This repository includes codes for Japanese CLIP (Contrastive Language-Image Pre-Training) variants by rinna Co., Ltd.
Table of Contents |
---|
News |
Pretrained Models |
Usage |
Citation |
License |
v0.2.0 was released!
- Both CLIP and CLOOB models were upgraded! Now,
rinna/japanese-cloob-vit-b-16
achieves 54.64. - Released our Japanese prompt templates and an example code (see
scripts/example.py
) for zero-shot ImageNet classification. Those templates were cleaned for Japanese based on the OpenAI 80 templates. - Changed the citation
Model Name | TOP1* | TOP5* |
---|---|---|
rinna/japanese-cloob-vit-b-16 | 54.64 | 72.86 |
rinna/japanese-clip-vit-b-16 | 50.69 | 72.35 |
sonoisa/clip-vit-b-32-japanese-v1 | 38.88 | 60.71 |
multilingual-CLIP | 14.36 | 27.28 |
*Zero-shot ImageNet validation set top-k accuracy.
- Install package
$ pip install git+https://github.com/rinnakk/japanese-clip.git
- Run
from PIL import Image
import torch
import japanese_clip as ja_clip
device = "cuda" if torch.cuda.is_available() else "cpu"
# ja_clip.available_models()
# ['rinna/japanese-clip-vit-b-16', 'rinna/japanese-cloob-vit-b-16']
# If you want v0.1.0 models, set `revision='v0.1.0'`
model, preprocess = ja_clip.load("rinna/japanese-clip-vit-b-16", cache_dir="/tmp/japanese_clip", device=device)
tokenizer = ja_clip.load_tokenizer()
image = preprocess(Image.open("./data/dog.jpeg")).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]]
To cite this repository:
@inproceedings{japanese-clip,
author = {シーン 誠, 趙 天雨, 沢田 慶},
title = {日本語における言語画像事前学習モデルの構築と公開},
booktitle= {The 25th Meeting on Image Recognition and Understanding},
year = 2022,
month = July,
}