README.md

---
tags:
- image-classification
- timm
- transformers
- animetimm
- dghs-imgutils
library_name: timm
license: gpl-3.0
datasets:
- animetimm/danbooru-wdtagger-v4-w640-ws-full
base_model:
- timm/resnet101.tv_in1k
---

# Anime Tagger resnet101.dbv4-full

## Model Details

- **Model Type:** Multilabel Image classification / feature backbone
- **Model Stats:**
  - Params: 68.1M
  - FLOPs / MACs: 46.0G / 22.9G
  - Image size: train = 384 x 384, test = 384 x 384
- **Dataset:** [animetimm/danbooru-wdtagger-v4-w640-ws-full](https://huggingface.co/datasets/animetimm/danbooru-wdtagger-v4-w640-ws-full)
  - Tags Count: 12476
    - General (#0) Tags Count: 9225
    - Character (#4) Tags Count: 3247
    - Rating (#9) Tags Count: 4

## Results

|     #      |    [email protected] (F1/MCC/P/R)    |    [email protected] (F1/MCC/P/R)    |  Macro@Best (F1/P/R)  |
|:----------:|:-----------------------------:|:-----------------------------:|:---------------------:|
| Validation | 0.436 / 0.448 / 0.535 / 0.395 | 0.622 / 0.622 / 0.672 / 0.578 |          ---          |
|    Test    | 0.437 / 0.448 / 0.535 / 0.396 | 0.622 / 0.623 / 0.672 / 0.579 | 0.481 / 0.509 / 0.482 |

* `Macro/[email protected]` means the metrics on the threshold 0.40.
* `Macro@Best` means the mean metrics on the tag-level thresholds on each tags, which should have the best F1 scores.

## Thresholds

|  Category  |   Name    |  Alpha  |  Threshold  |  Micro@Thr (F1/P/R)   |  [email protected] (F1/P/R)  |  Macro@Best (F1/P/R)  |
|:----------:|:---------:|:-------:|:-----------:|:---------------------:|:---------------------:|:---------------------:|
|     0      |  general  |    1    |    0.33     | 0.612 / 0.619 / 0.605 | 0.305 / 0.421 / 0.262 | 0.357 / 0.374 / 0.374 |
|     4      | character |    1    |    0.49     | 0.845 / 0.906 / 0.791 | 0.812 / 0.858 / 0.777 | 0.833 / 0.893 / 0.789 |
|     9      |  rating   |    1    |     0.4     | 0.800 / 0.755 / 0.851 | 0.805 / 0.778 / 0.837 | 0.806 / 0.771 / 0.848 |

* `Micro@Thr` means the metrics on the category-level suggested thresholds, which are listed in the table above.
* `[email protected]` means the metrics on the threshold 0.40.
* `Macro@Best` means the metrics on the tag-level thresholds on each tags, which should have the best F1 scores.

For tag-level thresholds, you can find them in [selected_tags.csv](https://huggingface.co/animetimm/resnet101.dbv4-full/resolve/main/selected_tags.csv).

## How to Use

We provided a sample image for our code samples, you can find it [here](https://huggingface.co/animetimm/resnet101.dbv4-full/blob/main/sample.webp).

### Use TIMM And Torch

Install [dghs-imgutils](https://github.com/deepghs/imgutils), [timm](https://github.com/huggingface/pytorch-image-models) and other necessary requirements with the following command

```shell
pip install 'dghs-imgutils>=0.17.0' torch huggingface_hub timm pillow pandas
```

After that you can load this model with timm library, and use it for train, validation and test, with the following code

```python
import json

import pandas as pd
import torch
from huggingface_hub import hf_hub_download
from imgutils.data import load_image
from imgutils.preprocess import create_torchvision_transforms
from timm import create_model

repo_id = 'animetimm/resnet101.dbv4-full'
model = create_model(f'hf-hub:{repo_id}', pretrained=True)
model.eval()

with open(hf_hub_download(repo_id=repo_id, repo_type='model', filename='preprocess.json'), 'r') as f:
    preprocessor = create_torchvision_transforms(json.load(f)['test'])
# Compose(
#     PadToSize(size=(512, 512), interpolation=bilinear, background_color=white)
#     Resize(size=384, interpolation=bilinear, max_size=None, antialias=True)
#     CenterCrop(size=[384, 384])
#     MaybeToTensor()
#     Normalize(mean=tensor([0.4850, 0.4560, 0.4060]), std=tensor([0.2290, 0.2240, 0.2250]))
# )

image = load_image('https://huggingface.co/animetimm/resnet101.dbv4-full/resolve/main/sample.webp')
input_ = preprocessor(image).unsqueeze(0)
# input_, shape: torch.Size([1, 3, 384, 384]), dtype: torch.float32
with torch.no_grad():
    output = model(input_)
    prediction = torch.sigmoid(output)[0]
# output, shape: torch.Size([1, 12476]), dtype: torch.float32
# prediction, shape: torch.Size([12476]), dtype: torch.float32

df_tags = pd.read_csv(
    hf_hub_download(repo_id=repo_id, repo_type='model', filename='selected_tags.csv'),
    keep_default_na=False
)
tags = df_tags['name']
mask = prediction.numpy() >= df_tags['best_threshold']
print(dict(zip(tags[mask].tolist(), prediction[mask].tolist())))
# {'general': 0.5100178718566895,
#  'sensitive': 0.5034157037734985,
#  '1girl': 0.9962267875671387,
#  'solo': 0.9669082760810852,
#  'looking_at_viewer': 0.8127952814102173,
#  'blush': 0.7912614941596985,
#  'smile': 0.9032713770866394,
#  'short_hair': 0.7837649583816528,
#  'shirt': 0.5146411657333374,
#  'long_sleeves': 0.7224600315093994,
#  'brown_hair': 0.5260339379310608,
#  'holding': 0.5752436518669128,
#  'dress': 0.5642756223678589,
#  'closed_mouth': 0.4826013743877411,
#  'purple_eyes': 0.7590888142585754,
#  'flower': 0.9180877208709717,
#  'braid': 0.9453270435333252,
#  'red_hair': 0.8512048721313477,
#  'blunt_bangs': 0.5289319753646851,
#  'bob_cut': 0.22592417895793915,
#  'plant': 0.5463797450065613,
#  'blue_flower': 0.6992892026901245,
#  'crown_braid': 0.7925195097923279,
#  'potted_plant': 0.5136846899986267,
#  'flower_pot': 0.4357028007507324,
#  'wiping_tears': 0.3059103488922119}
```
### Use ONNX Model For Inference

Install [dghs-imgutils](https://github.com/deepghs/imgutils) with the following command

```shell
pip install 'dghs-imgutils>=0.17.0'
```

Use `multilabel_timm_predict` function with the following code

```python
from imgutils.generic import multilabel_timm_predict

general, character, rating = multilabel_timm_predict(
    'https://huggingface.co/animetimm/resnet101.dbv4-full/resolve/main/sample.webp',
    repo_id='animetimm/resnet101.dbv4-full',
    fmt=('general', 'character', 'rating'),
)

print(general)
# {'1girl': 0.9962266683578491,
#  'solo': 0.96690833568573,
#  'braid': 0.9453268647193909,
#  'flower': 0.9180880784988403,
#  'smile': 0.9032710790634155,
#  'red_hair': 0.8512046337127686,
#  'looking_at_viewer': 0.8127949833869934,
#  'crown_braid': 0.792519211769104,
#  'blush': 0.7912609577178955,
#  'short_hair': 0.7837648391723633,
#  'purple_eyes': 0.7590886354446411,
#  'long_sleeves': 0.7224597930908203,
#  'blue_flower': 0.6992897391319275,
#  'holding': 0.5752434134483337,
#  'dress': 0.5642745494842529,
#  'plant': 0.5463811755180359,
#  'blunt_bangs': 0.5289315581321716,
#  'brown_hair': 0.5260326862335205,
#  'shirt': 0.5146413445472717,
#  'potted_plant': 0.5136858820915222,
#  'closed_mouth': 0.48260119557380676,
#  'flower_pot': 0.4357031583786011,
#  'wiping_tears': 0.30590835213661194,
#  'bob_cut': 0.22592449188232422}
print(character)
# {}
print(rating)
# {'general': 0.5100165009498596, 'sensitive': 0.5034170150756836}
```

For further information, see [documentation of function multilabel_timm_predict](https://dghs-imgutils.deepghs.org/main/api_doc/generic/multilabel_timm.html#multilabel-timm-predict).