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/eva02_large_patch14_448.mim_m38m_ft_in22k_in1k
---

# Anime Tagger eva02_large_patch14_448.dbv4-full

## Model Details

- **Model Type:** Multilabel Image classification / feature backbone
- **Model Stats:**
  - Params: 316.8M
  - FLOPs / MACs: 620.9G / 310.1G
  - Image size: train = 448 x 448, test = 448 x 448
- **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

|     #      |    Macro@0.40 (F1/MCC/P/R)    |    Micro@0.40 (F1/MCC/P/R)    |  Macro@Best (F1/P/R)  |
|:----------:|:-----------------------------:|:-----------------------------:|:---------------------:|
| Validation | 0.570 / 0.573 / 0.600 / 0.557 | 0.693 / 0.692 / 0.690 / 0.696 |          ---          |
|    Test    | 0.569 / 0.573 / 0.600 / 0.556 | 0.693 / 0.693 / 0.691 / 0.696 | 0.599 / 0.600 / 0.618 |

* `Macro/Micro@0.40` 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)   |  Macro@0.40 (F1/P/R)  |  Macro@Best (F1/P/R)  |
|:----------:|:---------:|:-------:|:-----------:|:---------------------:|:---------------------:|:---------------------:|
|     0      |  general  |    1    |    0.39     | 0.681 / 0.674 / 0.688 | 0.445 / 0.485 / 0.427 | 0.480 / 0.476 / 0.510 |
|     4      | character |    1    |    0.61     | 0.943 / 0.961 / 0.925 | 0.921 / 0.925 / 0.920 | 0.938 / 0.954 / 0.924 |
|     9      |  rating   |    1    |    0.38     | 0.832 / 0.801 / 0.865 | 0.838 / 0.817 / 0.860 | 0.839 / 0.819 / 0.861 |

* `Micro@Thr` means the metrics on the category-level suggested thresholds, which are listed in the table above.
* `Macro@0.40` 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/eva02_large_patch14_448.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/eva02_large_patch14_448.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/eva02_large_patch14_448.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=(448, 448), interpolation=bicubic, max_size=None, antialias=True)
#     CenterCrop(size=[448, 448])
#     MaybeToTensor()
#     Normalize(mean=tensor([0.4815, 0.4578, 0.4082]), std=tensor([0.2686, 0.2613, 0.2758]))
# )

image = load_image('https://huggingface.co/animetimm/eva02_large_patch14_448.dbv4-full/resolve/main/sample.webp')
input_ = preprocessor(image).unsqueeze(0)
# input_, shape: torch.Size([1, 3, 448, 448]), 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())))
# {'sensitive': 0.9555495381355286,
#  '1girl': 0.9977720379829407,
#  'solo': 0.9800751209259033,
#  'looking_at_viewer': 0.7236320972442627,
#  'blush': 0.7710952758789062,
#  'smile': 0.8856169581413269,
#  'short_hair': 0.803878128528595,
#  'long_sleeves': 0.3804128170013428,
#  'brown_hair': 0.6562796831130981,
#  'dress': 0.5758444666862488,
#  'sitting': 0.7712022066116333,
#  'purple_eyes': 0.5440564751625061,
#  'flower': 0.9287881851196289,
#  'braid': 0.8394284844398499,
#  'tears': 0.778815746307373,
#  'floral_print': 0.43895024061203003,
#  'plant': 0.6179906725883484,
#  'blue_flower': 0.30160021781921387,
#  'crown_braid': 0.40592360496520996,
#  'potted_plant': 0.5879666209220886,
#  'flower_pot': 0.49822214245796204,
#  'wiping_tears': 0.4761575758457184}
```
### 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/eva02_large_patch14_448.dbv4-full/resolve/main/sample.webp',
    repo_id='animetimm/eva02_large_patch14_448.dbv4-full',
    fmt=('general', 'character', 'rating'),
)

print(general)
# {'1girl': 0.9977719783782959,
#  'solo': 0.9800750613212585,
#  'flower': 0.9287877082824707,
#  'smile': 0.8856177926063538,
#  'braid': 0.8394323587417603,
#  'short_hair': 0.8038788437843323,
#  'tears': 0.7787976264953613,
#  'sitting': 0.7712044715881348,
#  'blush': 0.7710968255996704,
#  'looking_at_viewer': 0.7236329317092896,
#  'brown_hair': 0.6562790870666504,
#  'plant': 0.6180056929588318,
#  'potted_plant': 0.5879812836647034,
#  'dress': 0.5758441686630249,
#  'purple_eyes': 0.5440553426742554,
#  'flower_pot': 0.4982312321662903,
#  'wiping_tears': 0.47614389657974243,
#  'floral_print': 0.43895548582077026,
#  'crown_braid': 0.40593117475509644,
#  'long_sleeves': 0.3804135322570801,
#  'blue_flower': 0.3015919327735901}
print(character)
# {}
print(rating)
# {'sensitive': 0.9555498361587524}
```

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).