File size: 1,553 Bytes
9d5157a
a3ec9ad
 
605b122
 
a3ec9ad
 
 
 
 
 
9d5157a
a3ec9ad
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
---
language: ja
thumbnail: https://github.com/rinnakk/japanese-pretrained-models/blob/master/rinna.png
license: apache-2.0
tasks: Feature Extraction
tags:
- ja
- japanese
- clip
- cloob
- vision
---

# rinna/japanese-cloob-vit-b-16

![rinna-icon](./rinna.png)

This repository provides a Japanese [CLOOB (Contrastive Leave One Out Boost)](https://arxiv.org/abs/2110.11316) model. The model was trained by [rinna Co., Ltd.](https://corp.rinna.co.jp/)

# How to use the model


1. Install package

```shell
$ pip install git+https://github.com/rinnakk/japanese-clip.git
```

2. Run

```python
import io
import requests
from PIL import Image
import torch
import japanese_clip as ja_clip

device = "cuda" if torch.cuda.is_available() else "cpu"


model, preprocess = ja_clip.load("rinna/japanese-cloob-vit-b-16", device=device)
tokenizer = ja_clip.load_tokenizer()

img = Image.open(io.BytesIO(requests.get('https://images.pexels.com/photos/2253275/pexels-photo-2253275.jpeg?auto=compress&cs=tinysrgb&dpr=3&h=750&w=1260').content))
image = preprocess(img).unsqueeze(0).to(device)
encodings = ja_clip.tokenize(
    texts=["犬", "猫", "象"],
    max_seq_len=77,
    device=device,
    tokenizer=tokenizer, # this is optional. if you don't pass, load tokenizer each time
)

with torch.no_grad():
    image_features = model.get_image_features(image)
    text_features = model.get_text_features(**encodings)
    
    text_probs = (100.0 * image_features @ text_features.T).softmax(dim=-1)

print("Label probs:", text_probs)  # prints: [[1.0, 0.0, 0.0]]
```