---
title: tresnet_xl.miil_in1k
canonical_url: "https://www.modelscope.cn/models/timm/tresnet_xl.miil_in1k"
md_url: "https://www.modelscope.cn/models/timm/tresnet_xl.miil_in1k.md"
repository: timm/tresnet_xl.miil_in1k
last_updated: 2025-01-22
license: apache-2.0
pipeline_tag: image-classification
tasks:
  - image-classification
parameters: 78.6M
tensor_type:
  - I64
  - F32
library_name:
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 1815
stars: 0
tags:
  - image-classification
  - timm
  - transformers
---

# tresnet_xl.miil_in1k

> tresnet_xl.miil_in1k - timm 在 ModelScope 开源的模型。Model card for tresnetxl.miilin1k

timm/tresnet_xl.miil_in1k 是 ModelScope 魔搭社区上的 78.6M 参数image-classification模型，采用 apache-2.0 许可。

- **Repository**: timm/tresnet_xl.miil_in1k
- **License**: apache-2.0
- **Tasks**: image-classification
- **Parameters**: 78.6M
- **Tags**: image-classification, timm, transformers
- **Downloads**: 1815
- **Stars**: 0
- **Last updated**: 2025-01-22

Source: https://www.modelscope.cn/models/timm/tresnet_xl.miil_in1k

---

# Model card for tresnet_xl.miil_in1k

A TResNet image classification model. Trained on ImageNet-1k by paper authors.

The weights for this model have been remapped and modified from the originals to work with standard BatchNorm instead of InplaceABN. `inplace_abn` can be problematic to build recently and ends up slower with `memory_format=channels_last`, torch.compile(), etc.

## Model Details
- **Model Type:** Image classification / feature backbone
- **Model Stats:**
  - Params (M): 78.4
  - GMACs: 15.2
  - Activations (M): 15.3
  - Image size: 224 x 224
- **Papers:**
  - TResNet: High Performance GPU-Dedicated Architecture: https://arxiv.org/abs/2003.13630
- **Dataset:** ImageNet-1k
- **Original:** https://github.com/Alibaba-MIIL/TResNet

## Model Usage
### Image Classification
```python
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('tresnet_xl.miil_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
```python
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(
    'tresnet_xl.miil_in1k',
    pretrained=True,
    features_only=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

for o in output:
    # print shape of each feature map in output
    # e.g.:
    #  torch.Size([1, 83, 56, 56])
    #  torch.Size([1, 166, 28, 28])
    #  torch.Size([1, 1328, 14, 14])
    #  torch.Size([1, 2656, 7, 7])

    print(o.shape)
```

### Image Embeddings
```python
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(
    'tresnet_xl.miil_in1k',
    pretrained=True,
    num_classes=0,  # remove classifier nn.Linear
)
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))  # output is (batch_size, num_features) shaped tensor

# or equivalently (without needing to set num_classes=0)

output = model.forward_features(transforms(img).unsqueeze(0))
# output is unpooled, a (1, 2656, 7, 7) shaped tensor

output = model.forward_head(output, pre_logits=True)
# output is a (1, num_features) shaped tensor
```

## Citation
```bibtex
@misc{ridnik2020tresnet,
    title={TResNet: High Performance GPU-Dedicated Architecture},
    author={Tal Ridnik and Hussam Lawen and Asaf Noy and Itamar Friedman},
    year={2020},
    eprint={2003.13630},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}
```
