metadata
language: ja
thumbnail: https://github.com/rinnakk/japanese-pretrained-models/blob/master/rinna.png
license: apache-2.0
tags:
- feature-extraction
- ja
- japanese
- clip
- vision
rinna/japanese-clip-vit-b-16
This is a Japanese CLIP (Contrastive Language-Image Pre-Training) model trained by rinna Co., Ltd..
Please see japanese-clip for the other available models.
How to use the model
- Install package
$ pip install git+https://github.com/rinnakk/japanese-clip.git
- Run
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-clip-vit-b-16", cache_dir="/tmp/japanese_clip", 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]]
Model architecture
The model was trained a ViT-B/16 Transformer architecture as an image encoder and uses a 12-layer BERT as a text encoder. The image encoder was initialized from the AugReg vit-base-patch16-224
model.
Training
The model was trained on CC12M translated the captions to Japanese.