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 |
|---|---|---|---|
| Soul-AILab/SoulX-LiveAct SoulX-LiveAct is the official inference code for a real-time human animation framework that generates lifelike, audio/multimodal-controlled… | 51 | 1193 | active |
| GAIR-NLP/ASI-Arch A multi-agent framework that lets an LLM autonomously conduct end-to-end research on neural network architecture discovery, specifically li… | 40 | 1192 | active |
| warmshao/FasterLivePortrait A real-time portrait animation application based on LivePortrait that animates still photos or videos using a driving video, image, audio, … | 35 | 1191 | active |
| UniAnimate UniAnimate is the official code for a research paper on animating a reference human image into a video that follows a driving pose sequence… | 28 | 1190 | active |
| chakki-works/seqeval seqeval is a Python library for evaluating sequence labeling tasks such as named-entity recognition, part-of-speech tagging, and semantic r… | 23 | 1188 | stable |
| meta-pytorch/torchcodec TorchCodec is a PyTorch-native library for decoding and encoding videos, audio, and images into PyTorch tensors on CPU and CUDA GPU, built … | 87 | 1187 | active |
| SHI-Labs/Neighborhood-Attention-Transformer Official PyTorch implementation of the Neighborhood Attention Transformer (NAT/DiNAT), a family of efficient vision transformers with local… | 32 | 1186 | stable |
| msracver/Deformable-ConvNets Official MXNet implementation of Deformable Convolutional Networks (ICCV 2017) and R-FCN, including deformable convolution and ROI pooling … | 32 | 4120 | maintenance |
| csguoh/MambaIR MambaIR and MambaIRv2 are PyTorch-based image restoration models built on Mamba state-space models, published at ECCV 2024 and CVPR 2025. T… | 51 | 1182 | active |
| orpatashnik/StyleCLIP Official implementation of StyleCLIP, a method for text-driven manipulation of StyleGAN-generated imagery using CLIP. It provides three app… | 32 | 4117 | maintenance |
| Tencent-Hunyuan/MixGRPO MixGRPO is a research framework from Tencent Hunyuan implementing a mixed ODE-SDE GRPO algorithm for efficient reinforcement learning fine-… | 56 | 1181 | active |
| mlfoundations/open_flamingo OpenFlamingo is an open-source PyTorch implementation of DeepMind's Flamingo, a large multimodal vision-language model that interleaves ima… | 23 | 4114 | maintenance |
| balancap/SSD-Tensorflow A TensorFlow re-implementation of the Single Shot MultiBox Detector (SSD) for object detection, including VGG-based SSD-300 and SSD-512 net… | 32 | 4101 | maintenance |
| tjiiv-cprg/EPro-PnP EPro-PnP is a probabilistic Perspective-n-Points (PnP) layer for end-to-end 6DoF monocular object pose estimation networks, built on PyTorc… | 37 | 1177 | stable |
| magicleap/SuperGluePretrainedNetwork SuperGlue is a PyTorch implementation of a graph neural network with an optimal matching layer that matches sparse image features between t… | 32 | 4099 | maintenance |
| bytedance/1d-tokenizer A research repository from ByteDance containing code and pretrained model weights for 1D visual tokenizers (TiTok, TA-TiTok, FlowTok) and i… | 26 | 1176 | active |
| MIC-DKFZ/batchgenerators A Python framework for data augmentation of 2D and 3D images, developed by the German Cancer Research Center for medical image classificati… | 77 | 1175 | stable |
| iver56/torch-audiomentations A PyTorch library for fast audio data augmentation, inspired by audiomentations. It provides GPU-accelerated, differentiable audio transfor… | 54 | 1174 | active |
| princeton-nlp/MeZO MeZO is a memory-efficient zeroth-order optimizer that fine-tunes language models using only forward passes, with the same memory footprint… | 30 | 1174 | stable |
| wladradchenko/wunjo.wladradchenko.ru Wunjo CE is an open-source, locally-run AI media suite for face swap, lip sync, voice cloning, object/text/background removal, restyling, a… | 84 | 1173 | active |
| NVlabs/imaginaire NVIDIA's PyTorch library containing optimized implementations of image and video synthesis methods, including GAN-based image-to-image tran… | 32 | 4079 | maintenance |
| onnx/onnxmltools ONNXMLTools is a Python library that converts machine learning models from various toolkits (Keras/TensorFlow, scikit-learn, Core ML, XGBoo… | 77 | 1172 | active |
| LibCity LibCity is an open-source PyTorch library for urban spatial-temporal data mining, providing a unified pipeline for traffic prediction resea… | 23 | 1172 | active |
| geopavlakos/hamer HaMeR (Hand Mesh Recovery) is a transformer-based model that reconstructs 3D hand meshes from single monocular images using the MANO parame… | 53 | 1171 | active |
| GaParmar/clean-fid Clean-FID is a PyTorch library for computing the Frechet Inception Distance (FID) with correct image resizing and quantization steps, fixin… | 44 | 1171 | stable |
| SuperBruceJia/EEG-DL EEG-DL is a deep learning library built on TensorFlow for classifying EEG signals, supporting many architectures including CNNs, RNNs, GCNs… | 43 | 1171 | active |
| Tencent/hpc-ops HPC-Ops is a production-grade C++/CUDA operator library for high-performance LLM inference, developed by Tencent's Hunyuan AI Infra team. I… | 58 | 1170 | active |
| DepthAnything/PromptDA Prompt Depth Anything is a Python library implementing a CVPR 2025 method for high-resolution (up to 4K) accurate metric depth estimation. … | 47 | 1169 | active |
| warp-ctc A fast parallel implementation of the Connectionist Temporal Classification (CTC) loss function for CPU and CUDA GPU, with a simple C inter… | 32 | 4066 | maintenance |
| facebookresearch/encodec EnCodec is a deep learning based neural audio codec from Meta AI that compresses mono 24 kHz and stereo 48 kHz audio to bitrates from 1.5 t… | 32 | 4065 | maintenance |
| espressif/esp-dl ESP-DL is Espressif's lightweight neural network inference framework for ESP-series chips, with a custom .espdl model format, quantization … | 69 | 1168 | active |
| Dao-AILab/quack QuACK is a collection of high-performance GPU kernels (RMSNorm, LayerNorm, softmax, cross-entropy, GEMM with epilogues) written in NVIDIA's… | 83 | 1167 | active |
| extropic-ai/thrml THRML is a JAX library for building and sampling probabilistic graphical models, focused on efficient block Gibbs sampling of energy-based … | 70 | 1167 | active |
| Object Detection Metrics A Python toolkit implementing the most popular metrics (AP, mAP, precision-recall curves) used to evaluate object detection algorithms, wit… | 46 | 1167 | stable |
| deepmodeling/Uni-Mol Uni-Mol is a collection of 3D molecular representation learning frameworks and pretrained models for tasks like molecule property predictio… | 30 | 1167 | active |
| CyberAgentAILab/TANGO TANGO is a research library from CyberAgent AI Lab that generates co-speech gesture videos by reenactment, using hierarchical audio-motion … | 36 | 1166 | active |
| yosinski/deep-visualization-toolbox A GUI toolbox for visualizing and understanding deep neural networks, showing per-unit activations, backprop/deconv, and regularized-optimi… | 32 | 4052 | maintenance |
| sksq96/pytorch-summary A PyTorch library providing a Keras-style model.summary() that prints layer types, output shapes, parameter counts, and memory estimates. I… | 32 | 4050 | maintenance |
| cure-lab/MagicDrive MagicDrive is the official PyTorch implementation of an ICLR 2024 paper for controllable street view generation using diffusion models with… | 32 | 1165 | active |
| MCG-NKU/E2FGVI E2FGVI is the official PyTorch implementation of the CVPR 2022 paper 'Towards An End-to-End Framework for Flow-Guided Video Inpainting'. It… | 32 | 1165 | stable |
| facebookresearch/VideoPose3D A PyTorch implementation of CVPR 2019 research on 3D human pose estimation in video using temporal convolutions over 2D keypoint trajectori… | 10 | 4049 | maintenance |
| gemelo-ai/vocos Vocos is a fast neural vocoder that synthesizes audio waveforms from acoustic features such as mel-spectrograms or EnCodec tokens. It uses … | 62 | 1164 | stable |
| IST-DASLab/marlin Marlin is a highly optimized FP16xINT4 matrix multiplication CUDA kernel for LLM inference that achieves near-ideal 4x speedups at batch si… | 27 | 1164 | active |
| 3D ResNets for Action Recognition A PyTorch implementation of 3D ResNet and R(2+1)D models for video action recognition, accompanying CVPR 2018 and related papers. It includ… | 23 | 4038 | maintenance |
| Photo2Cartoon A Python deep-learning project from Minivision that converts real portrait photos into cartoon-style avatars using unpaired image translati… | 32 | 4032 | maintenance |
| fundamentalvision/Deformable-DETR Official PyTorch implementation of Deformable DETR, an efficient end-to-end object detector that uses deformable attention to fix DETR's sl… | 32 | 4028 | maintenance |
| chengtan9907/OpenSTL OpenSTL is a comprehensive benchmark and modular framework for spatio-temporal predictive learning, covering video prediction methods acros… | 50 | 1158 | active |
| Linaom1214/TensorRT-For-YOLO-Series A Python and C++ toolkit for running YOLO-series object detection models (YOLOv3 through YOLOv12, YOLOX) with NVIDIA TensorRT, including ON… | 40 | 1158 | active |
| sirius-ai/LPRNet_Pytorch A PyTorch implementation of LPRNet, a lightweight deep neural network for license plate recognition. It ships with pretrained weights focus… | 32 | 1158 | stable |
| JunMa11/SegLossOdyssey A curated collection of loss functions for medical image segmentation, accompanying the 'Loss Odyssey in Medical Image Segmentation' survey… | 32 | 4005 | maintenance |
| ShiqiYu/OpenGait OpenGait is a flexible and extensible Python framework for gait recognition research, providing implementations of state-of-the-art models … | 65 | 1154 | active |
| cvg/glue-factory Glue Factory is a PyTorch-based library for training and evaluating deep neural networks that detect and match local visual features (point… | 66 | 1153 | active |
| PKU-Alignment/omnisafe OmniSafe is a PyTorch-based infrastructural framework for safe reinforcement learning research, providing a unified modular toolkit and com… | 24 | 1153 | active |
| White-box Cartoonization Official TensorFlow implementation of the CVPR 2020 paper 'Learning to Cartoonize Using White-box Cartoon Representations', which converts … | 57 | 4000 | maintenance |
| quark0/darts DARTS is the official PyTorch implementation of the ICLR 2019 paper 'DARTS: Differentiable Architecture Search', which performs neural arch… | 32 | 3997 | maintenance |
| mgonzs13/yolo_ros A ROS 2 wrapper for Ultralytics YOLO models (YOLOv8 through YOLO26) providing object detection, tracking, instance segmentation, human pose… | 96 | 1152 | active |
| facebookresearch/fairseq2 fairseq2 is a PyTorch-based sequence modeling toolkit from Meta FAIR for training custom models for content generation tasks such as langua… | 86 | 1152 | active |
| open-gigaai/giga-world-1 GigaWorld-1 is an open-source framework providing training, inference, data processing, checkpoint conversion, and LoRA merge workflows for… | 52 | 1152 | active |
| JDAI-CV/fast-reid FastReID is a PyTorch-based research platform implementing state-of-the-art re-identification algorithms for persons, vehicles, and faces. … | 23 | 3995 | maintenance |
| MoonshotAI/MoonEP MoonEP is an Expert Parallelism communication library for Mixture-of-Experts training that keeps token loads perfectly balanced across rank… | 56 | 1146 | active |
| HengyiWang/spann3r Spann3R is a transformer-based model for dense 3D reconstruction from ordered or unordered image collections, built on the DUSt3R paradigm.… | 23 | 1145 | active |
| callous-youth/BOAT BOAT is a PyTorch-based library providing a compositional, operation-level toolbox for gradient-based bi-level optimization (BLO). It decom… | 81 | 1143 | active |
| noahcao/OC_SORT OC-SORT is a pure motion-model-based multi-object tracker for video, improving on SORT by fixing Kalman filter limitations to handle occlus… | 64 | 1143 | stable |
| rohitgandikota/sliders Official implementation of Concept Sliders, LoRA adaptors that enable precise, plug-and-play control of attributes in diffusion models like… | 49 | 1143 | active |
| LLM-Pruner LLM-Pruner is a PyTorch library implementing structural pruning of large language models based on gradient information, as published at Neu… | 30 | 1142 | active |
| clovaai/deep-text-recognition-benchmark Official PyTorch implementation of a four-stage scene text recognition (OCR) framework from an ICCV 2019 paper, with training and evaluatio… | 32 | 3941 | maintenance |
| andabi/deep-voice-conversion A TensorFlow implementation of deep neural networks for voice conversion (voice style transfer) that converts a source speaker's voice into… | 32 | 3941 | maintenance |
| rhymes-ai/Allegro Allegro is an open-source text-to-video generation model that produces high-quality 720p videos up to 6 seconds at 15 FPS from text prompts… | 23 | 1138 | active |
| flagos-ai/FlagGems FlagGems is a high-performance operator library for large language models written in the Triton language, providing backend-neutral GPU ker… | 88 | 1137 | active |
| OpenGVLab/SAM-Med2D Official implementation of SAM-Med2D, a fine-tuned Segment Anything Model (SAM) for 2D medical image segmentation, trained on the SA-Med2D-… | 29 | 1137 | active |
| alibaba-damo-academy/RynnVLA-002 RynnVLA-002 is a unified autoregressive Vision-Language-Action and world model that generates robot actions from text and image observation… | 40 | 1136 | active |
| InternRobotics/InternNav InternNav is an open-source PyTorch-based toolbox for building embodied navigation foundation models, supporting vision-language navigation… | 52 | 1135 | active |
| Janspiry/Image-Super-Resolution-via-Iterative-Refinement An unofficial PyTorch implementation of SR3 (Image Super-Resolution via Iterative Refinement), a diffusion-based model for image super-reso… | 32 | 3922 | maintenance |
| sooftware/conformer An unofficial PyTorch implementation of the Conformer architecture (convolution-augmented Transformer) from the INTERSPEECH 2020 paper, tar… | 65 | 1132 | active |
| localai-org/depth-anything.cpp A from-scratch C++17/ggml port of ByteDance's Depth Anything 2 and 3 models for dependency-free monocular metric depth and camera pose infe… | 56 | 1132 | active |
| OpenGVLab/VideoMamba VideoMamba is a state space model (Mamba-based) architecture for efficient video understanding, released with code and pretrained models fr… | 26 | 1130 | active |
| ikostrikov/pytorch-a2c-ppo-acktr-gail A PyTorch implementation of several deep reinforcement learning algorithms: A2C, PPO, ACKTR, and GAIL (imitation learning). It works with O… | 32 | 3907 | maintenance |
| MIT-SPARK/VGGT-SLAM VGGT-SLAM is a dense RGB SLAM system that performs real-time feed-forward 3D scene reconstruction, optimizing on the SL(4) manifold using t… | 56 | 1129 | active |
| PriesiaMioShirakana/DragonianVoice A C++ inference library for running ONNX-based TTS, SVC (singing voice conversion), and SVS (singing voice synthesis) models, supporting ar… | 36 | 1129 | active |
| Dobiasd/frugally-deep frugally-deep is a lightweight header-only C++ library for running inference (forward passes) on Keras/TensorFlow models without linking ag… | 79 | 1128 | active |
| AutoArk/open-audio-opd An industrial training stack for online policy distillation (OPD) of audio models, distilling compact ASR (and planned TTS) student models … | 49 | 1128 | active |
| nnsight nnsight is a Python library for interpreting and intervening on the internals of deep learning models, built on PyTorch. It lets researcher… | 96 | 1125 | active |
| xverse-engine/XScene-UEPlugin An Unreal Engine 5 plugin for real-time visualization, management, editing, and scalable hybrid rendering of 3D Gaussian Splatting models. … | 36 | 1125 | active |
| PRIME-RL/TTRL TTRL is an open-source implementation of Test-Time Reinforcement Learning, a method for training LLMs with RL on unlabeled data using major… | 49 | 1124 | active |
| THU-MIG/RepViT Official PyTorch implementation of RepViT, a family of lightweight CNNs designed by integrating efficient ViT architectural designs into Mo… | 20 | 1123 | stable |
| THUDM/SwissArmyTransformer SwissArmyTransformer (sat) is a PyTorch library for developing custom Transformer model variants where models like BERT, GPT, T5, GLM, and … | 23 | 1122 | active |
| hkchengrex/Cutie Cutie is a video object segmentation framework with object-level memory reading, a follow-up to XMem offering better consistency, robustnes… | 18 | 1122 | active |
| AdaptiveMotorControlLab/CEBRA CEBRA is a Python library for self-supervised learning of consistent latent embeddings from high-dimensional time-series recordings, using … | 87 | 1121 | active |
| MST (Spectral Compressive Imaging Toolbox) A Python toolbox for spectral compressive imaging reconstruction that implements over 15 algorithms including MST, CST, DAUHST, BiSCI, HDNe… | 49 | 1121 | active |
| FlagAI-Open/FlagAI FlagAI is a Python toolkit for training, fine-tuning, and deploying large-scale AI models across NLP, CV, and vision-language tasks. It int… | 61 | 3870 | maintenance |
| PaddlePaddle/PaddleHelix PaddleHelix is a bio-computing platform built on PaddlePaddle featuring large-scale representation learning and multi-task deep learning fo… | 52 | 1120 | active |
| yangxue0827/RotationDetection AlphaRotate is a TensorFlow-based benchmark and toolbox for rotated (oriented) object detection, implementing detectors such as R2CNN, Reti… | 23 | 1120 | active |
| HITsz-TMG/Uni-MoE Uni-MoE is a family of open-source Mixture-of-Experts (MoE) based omnimodal large language models that understand and generate across text,… | 65 | 1118 | active |
| SCLBD/DeepfakeBench DeepfakeBench is a comprehensive benchmark framework for deepfake detection, providing a unified platform for data management, implementati… | 33 | 1118 | active |
| luchris429/purejaxrl PureJaxRL provides end-to-end reinforcement learning training pipelines implemented entirely in JAX, including environments, enabling massi… | 31 | 1118 | active |
| Lasagne Lasagne is a lightweight Python library for building and training neural networks on top of Theano. It supports feed-forward, convolutional… | 23 | 3858 | maintenance |
| OpenMOSS/MOVA MOVA is an open-source foundation model and toolkit for joint video-audio generation, synthesizing synchronized video and audio in a single… | 59 | 1117 | active |
| MatchZoo MatchZoo is a Python toolkit for designing, comparing, and sharing deep text matching models. It provides a unified data pipeline, pre-buil… | 23 | 3849 | maintenance |
| openai/improved-diffusion The official codebase for OpenAI's Improved Denoising Diffusion Probabilistic Models paper, providing a Python package for training and sam… | 32 | 3848 | maintenance |
| yahoo/TensorFlowOnSpark TensorFlowOnSpark is a Python library that lets existing TensorFlow programs run distributed training and inference on Apache Spark and Had… | 23 | 3847 | maintenance |