Views
No views yet
Tu et al., 2022 — MaxViT: Multi-Axis Vision Transformer (arXiv:2204.01697)
timm/maxvit_large_tf_224.in1k,
converted to Lucid-native safetensors.| Tag | acc@1 | acc@5 | Params | GFLOPs | Size | Source |
|---|---|---|---|---|---|---|
IN1K (default) | 85.17 | 97.17 | 211.8M | — | 808.78 MB | timm |
1import lucid.models as models
2from lucid.models.weights import MaxViTLargeWeights
3
4# default tag
5model = models.maxvit_large_cls(pretrained=True)
6
7# explicit tag (enum or string)
8model = models.maxvit_large_cls(weights=MaxViTLargeWeights.IN1K)
9model = models.maxvit_large_cls(pretrained="IN1K")
10
11# preprocessing travels with the weights
12weights = MaxViTLargeWeights.IN1K
13preprocess = weights.transforms()
14logits = model(preprocess(image)[None]).logitstimm/maxvit_large_tf_224.in1k via
python -m tools.convert_weights maxvit_large --tag IN1K.
Key mapping + numerical parity verified against the source.apache-2.0 — inherited from the original weights.@inproceedings{tu2022maxvit,
title={MaxViT: Multi-Axis Vision Transformer},
author={Tu, Zhengzhong and Talebi, Hossein and Zhang, Han and Yang, Feng and Milanfar, Peyman and Bovik, Alan and Li, Yinxiao},
booktitle={ECCV}, year={2022}
}