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 |
|---|---|---|---|
| pykeio/ort ort is a safe, fast Rust binding for Microsoft's ONNX Runtime, providing hardware-accelerated machine learning inference and training for O… | 91 | 2569 | active |
| LTH14/JiT A PyTorch/GPU re-implementation of JiT (Just image Transformer), a minimalist pixel-space diffusion model for high-resolution image generat… | 39 | 2553 | active |
| DeepDetect DeepDetect is an open-source deep learning runtime, CLI, and REST server written in C++ for training and inference across images, text, tab… | 95 | 2551 | active |
| mozilla/TTS A deep learning library for advanced text-to-speech generation, built on PyTorch with models like Tacotron2, Glow-TTS, and various vocoders… | 23 | 10173 | maintenance |
| data-infra/cube-studio CubeStudio is an open-source, cloud-native, all-in-one AI platform covering the full machine learning lifecycle (MLOps/MaaS/LLMOps), includ… | 75 | 2546 | active |
| yifan123/flow_grpo Flow-GRPO is the official PyTorch implementation of a NeurIPS 2025 paper that trains flow matching models (e.g., SD3.5, FLUX.1, Qwen-Image,… | 52 | 2543 | active |
| SafeAILab/EAGLE EAGLE is the official implementation of the EAGLE family of speculative decoding algorithms (EAGLE-1/2/3) for accelerating LLM text generat… | 54 | 2542 | active |
| learning-at-home/hivemind Hivemind is a PyTorch library for decentralized deep learning across the Internet, enabling training of large models on hundreds of volunte… | 53 | 2541 | active |
| LyCORIS LyCORIS is a Python library implementing parameter-efficient fine-tuning algorithms (LoRA/LoCon, LoHa, LoKr, IA3, DyLoRA, and more) for Sta… | 99 | 2530 | active |
| rpautrat/SuperPoint A TensorFlow (with PyTorch conversion) implementation of the SuperPoint self-supervised interest point detector and descriptor network. It … | 38 | 2528 | stable |
| bilibili/Index-anisora Index-AniSora is Bilibili's open-source anime video generation model, capable of creating video shots in diverse anime styles from images, … | 60 | 2523 | active |
| luanfujun/deep-photo-styletransfer Reference implementation of the CVPR 2017 paper 'Deep Photo Style Transfer', performing photorealistic image style transfer using Torch wit… | 32 | 9982 | maintenance |
| google-deepmind/sonnet Sonnet is a Python library built on TensorFlow 2 that provides simple, composable abstractions (snt.Module) for constructing neural network… | 66 | 9973 | maintenance |
| sthalles/SimCLR A PyTorch reference implementation of SimCLR, a self-supervised contrastive learning framework for learning visual representations from unl… | 23 | 2494 | stable |
| google/tunix Tunix is a lightweight JAX-based library for post-training large language models, supporting supervised fine-tuning, preference optimizatio… | 81 | 2490 | active |
| GaParmar/img2img-turbo A research library implementing one-step image-to-image translation models (CycleGAN-Turbo and pix2pix-turbo) built on SD-Turbo diffusion m… | 38 | 2488 | active |
| ppogg/YOLOv5-Lite YOLOv5-Lite is a lightweight object detection model family evolved from YOLOv5, with models as small as ~900KB (int8) that run 10-15+ FPS o… | 23 | 2488 | active |
| xuebinqin/U-2-Net Official PyTorch implementation of U^2-Net, a nested U-structure deep network for salient object detection, published in Pattern Recognitio… | 32 | 9875 | maintenance |
| thuml/Autoformer Autoformer is the official PyTorch implementation of the NeurIPS 2021 paper 'Autoformer: Decomposition Transformers with Auto-Correlation f… | 32 | 2485 | stable |
| google/XNNPACK XNNPACK is a highly optimized C library of floating-point neural network inference operators for ARM, x86, WebAssembly, and RISC-V platform… | 77 | 2481 | active |
| alexjc/neural-doodle A Python implementation of Semantic Style Transfer (Champandard, 2016) based on the Neural Patches algorithm. It turns rough doodles into r… | 10 | 9851 | maintenance |
| Lightning-AI/torchmetrics TorchMetrics is a PyTorch-native library of machine learning metrics that work in distributed and scalable training pipelines. It provides … | 87 | 2477 | stable |
| roboflow/inference Roboflow Inference is a Python library and self-hostable inference server for deploying computer vision models on any computer or edge devi… | 91 | 2474 | active |
| jadore801120/attention-is-all-you-need-pytorch A PyTorch implementation of the Transformer model from the paper 'Attention Is All You Need', including training and translation scripts fo… | 32 | 9816 | maintenance |
| Alibaba-Quark/LiveAvatar LiveAvatar is an open-source implementation of an ECCV 2026 paper for streaming, real-time, infinite-length audio-driven avatar video gener… | 58 | 2456 | active |
| facebookresearch/pifuhd PIFuHD is a PyTorch implementation of a CVPR 2020 research model that reconstructs high-resolution 3D human body meshes from a single 2D im… | 10 | 9724 | maintenance |
| TorchIO-project/torchio TorchIO is a Python library for loading, augmenting, and processing 3D medical images (MRI, CT) within PyTorch deep learning pipelines. It … | 91 | 2446 | active |
| AI-Hypercomputer/maxtext MaxText is a high-performance, scalable open-source LLM training library written in pure Python/JAX, targeting Google Cloud TPUs and GPUs. … | 98 | 2444 | active |
| shenweichen/DeepMatch DeepMatch is a Python library of deep matching models for recommendations and advertising, built on TensorFlow/Keras. It lets users train m… | 66 | 2435 | active |
| resemble-ai/resemble-enhance Resemble Enhance is an AI-powered Python tool that improves speech quality through denoising and enhancement, using a denoiser module and a… | 18 | 2434 | active |
| wolny/pytorch-3dunet A PyTorch implementation of 3D U-Net and its variants (residual, squeeze-and-excitation) for volumetric semantic segmentation, with 2D U-Ne… | 57 | 2423 | active |
| HuCaoFighting/Swin-Unet Official PyTorch implementation of Swin-Unet, a U-shaped pure Transformer model for medical image segmentation, published at ECCV 2022 Medi… | 38 | 2420 | stable |
| tflearn/tflearn TFLearn is a modular deep learning library providing a higher-level, Keras-like API on top of TensorFlow for building and training neural n… | 23 | 9570 | maintenance |
| Olive Olive is Microsoft's AI model optimization toolkit for the ONNX Runtime, automating finetuning, conversion, quantization, and compression o… | 89 | 2399 | active |
| ailia models A collection of 400+ pre-trained state-of-the-art AI models (object detection, pose estimation, speech recognition, LLMs, image generation,… | 76 | 2393 | active |
| OpenGVLab/InternVideo InternVideo is a series of open-source video foundation models for multimodal video understanding, spanning generative and discriminative l… | 69 | 2393 | active |
| google/neural-tangents Neural Tangents is a Python library built on JAX for defining, training, and evaluating neural networks of both finite and infinite width. … | 10 | 2385 | stable |
| facebookresearch/perception_models Meta's Perception Models repository hosting state-of-the-art image, video, and audio encoders (Perception Encoder, PE) and a multimodal lan… | 51 | 2382 | active |
| jik876/hifi-gan The official PyTorch implementation of HiFi-GAN, a generative adversarial network that converts mel-spectrograms into high-fidelity 22.05 k… | 32 | 2377 | stable |
| apple/axlearn AXLearn is a Python deep learning library built on JAX and XLA for developing and training large-scale models, with an object-oriented conf… | 69 | 2375 | active |
| XiaomiMiMo/MiMo Xiaomi's MiMo is a 7B-parameter reasoning language model trained from pretraining through posttraining with reinforcement learning, release… | 28 | 2373 | active |
| FMInference/FlexLLMGen FlexLLMGen is a high-throughput generation engine for running large language models on a single GPU with limited memory. It uses IO-efficie… | 10 | 9346 | maintenance |
| tencent-ailab/V-Express V-Express is a Python research project from Tencent AI Lab that generates talking head portrait videos from a reference image, audio, and V… | 25 | 2357 | active |
| Neural Circuit Policies (ncps) A Python package providing PyTorch and TensorFlow/Keras implementations of Neural Circuit Policies (NCPs), including liquid time-constant (… | 23 | 2351 | active |
| google-deepmind/optax Optax is a gradient processing and optimization library for JAX, offering composable building blocks like optimizers and loss functions. It… | 80 | 2350 | stable |
| SkyworkAI/Matrix-Game Matrix-Game is Skywork AI's open-source series of interactive world foundation models that generate real-time, streaming video in response … | 50 | 2342 | active |
| facebookresearch/schedule_free A PyTorch library implementing schedule-free optimizers (SGD, AdamW, RAdam variants) that remove the need for learning rate schedules or sp… | 64 | 2329 | active |
| LabML A curated collection of over 60 PyTorch implementations of deep learning papers, complete with side‑by‑side explanatory notes. It serves as… | 26 | 2327 | active |
| LabML A Python library for tracking and monitoring deep learning experiments, with a self-hostable server app for viewing metrics and hardware us… | 26 | 2327 | active |
| ufoym/imbalanced-dataset-sampler A PyTorch dataset sampler (ImbalancedDatasetSampler, pip package torchsampler) that rebalances class distributions during training by overs… | 60 | 2326 | stable |
| Picotron Picotron is a minimalist, hackable distributed training framework for pre-training Llama-like large language models using 4D parallelism (d… | 37 | 2325 | active |
| PKU-YuanGroup/MoE-LLaVA MoE-LLaVA is an open-source Mixture-of-Experts based sparse large vision-language model, released with the MoE-Tuning training strategy fro… | 28 | 2324 | active |
| iver56/audiomentations Audiomentations is a Python library for audio data augmentation with an API inspired by albumentations. It provides fast CPU-based waveform… | 61 | 2323 | stable |
| facebookresearch/ImageBind A PyTorch library from Meta AI implementing ImageBind, a model that learns a joint embedding space across six modalities: images, text, aud… | 51 | 9076 | maintenance |
| LAION-AI/CLAP CLAP (Contrastive Language-Audio Pretraining) is a Python library from LAION that learns joint audio and text embeddings using contrastive … | 38 | 2312 | stable |
| IDEA-Research/detrex detrex is an open-source PyTorch-based research platform and toolbox for DETR-style Transformer detection algorithms, built on top of Detec… | 38 | 2311 | active |
| frgfm/torch-cam TorchCAM is a Python library that extracts class activation maps (CAMs) from PyTorch CNN classifiers, supporting many CAM variants such as … | 88 | 2307 | active |
| traveller59/spconv SpConv is a spatially sparse convolution library for deep learning on 3D point clouds and sparse tensors, distributed as PyPI packages with… | 32 | 2299 | active |
| THU-MIG/yoloe YOLOE is the official PyTorch implementation of an open-vocabulary object detection and segmentation model presented at ICCV 2025. It unifi… | 30 | 2297 | active |
| aigc-apps/VideoX-Fun VideoX-Fun is a Python-based video generation pipeline built on Diffusion Transformer models (CogVideoX-Fun, Wan-Fun) that generates videos… | 68 | 2292 | active |
| hkchengrex/MMAudio MMAudio is a PyTorch-based model for generating synchronized audio from video and/or text inputs, using multimodal joint training across au… | 40 | 2287 | active |
| dstack dstack is an open-source, vendor-agnostic control plane for GPU provisioning and orchestration that works across GPU clouds, Kubernetes, an… | 95 | 2277 | active |
| aixcoder-plugin/aiXcoder-7B Official repository for aiXcoder-7B, an open-weights 7B-parameter code large language model trained on 1.2T tokens for code completion, gen… | 26 | 2271 | active |
| RosettaCommons/RoseTTAFold RoseTTAFold is the official implementation of a deep learning system for predicting protein structures and interactions using a three-track… | 23 | 2269 | stable |
| Amazon SageMaker Containers & SDK The SageMaker Python SDK is an open-source Python library for training and deploying machine learning models on Amazon SageMaker. It suppor… | 94 | 2268 | active |
| ashawkey/stable-dreamfusion A PyTorch implementation of Dreamfusion that generates 3D models from text prompts or images using NeRF combined with Stable Diffusion guid… | 23 | 8865 | maintenance |
| aigc-apps/EasyAnimate EasyAnimate is an end-to-end Python pipeline for high-resolution, long video and image generation based on transformer diffusion (DiT) mode… | 17 | 2266 | active |
| 666DZY666/micronet micronet is a Python library for deep neural network model compression and deployment built on PyTorch. It provides quantization (QAT, PTQ,… | 38 | 2265 | active |
| galilai-group/stable-worldmodel A Python library providing a unified platform for reproducible world model research, covering data collection, training, and evaluation via… | 78 | 2260 | active |
| fishaudio/Bert-VITS2 Bert-VITS2 is a text-to-speech model implementation combining the VITS2 architecture with multilingual BERT embeddings, written in Python. … | 64 | 8805 | maintenance |
| facebookresearch/fvcore fvcore is a lightweight Python core library providing common functionality shared across FAIR's computer vision frameworks such as Detectro… | 73 | 2254 | stable |
| CoinCheung/pytorch-loss A PyTorch library providing a collection of loss functions (focal loss, triplet loss, AMSoftmax, label-smooth CE, dice loss, lovasz-softmax… | 32 | 2251 | active |
| Alpha-VLLM/Lumina-T2X Lumina-T2X is a unified framework for text-to-any-modality generation built on flow-based large diffusion transformers. It supports generat… | 25 | 2250 | active |
| kubeflow/trainer Kubeflow Trainer is a Kubernetes-native platform for distributed AI model training and LLM fine-tuning across frameworks like PyTorch, JAX,… | 93 | 2249 | active |
| ByteDance-Seed/VeOmni VeOmni is a PyTorch-native framework for single- and multi-modal model pre-training and post-training, with a modular, trainer-free design … | 87 | 2246 | active |
| Raster Vision Raster Vision is an open source Python library and low-code framework for building computer vision models on satellite, aerial, and other l… | 57 | 2241 | active |
| NVlabs/MambaVision MambaVision is NVIDIA's official PyTorch implementation of a hybrid Mamba-Transformer vision backbone, published at CVPR 2025. It provides … | 46 | 2235 | active |
| ellisdg/3DUnetCNN A PyTorch library for building, training, and applying 3D U-Net convolutional neural networks for medical image segmentation. It provides c… | 42 | 2235 | active |
| PyGAD PyGAD is an open-source Python 3 library for implementing the genetic algorithm to optimize single- and multi-objective problems. It can al… | 91 | 2230 | active |
| NVIDIA/vid2vid A PyTorch implementation of NVIDIA's video-to-video synthesis method for generating high-resolution (e.g., 2048x1024) photorealistic videos… | 32 | 8692 | maintenance |
| facebookresearch/DiT Official PyTorch implementation of Diffusion Transformers (DiT) from the paper 'Scalable Diffusion Models with Transformers', including mod… | 10 | 8676 | maintenance |
| lifeiteng/vall-e An unofficial PyTorch implementation of VALL-E, a zero-shot text-to-speech model that treats TTS as a conditional language modeling task ov… | 37 | 2221 | active |
| google-deepmind/tapnet Google DeepMind's official repository for Tracking Any Point (TAP), containing the TAP-Vid and TAPVid-3D benchmarks, the TAPIR and TAPNext … | 75 | 2217 | active |
| NX-AI/xlstm Official PyTorch implementation of xLSTM, an extended Long Short-Term Memory recurrent architecture with exponential gating and matrix memo… | 76 | 2205 | active |
| lucidrains/lion-pytorch A PyTorch implementation of the Lion optimizer (Evolved Sign Momentum), discovered by Google Brain via genetic algorithms and claimed to ou… | 59 | 2205 | active |
| AlphaGenome A Python SDK providing programmatic access to Google DeepMind's AlphaGenome model, which predicts genomic regulatory outputs such as gene e… | 89 | 2192 | active |
| TensorFlow Extended (TFX) TensorFlow Extended (TFX) is an end-to-end, Google-production-scale platform for building and deploying production machine learning pipelin… | 87 | 2191 | stable |
| MoonshotAI/MoBA MoBA (Mixture of Block Attention) is a PyTorch implementation of a trainable block-sparse attention mechanism for long-context large langua… | 24 | 2191 | active |
| yyfz/Pi3 Pi3 (π³) is a feed-forward neural network for visual geometry reconstruction that eliminates the need for a fixed reference view, using a p… | 56 | 2188 | active |
| tensorflow/quantum TensorFlow Quantum is a Python framework for hybrid quantum-classical machine learning that integrates Cirq quantum circuits with TensorFlo… | 80 | 2186 | active |
| Harry24k/adversarial-attacks-pytorch Torchattacks is a PyTorch library providing implementations of adversarial attacks to generate adversarial examples against deep learning m… | 23 | 2185 | active |
| ViTAE-Transformer/ViTPose Official PyTorch implementation of ViTPose and ViTPose++, Vision Transformer models for human and generic body pose estimation from NeurIPS… | 55 | 2173 | stable |
| Tencent-Hunyuan/HunyuanVideo-Avatar HunyuanVideo-Avatar is Tencent's open-source model and inference code for high-fidelity audio-driven human animation, generating talking av… | 42 | 2173 | active |
| SpikingJelly SpikingJelly is an open-source deep learning framework for Spiking Neural Networks (SNNs) built on PyTorch. It provides a beginner-friendly… | 77 | 2159 | active |
| PKU-YuanGroup/Helios Helios is a 14B autoregressive diffusion model for real-time, minute-scale video generation supporting text-to-video, image-to-video, and v… | 56 | 2158 | active |
| facebookresearch/mae A PyTorch/GPU re-implementation of the Masked Autoencoders (MAE) paper for self-supervised vision learning. It includes pre-training code, … | 10 | 8362 | maintenance |
| google/gin-config Gin-config is a lightweight Python configuration framework based on dependency injection. Functions or classes decorated with @gin.configur… | 67 | 2157 | stable |
| baowenbo/DAIN DAIN is a PyTorch-based deep learning model for depth-aware video frame interpolation, synthesizing intermediate frames between existing on… | 32 | 8316 | maintenance |
| google/trax Trax is an end-to-end deep learning library built on JAX and TensorFlow that focuses on clear code and speed, developed and maintained by t… | 10 | 8306 | maintenance |
| xinsir6/ControlNetPlus ControlNet++ is an all-in-one ControlNet model and architecture supporting 10+ control types for text-to-image generation and image editing… | 24 | 2143 | active |