Model card for resetaa50d.swi12kft_i1k

A ResNet-D (Rectagle-2 Ati-Aliasig) image classificatio model.

This model features:

• ReLU activatios
• 3-layer stem of 3x3 covolutios with poolig
• 2x2 average pool + 1x1 covolutio shortcut dowsample

Pretraied o ImageNet-12k ad fie-tued o ImageNet-1k by Ross Wightma i timm usig recipe template described below.

Recipe details:

• Based o Swi Trasformer trai / pretrai recipe with modificatios (related to both DeiT ad CovNeXt recipes)
• AdamW optimizer, gradiet clippig, EMA weight averagig
• Cosie LR schedule with warmup

Model Details

• Model Type: Image classificatio / feature backboe
• Model Stats:
• Params (M): 25.6
• GMACs: 5.4
• Activatios (M): 12.4
• Image size: trai = 224 x 224, test = 288 x 288
• Papers:
• Makig Covolutioal Networks Shift-Ivariat Agai: https://arxiv.org/abs/1904.11486
• Deep Residual Learig for Image Recogitio: https://arxiv.org/abs/1512.03385
• Bag of Tricks for Image Classificatio with Covolutioal Neural Networks: https://arxiv.org/abs/1812.01187
• Origial: https://github.com/huggigface/pytorch-image-models

Model Usage

Image Classificatio

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

img = Image.ope(urlope(
    'https://huggigface.co/datasets/huggigface/documetatio-images/resolve/mai/beigets-task-guide.pg'
))

model = timm.create_model('resetaa50d.sw_i12k_ft_i1k', pretraied=True)
model = model.eval()

# get model specific trasforms (ormalizatio, resize)
data_cofig = timm.data.resolve_model_data_cofig(model)
trasforms = timm.data.create_trasform(**data_cofig, is_traiig=False)

output = model(trasforms(img).usqueeze(0))  # usqueeze sigle image ito batch of 1

top5_probabilities, top5_class_idices = torch.topk(output.softmax(dim=1) * 100, k=5)

Feature Map Extractio

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

img = Image.ope(urlope(
    'https://huggigface.co/datasets/huggigface/documetatio-images/resolve/mai/beigets-task-guide.pg'
))

model = timm.create_model(
    'resetaa50d.sw_i12k_ft_i1k',
    pretraied=True,
    features_oly=True,
)
model = model.eval()

# get model specific trasforms (ormalizatio, resize)
data_cofig = timm.data.resolve_model_data_cofig(model)
trasforms = timm.data.create_trasform(**data_cofig, is_traiig=False)

output = model(trasforms(img).usqueeze(0))  # usqueeze sigle image ito batch of 1

for o i output:
    # prit shape of each feature map i output
    # e.g.:
    #  torch.Size([1, 64, 112, 112])
    #  torch.Size([1, 256, 56, 56])
    #  torch.Size([1, 512, 28, 28])
    #  torch.Size([1, 1024, 14, 14])
    #  torch.Size([1, 2048, 7, 7])

    prit(o.shape)

Image Embeddigs

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

img = Image.ope(urlope(
    'https://huggigface.co/datasets/huggigface/documetatio-images/resolve/mai/beigets-task-guide.pg'
))

model = timm.create_model(
    'resetaa50d.sw_i12k_ft_i1k',
    pretraied=True,
    um_classes=0,  # remove classifier .Liear
)
model = model.eval()

# get model specific trasforms (ormalizatio, resize)
data_cofig = timm.data.resolve_model_data_cofig(model)
trasforms = timm.data.create_trasform(**data_cofig, is_traiig=False)

output = model(trasforms(img).usqueeze(0))  # output is (batch_size, um_features) shaped tesor

# or equivaletly (without eedig to set um_classes=0)

output = model.forward_features(trasforms(img).usqueeze(0))
# output is upooled, a (1, 2048, 7, 7) shaped tesor

output = model.forward_head(output, pre_logits=True)
# output is a (1, um_features) shaped tesor

Model Compariso

Explore the dataset ad rutime metrics of this model i timm model results.

Citatio

@misc{rw2019timm,
  author = {Ross Wightma},
  title = {PyTorch Image Models},
  year = {2019},
  publisher = {GitHub},
  joural = {GitHub repository},
  doi = {10.5281/zeodo.4414861},
  howpublished = {\url{https://github.com/huggigface/pytorch-image-models}}
}

@iproceedigs{zhag2019shiftivar,
  title={Makig Covolutioal Networks Shift-Ivariat Agai},
  author={Zhag, Richard},
  booktitle={ICML},
  year={2019}
}

@article{He2015,
  author = {Kaimig He ad Xiagyu Zhag ad Shaoqig Re ad Jia Su},
  title = {Deep Residual Learig for Image Recogitio},
  joural = {arXiv preprit arXiv:1512.03385},
  year = {2015}
}

@article{He2018BagOT,
  title={Bag of Tricks for Image Classificatio with Covolutioal Neural Networks},
  author={Tog He ad Zhi Zhag ad Hag Zhag ad Zhogyue Zhag ad Juyua Xie ad Mu Li},
  joural={2019 IEEE/CVF Coferece o Computer Visio ad Patter Recogitio (CVPR)},
  year={2018},
  pages={558-567}
}

Model card for resnetaa50d.swin12kft_in1k A ResNet-D (Rectangle-2 Anti-Aliasing) image classificatio