Update model config and README
Browse files- README.md +244 -1
- config.json +3 -2
README.md
CHANGED
|
@@ -3,5 +3,248 @@ tags:
|
|
| 3 |
- image-classification
|
| 4 |
- timm
|
| 5 |
library_tag: timm
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
---
|
| 7 |
-
# Model card for maxvit_large_tf_384.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
- image-classification
|
| 4 |
- timm
|
| 5 |
library_tag: timm
|
| 6 |
+
license: apache-2.0
|
| 7 |
+
datasets:
|
| 8 |
+
- imagenet-1k
|
| 9 |
+
- imagenet-21k
|
| 10 |
---
|
| 11 |
+
# Model card for maxvit_large_tf_384.in21k_ft_in1k
|
| 12 |
+
|
| 13 |
+
An official MaxViT image classification model. Pretrained in tensorflow on ImageNet-21k (21843 Google specific instance of ImageNet-22k) and fine-tuned on ImageNet-1k by paper authors.
|
| 14 |
+
|
| 15 |
+
Ported from official Tensorflow implementation (https://github.com/google-research/maxvit) to PyTorch by Ross Wightman.
|
| 16 |
+
|
| 17 |
+
### Model Variants in [maxxvit.py](https://github.com/rwightman/pytorch-image-models/blob/main/timm/models/maxxvit.py)
|
| 18 |
+
|
| 19 |
+
MaxxViT covers a number of related model architectures that share a common structure including:
|
| 20 |
+
- CoAtNet - Combining MBConv (depthwise-separable) convolutional blocks in early stages with self-attention transformer blocks in later stages.
|
| 21 |
+
- MaxViT - Uniform blocks across all stages, each containing a MBConv (depthwise-separable) convolution block followed by two self-attention blocks with different partitioning schemes (window followed by grid).
|
| 22 |
+
- CoAtNeXt - A timm specific arch that uses ConvNeXt blocks in place of MBConv blocks in CoAtNet. All normalization layers are LayerNorm (no BatchNorm).
|
| 23 |
+
- MaxxViT - A timm specific arch that uses ConvNeXt blocks in place of MBConv blocks in MaxViT. All normalization layers are LayerNorm (no BatchNorm).
|
| 24 |
+
- MaxxViT-V2 - A MaxxViT variation that removes the window block attention leaving only ConvNeXt blocks and grid attention w/ more width to compensate.
|
| 25 |
+
|
| 26 |
+
Aside from the major variants listed above, there are more subtle changes from model to model. Any model name with the string `rw` are `timm` specific configs w/ modelling adjustments made to favour PyTorch eager use. These were created while training initial reproductions of the models so there are variations.
|
| 27 |
+
All models with the string `tf` are models exactly matching Tensorflow based models by the original paper authors with weights ported to PyTorch. This covers a number of MaxViT models. The official CoAtNet models were never released.
|
| 28 |
+
|
| 29 |
+
## Model Details
|
| 30 |
+
- **Model Type:** Image classification / feature backbone
|
| 31 |
+
- **Model Stats:**
|
| 32 |
+
- Params (M): 212.0
|
| 33 |
+
- GMACs: 132.6
|
| 34 |
+
- Activations (M): 445.8
|
| 35 |
+
- Image size: 384 x 384
|
| 36 |
+
- **Papers:**
|
| 37 |
+
- MaxViT: Multi-Axis Vision Transformer: https://arxiv.org/abs/2204.01697
|
| 38 |
+
- **Dataset:** ImageNet-1k
|
| 39 |
+
- **Pretrain Dataset:** ImageNet-21k
|
| 40 |
+
|
| 41 |
+
## Model Usage
|
| 42 |
+
### Image Classification
|
| 43 |
+
```python
|
| 44 |
+
from urllib.request import urlopen
|
| 45 |
+
from PIL import Image
|
| 46 |
+
import timm
|
| 47 |
+
|
| 48 |
+
img = Image.open(
|
| 49 |
+
urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'))
|
| 50 |
+
|
| 51 |
+
model = timm.create_model('maxvit_large_tf_384.in21k_ft_in1k', pretrained=True)
|
| 52 |
+
model = model.eval()
|
| 53 |
+
|
| 54 |
+
# get model specific transforms (normalization, resize)
|
| 55 |
+
data_config = timm.data.resolve_model_data_config(model)
|
| 56 |
+
transforms = timm.data.create_transform(**data_config, is_training=False)
|
| 57 |
+
|
| 58 |
+
output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1
|
| 59 |
+
|
| 60 |
+
top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5)
|
| 61 |
+
```
|
| 62 |
+
|
| 63 |
+
### Feature Map Extraction
|
| 64 |
+
```python
|
| 65 |
+
from urllib.request import urlopen
|
| 66 |
+
from PIL import Image
|
| 67 |
+
import timm
|
| 68 |
+
|
| 69 |
+
img = Image.open(
|
| 70 |
+
urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'))
|
| 71 |
+
|
| 72 |
+
model = timm.create_model(
|
| 73 |
+
'maxvit_large_tf_384.in21k_ft_in1k',
|
| 74 |
+
pretrained=True,
|
| 75 |
+
features_only=True,
|
| 76 |
+
)
|
| 77 |
+
model = model.eval()
|
| 78 |
+
|
| 79 |
+
# get model specific transforms (normalization, resize)
|
| 80 |
+
data_config = timm.data.resolve_model_data_config(model)
|
| 81 |
+
transforms = timm.data.create_transform(**data_config, is_training=False)
|
| 82 |
+
|
| 83 |
+
output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1
|
| 84 |
+
|
| 85 |
+
for o in output:
|
| 86 |
+
# print shape of each feature map in output
|
| 87 |
+
# e.g.:
|
| 88 |
+
# torch.Size([1, 128, 192, 192])
|
| 89 |
+
# torch.Size([1, 128, 96, 96])
|
| 90 |
+
# torch.Size([1, 256, 48, 48])
|
| 91 |
+
# torch.Size([1, 512, 24, 24])
|
| 92 |
+
# torch.Size([1, 1024, 12, 12])
|
| 93 |
+
print(o.shape)
|
| 94 |
+
```
|
| 95 |
+
|
| 96 |
+
### Image Embeddings
|
| 97 |
+
```python
|
| 98 |
+
from urllib.request import urlopen
|
| 99 |
+
from PIL import Image
|
| 100 |
+
import timm
|
| 101 |
+
|
| 102 |
+
img = Image.open(
|
| 103 |
+
urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'))
|
| 104 |
+
|
| 105 |
+
model = timm.create_model(
|
| 106 |
+
'maxvit_large_tf_384.in21k_ft_in1k',
|
| 107 |
+
pretrained=True,
|
| 108 |
+
num_classes=0, # remove classifier nn.Linear
|
| 109 |
+
)
|
| 110 |
+
model = model.eval()
|
| 111 |
+
|
| 112 |
+
# get model specific transforms (normalization, resize)
|
| 113 |
+
data_config = timm.data.resolve_model_data_config(model)
|
| 114 |
+
transforms = timm.data.create_transform(**data_config, is_training=False)
|
| 115 |
+
|
| 116 |
+
output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor
|
| 117 |
+
|
| 118 |
+
# or equivalently (without needing to set num_classes=0)
|
| 119 |
+
|
| 120 |
+
output = model.forward_features(transforms(img).unsqueeze(0))
|
| 121 |
+
# output is unpooled (ie.e a (batch_size, num_features, H, W) tensor
|
| 122 |
+
|
| 123 |
+
output = model.forward_head(output, pre_logits=True)
|
| 124 |
+
# output is (batch_size, num_features) tensor
|
| 125 |
+
```
|
| 126 |
+
|
| 127 |
+
## Model Comparison
|
| 128 |
+
### By Top-1
|
| 129 |
+
|model |top1 |top5 |samples / sec |Params (M) |GMAC |Act (M)|
|
| 130 |
+
|------------------------------------------------------------------------------------------------------------------------|----:|----:|--------------:|--------------:|-----:|------:|
|
| 131 |
+
|[maxvit_xlarge_tf_512.in21k_ft_in1k](https://huggingface.co/timm/maxvit_xlarge_tf_512.in21k_ft_in1k) |88.53|98.64| 21.76| 475.77|534.14|1413.22|
|
| 132 |
+
|[maxvit_xlarge_tf_384.in21k_ft_in1k](https://huggingface.co/timm/maxvit_xlarge_tf_384.in21k_ft_in1k) |88.32|98.54| 42.53| 475.32|292.78| 668.76|
|
| 133 |
+
|[maxvit_base_tf_512.in21k_ft_in1k](https://huggingface.co/timm/maxvit_base_tf_512.in21k_ft_in1k) |88.20|98.53| 50.87| 119.88|138.02| 703.99|
|
| 134 |
+
|[maxvit_large_tf_512.in21k_ft_in1k](https://huggingface.co/timm/maxvit_large_tf_512.in21k_ft_in1k) |88.04|98.40| 36.42| 212.33|244.75| 942.15|
|
| 135 |
+
|[maxvit_large_tf_384.in21k_ft_in1k](https://huggingface.co/timm/maxvit_large_tf_384.in21k_ft_in1k) |87.98|98.56| 71.75| 212.03|132.55| 445.84|
|
| 136 |
+
|[maxvit_base_tf_384.in21k_ft_in1k](https://huggingface.co/timm/maxvit_base_tf_384.in21k_ft_in1k) |87.92|98.54| 104.71| 119.65| 73.80| 332.90|
|
| 137 |
+
|[maxvit_rmlp_base_rw_384.sw_in12k_ft_in1k](https://huggingface.co/timm/maxvit_rmlp_base_rw_384.sw_in12k_ft_in1k) |87.81|98.37| 106.55| 116.14| 70.97| 318.95|
|
| 138 |
+
|[maxxvitv2_rmlp_base_rw_384.sw_in12k_ft_in1k](https://huggingface.co/timm/maxxvitv2_rmlp_base_rw_384.sw_in12k_ft_in1k) |87.47|98.37| 149.49| 116.09| 72.98| 213.74|
|
| 139 |
+
|[coatnet_rmlp_2_rw_384.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_rmlp_2_rw_384.sw_in12k_ft_in1k) |87.39|98.31| 160.80| 73.88| 47.69| 209.43|
|
| 140 |
+
|[maxvit_rmlp_base_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/maxvit_rmlp_base_rw_224.sw_in12k_ft_in1k) |86.89|98.02| 375.86| 116.14| 23.15| 92.64|
|
| 141 |
+
|[maxxvitv2_rmlp_base_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/maxxvitv2_rmlp_base_rw_224.sw_in12k_ft_in1k) |86.64|98.02| 501.03| 116.09| 24.20| 62.77|
|
| 142 |
+
|[maxvit_base_tf_512.in1k](https://huggingface.co/timm/maxvit_base_tf_512.in1k) |86.60|97.92| 50.75| 119.88|138.02| 703.99|
|
| 143 |
+
|[coatnet_2_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_2_rw_224.sw_in12k_ft_in1k) |86.57|97.89| 631.88| 73.87| 15.09| 49.22|
|
| 144 |
+
|[maxvit_large_tf_512.in1k](https://huggingface.co/timm/maxvit_large_tf_512.in1k) |86.52|97.88| 36.04| 212.33|244.75| 942.15|
|
| 145 |
+
|[coatnet_rmlp_2_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_rmlp_2_rw_224.sw_in12k_ft_in1k) |86.49|97.90| 620.58| 73.88| 15.18| 54.78|
|
| 146 |
+
|[maxvit_base_tf_384.in1k](https://huggingface.co/timm/maxvit_base_tf_384.in1k) |86.29|97.80| 101.09| 119.65| 73.80| 332.90|
|
| 147 |
+
|[maxvit_large_tf_384.in1k](https://huggingface.co/timm/maxvit_large_tf_384.in1k) |86.23|97.69| 70.56| 212.03|132.55| 445.84|
|
| 148 |
+
|[maxvit_small_tf_512.in1k](https://huggingface.co/timm/maxvit_small_tf_512.in1k) |86.10|97.76| 88.63| 69.13| 67.26| 383.77|
|
| 149 |
+
|[maxvit_tiny_tf_512.in1k](https://huggingface.co/timm/maxvit_tiny_tf_512.in1k) |85.67|97.58| 144.25| 31.05| 33.49| 257.59|
|
| 150 |
+
|[maxvit_small_tf_384.in1k](https://huggingface.co/timm/maxvit_small_tf_384.in1k) |85.54|97.46| 188.35| 69.02| 35.87| 183.65|
|
| 151 |
+
|[maxvit_tiny_tf_384.in1k](https://huggingface.co/timm/maxvit_tiny_tf_384.in1k) |85.11|97.38| 293.46| 30.98| 17.53| 123.42|
|
| 152 |
+
|[maxvit_large_tf_224.in1k](https://huggingface.co/timm/maxvit_large_tf_224.in1k) |84.93|96.97| 247.71| 211.79| 43.68| 127.35|
|
| 153 |
+
|[coatnet_rmlp_1_rw2_224.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_rmlp_1_rw2_224.sw_in12k_ft_in1k) |84.90|96.96| 1025.45| 41.72| 8.11| 40.13|
|
| 154 |
+
|[maxvit_base_tf_224.in1k](https://huggingface.co/timm/maxvit_base_tf_224.in1k) |84.85|96.99| 358.25| 119.47| 24.04| 95.01|
|
| 155 |
+
|[maxxvit_rmlp_small_rw_256.sw_in1k](https://huggingface.co/timm/maxxvit_rmlp_small_rw_256.sw_in1k) |84.63|97.06| 575.53| 66.01| 14.67| 58.38|
|
| 156 |
+
|[coatnet_rmlp_2_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_rmlp_2_rw_224.sw_in1k) |84.61|96.74| 625.81| 73.88| 15.18| 54.78|
|
| 157 |
+
|[maxvit_rmlp_small_rw_224.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_small_rw_224.sw_in1k) |84.49|96.76| 693.82| 64.90| 10.75| 49.30|
|
| 158 |
+
|[maxvit_small_tf_224.in1k](https://huggingface.co/timm/maxvit_small_tf_224.in1k) |84.43|96.83| 647.96| 68.93| 11.66| 53.17|
|
| 159 |
+
|[maxvit_rmlp_tiny_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_tiny_rw_256.sw_in1k) |84.23|96.78| 807.21| 29.15| 6.77| 46.92|
|
| 160 |
+
|[coatnet_1_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_1_rw_224.sw_in1k) |83.62|96.38| 989.59| 41.72| 8.04| 34.60|
|
| 161 |
+
|[maxvit_tiny_rw_224.sw_in1k](https://huggingface.co/timm/maxvit_tiny_rw_224.sw_in1k) |83.50|96.50| 1100.53| 29.06| 5.11| 33.11|
|
| 162 |
+
|[maxvit_tiny_tf_224.in1k](https://huggingface.co/timm/maxvit_tiny_tf_224.in1k) |83.41|96.59| 1004.94| 30.92| 5.60| 35.78|
|
| 163 |
+
|[coatnet_rmlp_1_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_rmlp_1_rw_224.sw_in1k) |83.36|96.45| 1093.03| 41.69| 7.85| 35.47|
|
| 164 |
+
|[maxxvitv2_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxxvitv2_nano_rw_256.sw_in1k) |83.11|96.33| 1276.88| 23.70| 6.26| 23.05|
|
| 165 |
+
|[maxxvit_rmlp_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxxvit_rmlp_nano_rw_256.sw_in1k) |83.03|96.34| 1341.24| 16.78| 4.37| 26.05|
|
| 166 |
+
|[maxvit_rmlp_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_nano_rw_256.sw_in1k) |82.96|96.26| 1283.24| 15.50| 4.47| 31.92|
|
| 167 |
+
|[maxvit_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_nano_rw_256.sw_in1k) |82.93|96.23| 1218.17| 15.45| 4.46| 30.28|
|
| 168 |
+
|[coatnet_bn_0_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_bn_0_rw_224.sw_in1k) |82.39|96.19| 1600.14| 27.44| 4.67| 22.04|
|
| 169 |
+
|[coatnet_0_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_0_rw_224.sw_in1k) |82.39|95.84| 1831.21| 27.44| 4.43| 18.73|
|
| 170 |
+
|[coatnet_rmlp_nano_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_rmlp_nano_rw_224.sw_in1k) |82.05|95.87| 2109.09| 15.15| 2.62| 20.34|
|
| 171 |
+
|[coatnext_nano_rw_224.sw_in1k](https://huggingface.co/timm/coatnext_nano_rw_224.sw_in1k) |81.95|95.92| 2525.52| 14.70| 2.47| 12.80|
|
| 172 |
+
|[coatnet_nano_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_nano_rw_224.sw_in1k) |81.70|95.64| 2344.52| 15.14| 2.41| 15.41|
|
| 173 |
+
|[maxvit_rmlp_pico_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_pico_rw_256.sw_in1k) |80.53|95.21| 1594.71| 7.52| 1.85| 24.86|
|
| 174 |
+
|
| 175 |
+
|
| 176 |
+
### By Throughput (samples / sec)
|
| 177 |
+
|model |top1 |top5 |samples / sec |Params (M) |GMAC |Act (M)|
|
| 178 |
+
|------------------------------------------------------------------------------------------------------------------------|----:|----:|--------------:|--------------:|-----:|------:|
|
| 179 |
+
|[coatnext_nano_rw_224.sw_in1k](https://huggingface.co/timm/coatnext_nano_rw_224.sw_in1k) |81.95|95.92| 2525.52| 14.70| 2.47| 12.80|
|
| 180 |
+
|[coatnet_nano_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_nano_rw_224.sw_in1k) |81.70|95.64| 2344.52| 15.14| 2.41| 15.41|
|
| 181 |
+
|[coatnet_rmlp_nano_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_rmlp_nano_rw_224.sw_in1k) |82.05|95.87| 2109.09| 15.15| 2.62| 20.34|
|
| 182 |
+
|[coatnet_0_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_0_rw_224.sw_in1k) |82.39|95.84| 1831.21| 27.44| 4.43| 18.73|
|
| 183 |
+
|[coatnet_bn_0_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_bn_0_rw_224.sw_in1k) |82.39|96.19| 1600.14| 27.44| 4.67| 22.04|
|
| 184 |
+
|[maxvit_rmlp_pico_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_pico_rw_256.sw_in1k) |80.53|95.21| 1594.71| 7.52| 1.85| 24.86|
|
| 185 |
+
|[maxxvit_rmlp_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxxvit_rmlp_nano_rw_256.sw_in1k) |83.03|96.34| 1341.24| 16.78| 4.37| 26.05|
|
| 186 |
+
|[maxvit_rmlp_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_nano_rw_256.sw_in1k) |82.96|96.26| 1283.24| 15.50| 4.47| 31.92|
|
| 187 |
+
|[maxxvitv2_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxxvitv2_nano_rw_256.sw_in1k) |83.11|96.33| 1276.88| 23.70| 6.26| 23.05|
|
| 188 |
+
|[maxvit_nano_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_nano_rw_256.sw_in1k) |82.93|96.23| 1218.17| 15.45| 4.46| 30.28|
|
| 189 |
+
|[maxvit_tiny_rw_224.sw_in1k](https://huggingface.co/timm/maxvit_tiny_rw_224.sw_in1k) |83.50|96.50| 1100.53| 29.06| 5.11| 33.11|
|
| 190 |
+
|[coatnet_rmlp_1_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_rmlp_1_rw_224.sw_in1k) |83.36|96.45| 1093.03| 41.69| 7.85| 35.47|
|
| 191 |
+
|[coatnet_rmlp_1_rw2_224.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_rmlp_1_rw2_224.sw_in12k_ft_in1k) |84.90|96.96| 1025.45| 41.72| 8.11| 40.13|
|
| 192 |
+
|[maxvit_tiny_tf_224.in1k](https://huggingface.co/timm/maxvit_tiny_tf_224.in1k) |83.41|96.59| 1004.94| 30.92| 5.60| 35.78|
|
| 193 |
+
|[coatnet_1_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_1_rw_224.sw_in1k) |83.62|96.38| 989.59| 41.72| 8.04| 34.60|
|
| 194 |
+
|[maxvit_rmlp_tiny_rw_256.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_tiny_rw_256.sw_in1k) |84.23|96.78| 807.21| 29.15| 6.77| 46.92|
|
| 195 |
+
|[maxvit_rmlp_small_rw_224.sw_in1k](https://huggingface.co/timm/maxvit_rmlp_small_rw_224.sw_in1k) |84.49|96.76| 693.82| 64.90| 10.75| 49.30|
|
| 196 |
+
|[maxvit_small_tf_224.in1k](https://huggingface.co/timm/maxvit_small_tf_224.in1k) |84.43|96.83| 647.96| 68.93| 11.66| 53.17|
|
| 197 |
+
|[coatnet_2_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_2_rw_224.sw_in12k_ft_in1k) |86.57|97.89| 631.88| 73.87| 15.09| 49.22|
|
| 198 |
+
|[coatnet_rmlp_2_rw_224.sw_in1k](https://huggingface.co/timm/coatnet_rmlp_2_rw_224.sw_in1k) |84.61|96.74| 625.81| 73.88| 15.18| 54.78|
|
| 199 |
+
|[coatnet_rmlp_2_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_rmlp_2_rw_224.sw_in12k_ft_in1k) |86.49|97.90| 620.58| 73.88| 15.18| 54.78|
|
| 200 |
+
|[maxxvit_rmlp_small_rw_256.sw_in1k](https://huggingface.co/timm/maxxvit_rmlp_small_rw_256.sw_in1k) |84.63|97.06| 575.53| 66.01| 14.67| 58.38|
|
| 201 |
+
|[maxxvitv2_rmlp_base_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/maxxvitv2_rmlp_base_rw_224.sw_in12k_ft_in1k) |86.64|98.02| 501.03| 116.09| 24.20| 62.77|
|
| 202 |
+
|[maxvit_rmlp_base_rw_224.sw_in12k_ft_in1k](https://huggingface.co/timm/maxvit_rmlp_base_rw_224.sw_in12k_ft_in1k) |86.89|98.02| 375.86| 116.14| 23.15| 92.64|
|
| 203 |
+
|[maxvit_base_tf_224.in1k](https://huggingface.co/timm/maxvit_base_tf_224.in1k) |84.85|96.99| 358.25| 119.47| 24.04| 95.01|
|
| 204 |
+
|[maxvit_tiny_tf_384.in1k](https://huggingface.co/timm/maxvit_tiny_tf_384.in1k) |85.11|97.38| 293.46| 30.98| 17.53| 123.42|
|
| 205 |
+
|[maxvit_large_tf_224.in1k](https://huggingface.co/timm/maxvit_large_tf_224.in1k) |84.93|96.97| 247.71| 211.79| 43.68| 127.35|
|
| 206 |
+
|[maxvit_small_tf_384.in1k](https://huggingface.co/timm/maxvit_small_tf_384.in1k) |85.54|97.46| 188.35| 69.02| 35.87| 183.65|
|
| 207 |
+
|[coatnet_rmlp_2_rw_384.sw_in12k_ft_in1k](https://huggingface.co/timm/coatnet_rmlp_2_rw_384.sw_in12k_ft_in1k) |87.39|98.31| 160.80| 73.88| 47.69| 209.43|
|
| 208 |
+
|[maxxvitv2_rmlp_base_rw_384.sw_in12k_ft_in1k](https://huggingface.co/timm/maxxvitv2_rmlp_base_rw_384.sw_in12k_ft_in1k) |87.47|98.37| 149.49| 116.09| 72.98| 213.74|
|
| 209 |
+
|[maxvit_tiny_tf_512.in1k](https://huggingface.co/timm/maxvit_tiny_tf_512.in1k) |85.67|97.58| 144.25| 31.05| 33.49| 257.59|
|
| 210 |
+
|[maxvit_rmlp_base_rw_384.sw_in12k_ft_in1k](https://huggingface.co/timm/maxvit_rmlp_base_rw_384.sw_in12k_ft_in1k) |87.81|98.37| 106.55| 116.14| 70.97| 318.95|
|
| 211 |
+
|[maxvit_base_tf_384.in21k_ft_in1k](https://huggingface.co/timm/maxvit_base_tf_384.in21k_ft_in1k) |87.92|98.54| 104.71| 119.65| 73.80| 332.90|
|
| 212 |
+
|[maxvit_base_tf_384.in1k](https://huggingface.co/timm/maxvit_base_tf_384.in1k) |86.29|97.80| 101.09| 119.65| 73.80| 332.90|
|
| 213 |
+
|[maxvit_small_tf_512.in1k](https://huggingface.co/timm/maxvit_small_tf_512.in1k) |86.10|97.76| 88.63| 69.13| 67.26| 383.77|
|
| 214 |
+
|[maxvit_large_tf_384.in21k_ft_in1k](https://huggingface.co/timm/maxvit_large_tf_384.in21k_ft_in1k) |87.98|98.56| 71.75| 212.03|132.55| 445.84|
|
| 215 |
+
|[maxvit_large_tf_384.in1k](https://huggingface.co/timm/maxvit_large_tf_384.in1k) |86.23|97.69| 70.56| 212.03|132.55| 445.84|
|
| 216 |
+
|[maxvit_base_tf_512.in21k_ft_in1k](https://huggingface.co/timm/maxvit_base_tf_512.in21k_ft_in1k) |88.20|98.53| 50.87| 119.88|138.02| 703.99|
|
| 217 |
+
|[maxvit_base_tf_512.in1k](https://huggingface.co/timm/maxvit_base_tf_512.in1k) |86.60|97.92| 50.75| 119.88|138.02| 703.99|
|
| 218 |
+
|[maxvit_xlarge_tf_384.in21k_ft_in1k](https://huggingface.co/timm/maxvit_xlarge_tf_384.in21k_ft_in1k) |88.32|98.54| 42.53| 475.32|292.78| 668.76|
|
| 219 |
+
|[maxvit_large_tf_512.in21k_ft_in1k](https://huggingface.co/timm/maxvit_large_tf_512.in21k_ft_in1k) |88.04|98.40| 36.42| 212.33|244.75| 942.15|
|
| 220 |
+
|[maxvit_large_tf_512.in1k](https://huggingface.co/timm/maxvit_large_tf_512.in1k) |86.52|97.88| 36.04| 212.33|244.75| 942.15|
|
| 221 |
+
|[maxvit_xlarge_tf_512.in21k_ft_in1k](https://huggingface.co/timm/maxvit_xlarge_tf_512.in21k_ft_in1k) |88.53|98.64| 21.76| 475.77|534.14|1413.22|
|
| 222 |
+
|
| 223 |
+
## Citation
|
| 224 |
+
```bibtex
|
| 225 |
+
@misc{rw2019timm,
|
| 226 |
+
author = {Ross Wightman},
|
| 227 |
+
title = {PyTorch Image Models},
|
| 228 |
+
year = {2019},
|
| 229 |
+
publisher = {GitHub},
|
| 230 |
+
journal = {GitHub repository},
|
| 231 |
+
doi = {10.5281/zenodo.4414861},
|
| 232 |
+
howpublished = {\url{https://github.com/rwightman/pytorch-image-models}}
|
| 233 |
+
}
|
| 234 |
+
```
|
| 235 |
+
```bibtex
|
| 236 |
+
@article{tu2022maxvit,
|
| 237 |
+
title={MaxViT: Multi-Axis Vision Transformer},
|
| 238 |
+
author={Tu, Zhengzhong and Talebi, Hossein and Zhang, Han and Yang, Feng and Milanfar, Peyman and Bovik, Alan and Li, Yinxiao},
|
| 239 |
+
journal={ECCV},
|
| 240 |
+
year={2022},
|
| 241 |
+
}
|
| 242 |
+
```
|
| 243 |
+
```bibtex
|
| 244 |
+
@article{dai2021coatnet,
|
| 245 |
+
title={CoAtNet: Marrying Convolution and Attention for All Data Sizes},
|
| 246 |
+
author={Dai, Zihang and Liu, Hanxiao and Le, Quoc V and Tan, Mingxing},
|
| 247 |
+
journal={arXiv preprint arXiv:2106.04803},
|
| 248 |
+
year={2021}
|
| 249 |
+
}
|
| 250 |
+
```
|
config.json
CHANGED
|
@@ -4,6 +4,7 @@
|
|
| 4 |
"num_features": 1024,
|
| 5 |
"global_pool": "avg",
|
| 6 |
"pretrained_cfg": {
|
|
|
|
| 7 |
"custom_load": false,
|
| 8 |
"input_size": [
|
| 9 |
3,
|
|
@@ -26,8 +27,8 @@
|
|
| 26 |
],
|
| 27 |
"num_classes": 1000,
|
| 28 |
"pool_size": [
|
| 29 |
-
|
| 30 |
-
|
| 31 |
],
|
| 32 |
"first_conv": "stem.conv1",
|
| 33 |
"classifier": "head.fc"
|
|
|
|
| 4 |
"num_features": 1024,
|
| 5 |
"global_pool": "avg",
|
| 6 |
"pretrained_cfg": {
|
| 7 |
+
"tag": "in21k_ft_in1k",
|
| 8 |
"custom_load": false,
|
| 9 |
"input_size": [
|
| 10 |
3,
|
|
|
|
| 27 |
],
|
| 28 |
"num_classes": 1000,
|
| 29 |
"pool_size": [
|
| 30 |
+
12,
|
| 31 |
+
12
|
| 32 |
],
|
| 33 |
"first_conv": "stem.conv1",
|
| 34 |
"classifier": "head.fc"
|