The largest collection of PyTorch image encoders / backbones. Including train, eval, inference, export scripts, and pretrained weights -- ResNet, ResNeXT, EfficientNet, NFNet, Vision Transformer (ViT), MobileNetV4, MobileNet-V3 & V2, RegNet, DPN, CSPNet, Swin Transformer, MaxViT, CoAtNet, ConvNeXt, and more
- pytorch
- resnet
- pretrained-models
- pretrained-weights
A reference containing Styles and Keywords that you can use with MidJourney AI. There are also pages showing resolution comparison, image weights, and much more!
Device-native foundation models: 145B on a laptop, 35B on a phone. Latent MoE, SSD-streamed weights, a router that predicts loads a step ahead. Mac, iPhone, Android, Windows.
- edge-ai
- foundation-model
- on-device-ai
Semantic segmentation models with 500+ pretrained convolutional and transformer-based backbones.
PyTorch implementation of the U-Net for image semantic segmentation with high quality images
- pytorch
- unet
- pytorch-unet
- semantic-segmentation
The AI developer platform. Use Weights & Biases to train and fine-tune models, and manage models from experimentation to production.
Keras code and weights files for popular deep learning models.
Translate darknet to tensorflow. Load trained weights, retrain/fine-tune using tensorflow, export constant graph def to mobile devices
The pytorch re-implement of the official efficientdet with SOTA performance in real time and pretrained weights.
Rapid-MLX is an open-source (Apache 2.0) OpenAI- and Anthropic-compatible LLM inference server for Apple Silicon, built on MLX, focused on reliable tool calling for coding agents. Up to 4× faster than Apple's MLX (mlx-lm) on the same weights.
- apple-silicon
- local-llm
- macos
- mlx
A more memory-efficient rewrite of the HF transformers implementation of Llama for use with quantized weights.
Run the full 2.78-trillion-parameter Kimi K3 model, DeepSeek V4.1 Flash or GLM-5.3-Flash beyond available RAM by streaming activated weights directly from NVMe. A dependency-free, embeddable C inference engine.
One-click character swap for videos: an open-source, local Higgsfield Genjutsu alternative built on ComfyUI and MiniMax H3, with FLUX.3 Klein relighting.
- character-swap
- comfyui
- minimax-h3
- nextjs
Typeface for setting text in many sizes, weights, and languages. Designed to complement Source Sans.
- opentype-fonts
- otf
- afdko
- ttf-fonts
YOLOv4, YOLOv4-tiny, YOLOv3, YOLOv3-tiny Implemented in Tensorflow 2.0, Android. Convert YOLO v4 .weights tensorflow, tensorrt and tflite
Caffe implementation of Google MobileNet SSD detection network, with pretrained weights on VOC0712 and mAP=0.727.
- ssd
- mobilenet-ssd
- caffe
- mobilenet
Featherlight is a very lightweight jQuery lightbox plugin. It's simple yet flexible and easy to use. Featherlight has minimal css and uses no inline styles, everything is name-spaced, it's completely customizable via config object and offers image, ajax and iframe support out of the box. Featherlights small footprint weights about 4kB – in total.
- lightbox
- javascript
- ajax
- jquery
Magenta RealTime 2: An Open-Weights Live Music Model
Open decision models from 0.8B to 27B: a calibrated probability for every option in one pass, and a short thought only when unsure. Engine for vLLM, llama.cpp and MLX.
- agents
- calibration
- classification
- decision-making
A PyTorch impl of EfficientDet faithful to the original Google impl w/ ported weights
A 2D Unity simulation in which cars learn to navigate themselves through different courses. The cars are steered by a feedforward neural network. The weights of the network are trained using a modified genetic algorithm.
- artificial-neural-networks
- neural-networks
- evolutionary-algorithms
- genetic-algorithm
Open-source Higgsfield on ComfyUI: 120 image, video, VFX and audio apps, every model open weights; runs on Modal or your own GPU
Keras implementation of Deeplab v3+ with pretrained weights
Open-source pan-CJK font families with enormous range of weights, the CJK sans and serif you have dreamed of.
- source-han-sans
- source-han-serif
- source-han-truetype