dm nfnet f2.dm in1k EU AI Act Compliance Profile

Model card for dm_nfnet_f2.dm_in1k

A NFNet (Normalization Free Network) image classification model. Trained on ImageNet-1k by paper authors.

Normalization Free Networks are (pre-activation) ResNet-like models without any normalization layers. Instead of Batch Normalization or alternatives, they use Scaled Weight Standardization and specifically placed scalar gains in residual path and at non-linearities based on signal propagation analysis.

Model Details

• Model Type: Image classification / feature backbone
• Model Stats:
  • Params (M): 193.8
  • GMACs: 33.8
  • Activations (M): 41.8
  • Image size: train = 256 x 256, test = 352 x 352
• Papers:
  • High-Performance Large-Scale Image Recognition Without Normalization: https://arxiv.org/abs/2102.06171
  • Characterizing signal propagation to close the performance gap in unnormalized ResNets: https://arxiv.org/abs/2101.08692
• Original: https://github.com/deepmind/deepmind-research/tree/master/nfnets
• Dataset: ImageNet-1k

Model Usage

Image Classification

from urllib.request import urlopen
from PIL import Image
import timm

img = Image.open(urlopen(
    'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))

model = timm.create_model('dm_nfnet_f2.dm_in1k', pretrained=True)
model = model.eval()

# get model specific transforms (normalization, resize)
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)

output = model(transforms(img).unsqueeze(0))  # unsqueeze single image into batch of 1

top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5)

Feature Map Extraction

from urllib.request import urlopen
from PIL import Image
import timm

img = Image.open(urlopen(
    'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))

model = timm.create_model(

Your obligations as a dm nfnet f2.dm in1k deployer

Track your progress

Sign up to track obligations, receive deadline reminders, and generate audit packages.

Tags

Library