PyTorch HubFor Researchers – PyTorch (original) (raw)
Explore and extend models from the latest cutting edge research.
Discover and publish models to a pre-trained model repository designed for research exploration. Check out the models for Researchers, or learn How It Works. Contribute Models.
*This is a beta release – we will be collecting feedback and improving the PyTorch Hub over the coming months.
YOLOv5
Ultralytics YOLOv5 🚀 for object detection, instance segmentation and image classification.
RoBERTa
A Robustly Optimized BERT Pretraining Approach
ResNet
Deep residual networks pre-trained on ImageNet
ResNext
Next generation ResNets, more efficient and accurate
ShuffleNet v2
An efficient ConvNet optimized for speed and memory, pre-trained on ImageNet
Deeplabv3
DeepLabV3 models with ResNet-50, ResNet-101 and MobileNet-V3 backbones
AlexNet
The 2012 ImageNet winner achieved a top-5 error of 15.3%, more than 10.8 percentage points lower than that of the runner up.
Densenet
Dense Convolutional Network (DenseNet), connects each layer to every other layer in a feed-forward fashion.
vgg-nets
Award winning ConvNets from 2014 ImageNet ILSVRC challenge
FCN
Fully-Convolutional Network model with ResNet-50 and ResNet-101 backbones
GoogLeNet
GoogLeNet was based on a deep convolutional neural network architecture codenamed “Inception” which won ImageNet 2014.
Inception_v3
Also called GoogleNetv3, a famous ConvNet trained on ImageNet from 2015
MobileNet v2
Efficient networks optimized for speed and memory, with residual blocks
SSD
Single Shot MultiBox Detector model for object detection
Tacotron 2
The Tacotron 2 model for generating mel spectrograms from text
WaveGlow
WaveGlow model for generating speech from mel spectrograms (generated by Tacotron2)
GPUNet
GPUNet is a new family of Convolutional Neural Networks designed to max out the performance of NVIDIA GPU and TensorRT.
EfficientNet
EfficientNets are a family of image classification models, which achieve state-of-the-art accuracy, being an order-of-magnitude smaller and faster. Trained with mixed precision using Tensor Cores.
HiFi GAN
The HiFi GAN model for generating waveforms from mel spectrograms
ResNet50
ResNet50 model trained with mixed precision using Tensor Cores.
ResNeXt101
ResNet with bottleneck 3×3 Convolutions substituted by 3×3 Grouped Convolutions, trained with mixed precision using Tensor Cores.
SE-ResNeXt101
ResNeXt with Squeeze-and-Excitation module added, trained with mixed precision using Tensor Cores.
MiDaS
MiDaS models for computing relative depth from a single image.
SNNMLP
Brain-inspired Multilayer Perceptron with Spiking Neurons
GhostNet
Efficient networks by generating more features from cheap operations
3D ResNet
Resnet Style Video classification networks pretrained on the Kinetics 400 dataset
SlowFast
SlowFast networks pretrained on the Kinetics 400 dataset
X3D
X3D networks pretrained on the Kinetics 400 dataset
YOLOP
YOLOP pretrained on the BDD100K dataset
Once-for-All
Once-for-all (OFA) decouples training and search, and achieves efficient inference across various edge devices and resource constraints.
ProxylessNAS
Proxylessly specialize CNN architectures for different hardware platforms.
IBN-Net
Networks with domain/appearance invariance
U-Net for brain MRI
U-Net with batch normalization for biomedical image segmentation with pretrained weights for abnormality segmentation in brain MRI
MEAL_V2
Boosting Tiny and Efficient Models using Knowledge Distillation.
ResNext WSL
ResNext models trained with billion scale weakly-supervised data.
HarDNet
Harmonic DenseNet pre-trained on ImageNet
SimpleNet
Lets Keep it simple, Using simple architectures to outperform deeper and more complex architectures
ntsnet
classify birds using this fine-grained image classifier