EfficientNet

Category: Efficient/Mobile · Code: timm/models/efficientnet.py · Architectures: 115 · Pretrained weights: 119 (113 with ImageNet-1k results)

EfficientNet and EfficientNetV2, and the related mobile architectures built from the same depthwise separable, inverted residual (MBConv) and fused MBConv blocks: MixNet, MnasNet, FBNet, Single-Path NAS, MobileNet-V1 / V2, EfficientNet-Lite, EdgeTPU and TinyNet.

Papers

• EfficientNetV2: Smaller Models and Faster Training
• EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
• CondConv: Conditionally Parameterized Convolutions for Efficient Inference
• Adversarial Examples Improve Image Recognition
• Self-training with Noisy Student improves ImageNet classification
• Knapsack Pruning with Inner Distillation
• Evolving Normalization-Activation Layers
• Searching for Fast Model Families on Datacenter Accelerators
• MixConv: Mixed Depthwise Convolutional Kernels
• MnasNet: Platform-Aware Neural Architecture Search for Mobile
• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
• MobileNetV2: Inverted Residuals and Linear Bottlenecks
• Searching for Efficient Neural Architectures for On-Device ML on Edge TPUs
• FBNet: Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Search
• Single-Path NAS: Designing Hardware-Efficient ConvNets in less than 4 Hours
• Model Rubik’s Cube: Twisting Resolution, Depth and Width for TinyNets

Usage

See the Quickstart for loading pretrained weights, image preprocessing and fine-tuning, and Feature Extraction for using the models as backbones.

>>> import timm
>>> model = timm.create_model('efficientnet_b0.ra_in1k', pretrained=True)
>>> data_config = timm.data.resolve_model_data_config(model)

Model API support

✓ supported, ✓ (square) works for square sizes only, ✗ not supported, – not needed (the model already accepts any input size). Checked by running each feature on a randomly initialized model (EfficientNet with mobilenetv2_035 at 224), other architectures of the same class are expected to behave the same. See the Model API page for details.

Architectures

GMACs and parameter counts are for the default input size and classifier of each architecture, from the benchmark results.

Pretrained weights

Top-1 is ImageNet-1k validation accuracy from results-imagenet.csv at the listed eval size, - if the weight has no ImageNet-1k result (e.g. pretraining or image-text weights). Classes is the size of the pretrained classifier, none for weights without one.

Weights are grouped by source: the paper or origin given in the pretrained config, else a Hub organization other than timm, else the first part of the pretrained tag (architecture.tag). Sources with a single weight are listed under other.

ra tags

ns tags

ap tags

aa tags

ra4 tags

ra2 tags

rmsp tags

sw tags

agc tags

lamb tags

Other