG-style ConvNet architecture from the CVPR 2021 paper, achieving over 84% top-1 ImageNet accura… | 32 | 3477 | maintenance |
| open-mmlab/mmyolo MMYOLO is the OpenMMLab toolbox and benchmark for the YOLO series of object detection models, implemented on PyTorch. It provides unified i… | 23 | 3477 | maintenance |
| fla-org/native-sparse-attention Efficient Triton kernel implementations of Native Sparse Attention (NSA), a hardware-aligned, natively trainable sparse attention mechanism… | 47 | 1023 | active |
| MeshAnything MeshAnything is an autoregressive transformer model that generates artist-created 3D meshes (up to 1600 faces in V2) aligned with a given s… | 28 | 1023 | active |
| TencentARC/SEED-Voken SEED-Voken is a collection of visual tokenizers (Open-MAGVIT2 and IBQ) that convert images and videos into discrete tokens for autoregressi… | 45 | 1022 | active |
| JiahuiYu/generative_inpainting An open-source implementation of DeepFill v1/v2 generative image inpainting models, featuring Contextual Attention (CVPR 2018) and Gated Co… | 32 | 3471 | maintenance |
| tensorlayer/SRGAN Reference implementation of SRGAN, a generative adversarial network for photo-realistic single image super-resolution, built on TensorLayer… | 23 | 3467 | maintenance |
| AI-Optimizer AI-Optimizer is a deep reinforcement learning toolkit from TJU's DRL lab offering algorithm libraries spanning model-free, model-based, off… | 32 | 3465 | maintenance |
| richzhang/colorization A Python library implementing automatic colorization of grayscale photos using deep neural networks from the ECCV 2016 'Colorful Image Colo… | 32 | 3463 | maintenance |
| deepseek-ai/DeepSeek-Math DeepSeekMath is a 7B open language model specialized in mathematical reasoning, initialized from DeepSeek-Coder and trained on 500B math-re… | 26 | 3457 | maintenance |
| ShangtongZhang/DeepRL A modularized PyTorch implementation of popular deep reinforcement learning algorithms including DQN variants, PPO, DDPG, TD3, A2C, and Opt… | 23 | 3456 | maintenance |
| google/prompt-to-prompt Google's official implementation of the Prompt-to-Prompt paper, which enables text-driven image editing in Latent Diffusion and Stable Diff… | 10 | 3456 | maintenance |
| Alpha-VLLM/Lumina-Image-2.0 Lumina-Image 2.0 is an open-source 2.6B-parameter text-to-image generation framework built on a unified Next-DiT architecture with a unifie… | 55 | 1017 | active |
| Towhee Towhee is a Python framework for building ETL pipelines that process unstructured data (images, video, text, audio) into embeddings using s… | 23 | 3451 | maintenance |
| tensorflow/adanet AdaNet is a lightweight TensorFlow-based AutoML framework that automatically learns high-quality neural network architectures and ensembles… | 10 | 3451 | maintenance |
| HobbitLong/SupContrast A PyTorch reference implementation of the Supervised Contrastive Learning paper (SupCon loss) that also supports SimCLR when labels are omi… | 32 | 3449 | maintenance |
| EvolvingLMMs-Lab/Otter Otter is a multi-modal vision-language model built on OpenFlamingo, instruction-tuned on the MIMIC-IT dataset with image and video understa… | 21 | 3445 | maintenance |
| DyNet DyNet is a C++ neural network toolkit with Python bindings, designed for efficient CPU/GPU training of networks with dynamic per-instance s… | 23 | 3429 | maintenance |
| TensorSpeech/TensorFlowASR TensorFlowASR is a Python library implementing automatic speech recognition architectures such as DeepSpeech2, Jasper, RNN Transducer, Cont… | 86 | 1011 | active |
| huawei-noah/noah-research A collection of research code subprojects released by Huawei Noah's Ark Lab, each in its own directory. It is not an official Huawei produc… | 73 | 1011 | active |
| mni-ml/framework A TypeScript machine learning framework with Rust native backends for CPU, CUDA, and WebGPU. It provides autograd, tensor operations, built… | 48 | 1010 | active |
| sail-sg/EditAnything Edit Anything is a Python application for text-guided image editing and generation, combining Segment Anything, ControlNet, BLIP2, and Stab… | 31 | 3422 | maintenance |
| davidtvs/pytorch-lr-finder A PyTorch library implementing the learning rate range test from Leslie Smith's cyclical learning rates paper, including the fastai-tweaked… | 32 | 1009 | active |
| facebookresearch/Mask2Former Mask2Former is the official PyTorch implementation of the CVPR 2022 paper 'Masked-attention Mask Transformer for Universal Image Segmentati… | 10 | 3417 | maintenance |
| shaoshengsong/DeepSORT A real-time multi-object tracking application in C++17 using Qt 6, OpenCV, and ONNX Runtime, running YOLO detection with four pluggable tra… | 83 | 1008 | active |
| shaoanlu/faceswap-GAN A Jupyter Notebook-based implementation of face swapping using a denoising autoencoder architecture enhanced with adversarial losses, VGGFa… | 32 | 3415 | maintenance |
| makegirlsmoe/makegirlsmoe_web The React front-end for MakeGirlsMoe, a web app that generates anime character portraits using a GAN model. It provides the interactive UI … | 32 | 3413 | maintenance |
| SparseDrive SparseDrive is a research codebase implementing an end-to-end autonomous driving framework based on sparse scene representation. It unifies… | 58 | 1006 | active |
| minimaxir/gpt-2-simple A Python package that simplifies fine-tuning OpenAI's GPT-2 text-generation model (124M/355M) on custom text and generating text from the r… | 23 | 3405 | maintenance |
| henry123-boy/SpaTrackerV2 SpatialTrackerV2 is a research library from ICCV 2025 that performs 3D point tracking in videos, lifting tracked points into 3D using depth… | 47 | 1005 | active |
| facebookresearch/fairscale FairScale is a PyTorch extension library providing composable modules and APIs for high-performance, large-scale distributed training, incl… | 27 | 3403 | maintenance |
| meetps/pytorch-semseg A PyTorch library implementing popular semantic segmentation architectures such as FCN, U-Net, SegNet, PSPNet, ICNet, FRRN, and LinkNet, wi… | 23 | 3403 | maintenance |
| pjlab-sys4nlp/llama-moe LLaMA-MoE is a Python toolkit and model series for building Mixture-of-Experts (MoE) language models from dense LLaMA models via expert con… | 20 | 1004 | active |
| SciSharp/TensorFlow.NET TensorFlow.NET provides .NET Standard bindings for Google's TensorFlow, aiming to implement the complete TensorFlow API in C# and F#. It in… | 23 | 3395 | maintenance |
| siyuanchen0214/Scam-AI-Multi-modal-Evaluation-System A Python-based multi-modal AI system for detecting fraudulent content across text, image, audio, and video, with provenance tracing and cro… | 36 | 1002 | active |
| Liuhong99/Sophia Official PyTorch implementation of Sophia-G, a scalable stochastic second-order optimizer for language model pre-training, from the paper b… | 30 | 1001 | active |
| catalyst-team/catalyst Catalyst is a high-level PyTorch framework for deep learning research and development, focused on reproducibility, rapid experimentation, a… | 60 | 3386 | maintenance |
| Image Captioning A PyTorch research codebase for image captioning, serving as the widely used unofficial implementation of Self-critical Sequence Training (… | 23 | 999 | stable |
| PAIR-code/saliency A framework-agnostic Python library implementing state-of-the-art saliency methods for visualizing and explaining deep neural network predi… | 32 | 996 | stable |
| mit-han-lab/efficientvit A collection of efficient vision foundation models from MIT Han Lab, including EfficientViT backbones for perception, EfficientViT-SAM for … | 45 | 3361 | maintenance |
| shelhamer/fcn.berkeleyvision.org Reference implementation of Fully Convolutional Networks (FCN) for semantic segmentation from the CVPR 2015 / PAMI 2016 papers, built on Ca… | 32 | 3350 | maintenance |
| eladrich/pixel2style2pixel Official PyTorch implementation of pixel2style2pixel (pSp), a StyleGAN encoder from CVPR 2021 that maps real images directly into the W+ la… | 32 | 3349 | maintenance |
| tianzhi0549/FCOS Official PyTorch implementation of FCOS, a fully convolutional one-stage, anchor-free object detector published at ICCV 2019. It provides t… | 32 | 3344 | maintenance |
| tamarott/SinGAN Official PyTorch implementation of SinGAN, an ICCV 2019 best-paper generative model trained on a single natural image. It learns patch stat… | 32 | 3341 | maintenance |
| HRNet/HRNet-Semantic-Segmentation Official PyTorch implementation of HRNet (High-Resolution Network) and the Segmentation Transformer (OCR) approach for semantic segmentatio… | 32 | 3337 | maintenance |
| NVlabs/eg3d Official PyTorch implementation of EG3D, an efficient geometry-aware 3D generative adversarial network from NVIDIA Research (CVPR 2022). It… | 32 | 3336 | maintenance |
| run-youngjoo/SC-FEGAN SC-FEGAN is a GUI application that uses a generative adversarial network (SN-patchGAN discriminator with a U-Net generator) to edit face im… | 32 | 3328 | maintenance |
| JIA-Lab-research/MGM Official PyTorch implementation of Mini-Gemini, a multimodal vision-language model framework built on LLaVA that supports dense and MoE LLM… | 26 | 3322 | maintenance |
| pytorch-yolo-v3 A minimal PyTorch implementation of the YOLO v3 object detection algorithm, supporting detection on images and video with configurable reso… | 32 | 3314 | maintenance |
| PixArt-alpha/PixArt-alpha PixArt-α is a Transformer-based text-to-image diffusion model with PyTorch model definitions, pre-trained weights, and inference/training c… | 28 | 3306 | maintenance |
| open-mmlab/mmselfsup OpenMMLab's PyTorch-based toolbox and benchmark for self-supervised and unsupervised visual representation learning. It provides implementa… | 23 | 3303 | maintenance |
| NVIDIA/flownet2-pytorch A PyTorch implementation of FlowNet 2.0 for deep-learning-based optical flow estimation, released by NVIDIA. It provides multiple network a… | 62 | 3292 | maintenance |
| bytedance/lightseq LightSeq is a high-performance CUDA-based library for training and inference of sequence models like Transformer, BERT, GPT, and BART, with… | 10 | 3292 | maintenance |
| AlphaGenome Haiku is a simple neural network library for JAX from Google DeepMind, providing an object-oriented module abstraction (hk.Module) and a fu… | 86 | 3291 | maintenance |
| facebookresearch/vissl VISSL is Facebook AI Research's extensible, modular and scalable PyTorch library for state-of-the-art self-supervised learning with images.… | 10 | 3290 | maintenance |
| google-research/albert Official TensorFlow implementation and pretrained checkpoints of ALBERT, a lite version of BERT for self-supervised learning of language re… | 10 | 3278 | maintenance |
| zhanghang1989/ResNeSt ResNeSt is a PyTorch implementation of the Split-Attention Network, a ResNet variant that applies channel-wise attention across network bra… | 65 | 3259 | maintenance |
| martinarjovsky/WassersteinGAN Reference PyTorch implementation of the Wasserstein GAN paper, providing training scripts for DCGAN and MLP architectures on datasets like … | 32 | 3247 | maintenance |
| anandpawara/Real_Time_Image_Animation A real-time Python application that animates a still image (e.g., a portrait) using facial motion from a live camera or video file, built o… | 32 | 3246 | maintenance |
| google/model_search Model Search is a Google AutoML framework that implements neural architecture search algorithms at scale to find optimal DNN architectures … | 10 | 3235 | maintenance |
| mkocabas/VIBE Official PyTorch implementation of VIBE (CVPR 2020), a video-based method for 3D human body pose and shape estimation that predicts SMPL bo… | 23 | 3210 | maintenance |
| tomgoldstein/loss-landscape PyTorch code for visualizing the loss landscape of neural networks, from the NIPS 2018 paper 'Visualizing the Loss Landscape of Neural Nets… | 32 | 3205 | maintenance |
| tinyvision/DAMO-YOLO DAMO-YOLO is a fast and accurate object detection framework built on PyTorch, featuring NAS-searched backbones, RepGFPN, a lightweight Zero… | 32 | 3193 | maintenance |
| yusugomori/DeepLearning A multi-language (Python, C, C++, Java, Scala, Go) educational implementation of classic deep learning algorithms such as Deep Belief Nets,… | 32 | 3184 | maintenance |
| adambielski/siamese-triplet A PyTorch library implementing siamese and triplet networks for learning image embeddings, with online pair/triplet mining strategies. It p… | 32 | 3175 | maintenance |
| jettify/pytorch-optimizer torch-optimizer is a Python library providing a collection of additional optimization algorithms for PyTorch, drop-in compatible with the t… | 23 | 3172 | maintenance |
| farizrahman4u/seq2seq A sequence-to-sequence learning add-on library for Keras, providing modular encoder-decoder layers and ready-made Seq2Seq models. It suppor… | 32 | 3166 | maintenance |
| huawei-noah/Pretrained-Language-Model A collection of pretrained language models and optimization techniques from Huawei Noah's Ark Lab, including PanGu-α (200B-parameter Chines… | 32 | 3166 | maintenance |
| migueldeicaza/TensorFlowSharp TensorFlowSharp provides .NET bindings to the TensorFlow C API, exposing a strongly-typed low-level API for C# and F#. It is designed mainl… | 10 | 3147 | maintenance |
| microsoft/torchscale A PyTorch library from Microsoft implementing foundation Transformer architectures such as DeepNet, Magneto, RetNet, LongNet, BitNet, and X… | 32 | 3143 | maintenance |
| open-mmlab/mmdeploy MMDeploy is the OpenMMLab model deployment framework that converts PyTorch-based OpenMMLab models (mmdetection, mmsegmentation, etc.) into … | 23 | 3141 | maintenance |
| FudanNLP/fastNLP fastNLP is a lightweight, modularized and extensible NLP framework in Python that reduces engineering boilerplate such as data processing l… | 23 | 3137 | maintenance |
| open-mmlab/mmskeleton MMSkeleton is an OpenMMLAB toolbox for skeleton-based human understanding, built on PyTorch. It supports 2D pose estimation, skeleton-based… | 23 | 3131 | maintenance |
| google-deepmind/trfl TRFL is a Python library built on TensorFlow that provides building-block loss operations (e.g., Q-learning, TD learning, distributional RL… | 32 | 3130 | maintenance |
| diegoantognini/pyGAT A PyTorch implementation of the Graph Attention Network (GAT) model from Veličković et al. (2017), including a sparse-matrix variant. It re… | 32 | 3124 | maintenance |
| dbiir/UER-py UER-py is a PyTorch framework for pre-training transformer language models (BERT, GPT-2, T5, ELMo, etc.) and fine-tuning them on downstream… | 32 | 3114 | maintenance |
| lucasjinreal/yolov7_d2 A detectron2-based implementation of YOLOv7 that extends YOLO-style detection to instance segmentation, keypoint detection, and multi-head … | 23 | 3112 | maintenance |
| cysmith/neural-style-tf A TensorFlow implementation of neural style transfer based on Gatys et al.'s convolutional neural network approach, with support for video … | 32 | 3104 | maintenance |
| tusen-ai/simpledet SimpleDet is a Python framework built on MXNet for object detection and instance recognition. It provides state-of-the-art detection models… | 32 | 3085 | maintenance |
| Tramac/awesome-semantic-segmentation-pytorch A PyTorch library providing concise, modifiable reference implementations of many semantic segmentation models such as FCN, PSPNet, DeepLab… | 32 | 3070 | maintenance |
| cvlab-columbia/zero123 Zero-1-to-3 is a research codebase and pretrained diffusion model from Columbia CVLab that changes the camera viewpoint of an object from a… | 31 | 3067 | maintenance |
| stellargraph/stellargraph StellarGraph is a Python library for machine learning on graphs and networks, offering state-of-the-art graph neural network algorithms suc… | 23 | 3059 | maintenance |
| argman/EAST A TensorFlow re-implementation of the EAST (Efficient and Accurate Scene Text Detector) deep learning model for detecting text in natural s… | 32 | 3056 | maintenance |
| rinongal/textual_inversion Official implementation of the Textual Inversion paper, which learns new word embeddings in a frozen text-to-image (Latent Diffusion) model… | 32 | 3055 | maintenance |
| avinashpaliwal/Super-SloMo A PyTorch implementation of the Super SloMo paper for high-quality video frame interpolation, generating multiple intermediate frames to co… | 10 | 3022 | maintenance |
| microsoft/human-pose-estimation.pytorch Official PyTorch implementation of the ECCV 2018 paper 'Simple Baselines for Human Pose Estimation and Tracking' from Microsoft. It provide… | 10 | 3007 | maintenance |
| T5 T5X is a modular, composable framework built on JAX and Flax for high-performance training, evaluation, and inference of sequence models at… | 71 | 3002 | maintenance |
| divamgupta/image-segmentation-keras A Keras library implementing popular deep learning semantic image segmentation models including SegNet, FCN, U-Net, and PSPNet. It provides… | 23 | 3002 | maintenance |
| jfzhang95/pytorch-deeplab-xception A PyTorch implementation of the DeepLab v3+ semantic segmentation model with support for multiple backbones (Xception, ResNet, MobileNet, D… | 32 | 2998 | maintenance |
| keithito/tacotron An unofficial open-source TensorFlow implementation of Google's Tacotron end-to-end neural text-to-speech model, with a pre-trained model a… | 23 | 2998 | maintenance |
| ffcv FFCV is a fast data loading system for PyTorch that dramatically increases data throughput in model training by replacing standard data loa… | 23 | 2996 | maintenance |
| yangjianxin1/GPT2-chitchat A GPT2-based Chinese chitchat dialogue model project built on HuggingFace transformers, including training, preprocessing, and interactive … | 32 | 2995 | maintenance |