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 |
|---|---|---|---|
| yenchenlin/nerf-pytorch A PyTorch implementation of NeRF (Neural Radiance Fields) that reproduces the original paper's results for synthesizing novel views of comp… | 32 | 6052 | maintenance |
| CoinCheung/BiSeNet A PyTorch implementation of the BiSeNet V1 and V2 real-time semantic segmentation models, with pretrained weights for Cityscapes, COCO-Stuf… | 67 | 1639 | active |
| Higgsfield Higgsfield is an open-source GPU orchestration and machine learning framework for fault-tolerant, distributed training of very large models… | 67 | 6047 | maintenance |
| luanfujun/deep-painterly-harmonization Research code implementing the 'Deep Painterly Harmonization' algorithm, which seamlessly blends a pasted object into a painting's style us… | 32 | 6033 | maintenance |
| yoshitomo-matsubara/torchdistill torchdistill is a modular, configuration-driven PyTorch framework for knowledge distillation and general deep learning experiments, requiri… | 83 | 1633 | active |
| jeffffffli/HybrIK HybrIK is the official PyTorch implementation of a hybrid analytical-neural inverse kinematics method for 3D human pose and shape estimatio… | 23 | 1632 | stable |
| Fantasy-AMAP/fantasy-talking FantasyTalking is a research codebase and model for generating realistic talking portrait videos from a single image and an audio clip, bui… | 45 | 1631 | active |
| ZhangGe6/onnx-modifier onnx-modifier is a web-based visual editor for ONNX neural network models, built on Netron for graph visualization and Flask as the backend… | 69 | 1629 | active |
| SyncTalk SyncTalk is the official PyTorch implementation of a CVPR 2024 paper that synthesizes speech-driven, synchronized talking head videos using… | 43 | 1628 | active |
| Picsart-AI-Research/StreamingT2V StreamingT2V is a research codebase (CVPR 2025) implementing an autoregressive technique that turns short text-to-video diffusion models li… | 28 | 1628 | active |
| pq-yang/MatAnyone MatAnyone is a CVPR 2025 human video matting framework that extracts alpha mattes of target people from video using consistent memory propa… | 51 | 1625 | active |
| PaddlePaddle/PaddleSlim PaddleSlim is an open-source library built on PaddlePaddle for deep learning model compression and architecture search. It provides low-bit… | 46 | 1612 | active |
| gorgonia/gorgonia Gorgonia is a Go library for machine learning that lets you define and evaluate mathematical equations over multidimensional arrays using a… | 23 | 5927 | maintenance |
| Chainer Chainer is a Python-based deep learning framework that pioneered the define-by-run approach with dynamic computational graphs and automatic… | 23 | 5919 | maintenance |
| dmlc/gluon-cv GluonCV is a deep learning toolkit providing state-of-the-art computer vision model implementations with 170+ pre-trained models. It suppor… | 23 | 5914 | maintenance |
| OpenGVLab/LLaMA-Adapter Official implementation of LLaMA-Adapter and its V2 successor, a parameter-efficient fine-tuning method that adapts LLaMA models to follow … | 21 | 5913 | maintenance |
| MoonshotAI/Kimi-Linear Kimi Linear is a hybrid linear attention architecture (Kimi Delta Attention, based on Gated DeltaNet) released by Moonshot AI with 48B-para… | 37 | 1607 | active |
| autonomousvision/transfuser Official PyTorch implementation of TransFuser, a transformer-based multi-modal sensor fusion model for end-to-end autonomous driving, publi… | 75 | 1606 | stable |
| AlibabaResearch/DAMO-ConvAI The official codebase for Alibaba DAMO Academy's Conversational AI research, containing implementations of models from their published pape… | 75 | 1605 | active |
| Tencent-Hunyuan/HY-WorldPlay HY-WorldPlay (HY-World 1.5) is Tencent Hunyuan's open-source framework for interactive 3D world modeling, generating explorable 3D scenes f… | 52 | 1604 | active |
| ml4a/ml4a ml4a is a Python library and collection of Jupyter notebooks for making art with machine learning. It wraps popular deep learning models li… | 32 | 1604 | active |
| meituan/YOLOv6 YOLOv6 is a single-stage object detection framework implemented in PyTorch, designed for industrial applications with a family of pretraine… | 23 | 5892 | maintenance |
| Drexubery/ViewCrafter ViewCrafter is a research codebase that uses video diffusion models to synthesize high-fidelity novel views of scenes from a single or spar… | 45 | 1602 | active |
| google/uncertainty-baselines A library of high-quality, minimal-dependency implementations of standard and state-of-the-art uncertainty and robustness methods for deep … | 77 | 1601 | active |
| vibevoice-community/VibeVoice VibeVoice is a community-maintained fork of Microsoft's long-form conversational text-to-speech model, generating expressive multi-speaker … | 59 | 1601 | active |
| xLLM-AI/xllm xLLM is a high-performance C++ inference engine for LLM, VLM, DiT and recommendation models, optimized for heterogeneous AI accelerators su… | 80 | 1600 | active |
| SalesforceAIResearch/uni2ts Uni2TS is a PyTorch library for unified pre-training, fine-tuning, inference, and evaluation of universal time series forecasting transform… | 54 | 1599 | active |
| alibaba/Pai-Megatron-Patch Pai-Megatron-Patch is an open-source deep learning training toolkit from Alibaba Cloud for large-scale training and inference of LLMs and V… | 50 | 1599 | active |
| Tencent-Hunyuan/HunyuanWorld-Voyager HunyuanWorld-Voyager is a video diffusion framework from Tencent Hunyuan that generates world-consistent RGBD video and 3D point-cloud sequ… | 49 | 1599 | active |
| XPixelGroup/HAT HAT (Hybrid Attention Transformer) is a PyTorch implementation of a state-of-the-art transformer model for image super-resolution and resto… | 32 | 1598 | stable |
| X-LANCE/AniTalker AniTalker is the official PyTorch implementation of an ACM MM 2024 paper that animates a single static portrait into a vivid talking-face v… | 25 | 1598 | active |
| SonyResearch/micro_diffusion Official implementation of Sony Research's micro-budget approach to training large-scale text-to-image diffusion transformer models from sc… | 24 | 1598 | active |
| databricks/megablocks MegaBlocks is a lightweight Python library for efficient training of mixture-of-experts (MoE) models, built around its dropless-MoE (dMoE) … | 58 | 1596 | active |
| pytorch/FBGEMM FBGEMM is a collection of highly optimized low-precision matrix multiplication and convolution kernels for server-side deep learning infere… | 91 | 1595 | active |
| GHOST GHOST (Generative High-fidelity One Shot Transfer) is a one-shot face swap pipeline for images and videos, published as an IEEE paper and i… | 23 | 1593 | active |
| facebookresearch/fast3r Fast3R is the official PyTorch implementation of a CVPR 2025 model from Meta FAIR that reconstructs 3D scenes and estimates camera poses fr… | 10 | 1592 | active |
| FoundationVision/Infinity Infinity is a bitwise autoregressive text-to-image generation model (CVPR 2025 Oral) with released training and inference code, checkpoints… | 53 | 1589 | active |
| DeepGraphLearning/torchdrug TorchDrug is a PyTorch-based machine learning library for drug discovery, covering graph neural networks, deep generative models, and reinf… | 23 | 1589 | active |
| meta-pytorch/torchtune Torchtune is a PyTorch-native library for authoring, post-training, and experimenting with large language models. It provides hackable trai… | 67 | 5819 | maintenance |
| RWKV/rwkv.cpp A C++ port of the RWKV language model to the ggml tensor library, providing FP32, FP16, and quantized INT4/INT5/INT8 inference focused on C… | 34 | 1583 | active |
| Xilinx/brevitas Brevitas is a PyTorch library for neural network quantization supporting both post-training quantization (PTQ) and quantization-aware train… | 93 | 1582 | active |
| microsoft/MMdnn MMdnn is a Microsoft toolkit for converting, visualizing, and diagnosing deep learning models across frameworks such as TensorFlow, PyTorch… | 35 | 5803 | maintenance |
| google-ai-edge/model-explorer Model Explorer is a model graph visualizer and debugger from Google AI Edge that renders neural network graphs hierarchically with expandab… | 92 | 1581 | active |
| DiffEqML/torchdyn Torchdyn is a PyTorch library dedicated to numerical deep learning, providing tools for neural differential equations, implicit models, and… | 23 | 1580 | active |
| tensorflow/model-optimization The TensorFlow Model Optimization Toolkit (tfmot) is a Python library providing tools to optimize machine learning models for deployment, i… | 79 | 1579 | stable |
| menyifang/MIMO MIMO is the official PyTorch implementation of a CVPR 2025 paper on controllable character video synthesis using spatially decomposed model… | 32 | 1578 | active |
| BloodAxe/pytorch-toolbelt A Python library of PyTorch extensions providing building blocks for fast R&D prototyping, including encoder-decoder architectures, special… | 40 | 1575 | active |
| Flux Zygote.jl is a source-to-source automatic differentiation library for Julia that hooks into the Julia compiler to generate gradient (backwa… | 97 | 1574 | active |
| ZFTurbo/Music-Source-Separation-Training A Python training framework for music source separation models, supporting many architectures such as MDX23C, Demucs, Band Split RoFormer, … | 87 | 1569 | active |
| Chinese-Text-Classification-Pytorch A PyTorch-based collection of ready-to-run Chinese text classification models including TextCNN, TextRNN, FastText, TextRCNN, BiLSTM with a… | 32 | 5728 | maintenance |
| mosaicml/streaming StreamingDataset is a Python library from MosaicML for fast, accurate streaming of training data from cloud object storage (S3, GCS, Azure,… | 66 | 1560 | active |
| ByteDance-Seed/Triton-distributed Triton-distributed is a distributed compiler built on OpenAI Triton for computation-communication overlapping on multi-GPU systems. It lets… | 61 | 1559 | active |
| hustvl/MapTR MapTR is an end-to-end transformer-based framework for online vectorized HD map construction from camera imagery in autonomous driving. It … | 23 | 1558 | active |
| decoderesearch/SAELens SAELens is a Python library for training sparse autoencoders (SAEs) on language model activations and analyzing them for mechanistic interp… | 90 | 1555 | active |
| ZeyueT/AudioX AudioX is a unified multimodal framework for anything-to-audio generation, producing audio and music conditioned on text, video, image, or … | 49 | 1552 | active |
| MST (Spectral Compressive Imaging Toolbox) Retinexformer is a one-stage Retinex-based Transformer model and toolbox for low-light image enhancement, published at ICCV 2023. It suppor… | 63 | 1548 | active |
| hustvl/LightningDiT LightningDiT is a research codebase for latent diffusion models implementing VA-VAE and LightningDiT, achieving FID 1.35 on ImageNet-256 wi… | 44 | 1547 | active |
| NirAharon/BoT-SORT BoT-SORT is a state-of-the-art multi-object tracker that combines motion and appearance information with camera motion compensation and an … | 32 | 1547 | active |
| tensorflow/gnn TensorFlow GNN is a Python library for building Graph Neural Networks on TensorFlow, including a GraphTensor type for heterogeneous graphs,… | 66 | 1546 | active |
| kijai/ComfyUI-CogVideoXWrapper A ComfyUI custom node wrapper for CogVideoX and related video generation models (including Fun variants, CogVideoX 1.5, and Go-with-the-Flo… | 36 | 1546 | active |
| MoonshotAI/Moonlight Moonlight is a 3B/16B Mixture-of-Experts LLM trained with the Muon optimizer, released by Moonshot AI along with a memory- and communicatio… | 33 | 1546 | active |
| lucidrains/soundstorm-pytorch A PyTorch implementation of SoundStorm, Google DeepMind's efficient parallel audio generation model that applies MaskGiT-style masked gener… | 32 | 1545 | active |
| facebookresearch/mmf MMF is a modular PyTorch framework for vision and language multimodal research from Facebook AI Research. It ships reference implementation… | 67 | 5637 | maintenance |
| ROMP ROMP is a PyTorch-based library and pip-installable API (simple-romp) for real-time monocular multi-person 3D human mesh recovery, implemen… | 23 | 1543 | stable |
| lucidrains/DALLE-pytorch A PyTorch implementation/replication of OpenAI's DALL-E, a text-to-image transformer, including a discrete VAE and optional CLIP for rankin… | 23 | 5627 | maintenance |
| JingyunLiang/SwinIR Official PyTorch implementation of SwinIR, a Swin Transformer-based model for image restoration tasks including super-resolution, denoising… | 23 | 5611 | maintenance |
| allenzren/open-pi-zero An open-source re-implementation of the pi0 vision-language-action (VLA) model from Physical Intelligence, built on a pre-trained PaliGemma… | 24 | 1533 | active |
| Leela Zero Leela Zero is an open-source Go engine that reimplements AlphaGo Zero, combining Monte Carlo Tree Search with a deep residual convolutional… | 23 | 5591 | maintenance |
| WenmuZhou/PytorchOCR A PyTorch-based OCR toolkit that ports PaddleOCR models to PyTorch, supporting common text detection and recognition algorithms like the PP… | 56 | 1527 | active |
| cchen156/Learning-to-See-in-the-Dark TensorFlow implementation of 'Learning to See in the Dark' (CVPR 2018), a deep learning model that brightens very dark, short-exposure RAW … | 48 | 5567 | maintenance |
| NVIDIA/DLSS NVIDIA DLSS is the public SDK repository for NVIDIA's RTX Deep Learning Super Sampling, a neural network that boosts game frame rates and g… | 91 | 1526 | active |
| Soul-AILab/SoulX-FlashTalk SoulX-FlashTalk is a 14B audio-driven talking avatar model that streams infinite real-time video from a reference image and audio, achievin… | 56 | 1526 | active |
| keras-rl/keras-rl A Python library implementing state-of-the-art deep reinforcement learning algorithms (DQN, DDPG, SARSA, and more) that integrates seamless… | 23 | 5549 | maintenance |
| Phantom-video/Phantom Phantom is a subject-consistent video generation model from ByteDance that preserves reference subject identity via cross-modal alignment. … | 36 | 1517 | active |
| mattmireles/gemma-tuner-multimodal A Python tool for LoRA fine-tuning of Gemma 4 and 3n models on text, images, and audio using Apple Silicon's Metal Performance Shaders. It … | 60 | 1516 | active |
| hustvl/DiffusionDrive DiffusionDrive is a truncated diffusion model for real-time end-to-end autonomous driving, released as the official PyTorch implementation … | 41 | 1513 | active |
| k2-fsa/icefall Icefall is a collection of speech recognition (ASR) and TTS training recipes built on the k2 and lhotse libraries, implemented in Python wi… | 67 | 1512 | active |
| imoneoi/openchat OpenChat is a library of open-source large language models fine-tuned with C-RLFT, an offline reinforcement learning strategy that learns f… | 20 | 5485 | maintenance |
| charlesq34/pointnet Reference implementation of PointNet, a neural network architecture that directly consumes unordered 3D point clouds for classification and… | 32 | 5482 | maintenance |
| wb14123/seq2seq-couplet A deep learning project that generates Chinese couplets (对联) using a seq2seq model built with TensorFlow. It includes training scripts, a w… | 32 | 5480 | maintenance |
| czczup/ViT-Adapter Official PyTorch implementation of ViT-Adapter, an ICLR 2023 Spotlight paper introducing a pre-training-free adapter that lets plain Vision… | 30 | 1506 | stable |
| graphdeco-inria/diff-gaussian-rasterization A CUDA-based differentiable rasterization engine for 3D Gaussian Splatting, used in the SIGGRAPH 2023 paper '3D Gaussian Splatting for Real… | 29 | 1506 | stable |
| Flashlight wav2letter++ is Facebook AI Research's end-to-end automatic speech recognition (ASR) toolkit written in C++. It has been consolidated into … | 67 | 5476 | maintenance |
| TencentARC/MotionCtrl MotionCtrl is the official implementation of a SIGGRAPH 2024 paper providing a unified and flexible motion controller for video generation … | 27 | 1500 | active |
| CUT3R/CUT3R CUT3R is the official PyTorch implementation of 'Continuous 3D Perception Model with Persistent State' (CVPR 2025 Oral), a stateful recurre… | 35 | 1499 | active |
| open-mmlab/mmengine MMEngine is the foundational training engine library for OpenMMLab projects, providing a unified training loop, config system, registry, ho… | 68 | 1496 | active |
| NVlabs/Fast-FoundationStereo Fast-FoundationStereo is NVIDIA's official PyTorch implementation of a real-time zero-shot stereo matching model family, accepted to CVPR 2… | 59 | 1496 | active |
| ucb-bar/gemmini Gemmini is Berkeley's open-source generator for parameterizable systolic-array DNN hardware accelerators, written in Chisel (Scala) and int… | 66 | 1493 | active |
| ibab/tensorflow-wavenet A TensorFlow implementation of DeepMind's WaveNet generative neural network architecture for raw audio waveform generation. It provides tra… | 32 | 5422 | maintenance |
| bert4keras A lightweight, clean reimplementation of BERT and other transformer models (RoBERTa, ALBERT, T5, GPT, ELECTRA, NEZHA) for Keras/tf.keras. I… | 23 | 5409 | maintenance |
| google-deepmind/graph_nets DeepMind's library for building graph networks (graph neural networks) in TensorFlow and Sonnet, based on the 'Relational inductive biases,… | 32 | 5407 | maintenance |
| alxndrTL/mamba.py A simple, readable pure-PyTorch (plus MLX) implementation of the Mamba state-space model architecture with a parallel scan for efficient tr… | 50 | 1480 | active |
| autonomousvision/mip-splatting Mip-Splatting is a research implementation of alias-free 3D Gaussian Splatting, introducing a 3D smoothing filter and 2D Mip filter to elim… | 27 | 1480 | active |
| sczhou/Upscale-A-Video Upscale-A-Video is a diffusion-based model for real-world video super-resolution that takes low-resolution videos and text prompts as input… | 27 | 1477 | active |
| Lightning-AI/lightning-thunder Lightning Thunder is a source-to-source deep learning compiler for PyTorch that optimizes models for training and inference. It provides a … | 70 | 1475 | active |
| Om-Alve/smolGPT A minimal pure-PyTorch implementation for training small GPT-style LLMs from scratch, featuring flash attention, RMSNorm, SwiGLU, RoPE, and… | 21 | 1474 | active |
| hojonathanho/diffusion The official reference implementation of Denoising Diffusion Probabilistic Models (DDPM) from the 2020 paper by Jonathan Ho et al., written… | 32 | 5341 | maintenance |
| mit-han-lab/torchsparse TorchSparse is a high-performance PyTorch library for sparse convolution on 3D point clouds, with optimized GPU kernels for both training a… | 23 | 1473 | active |
| FeiYull/TensorRT-Alpha A C++/CUDA library providing TensorRT-accelerated deployment for 30+ popular computer vision models including YOLOv3-v8, YOLOv8-Pose/Seg/Cl… | 100 | 1464 | active |
| tianweiy/DMD2 DMD2 is the official PyTorch implementation of Improved Distribution Matching Distillation, a NeurIPS 2024 method that distills diffusion m… | 25 | 1464 | active |