标签:

• clip 库名称: open_clip 管道标签: 零样本图像分类 许可证: apache-2.0 数据集:
• mlfoundations/datacomp_1b

ViT-bigG-14-CLIPA-336-datacomp1B 模型卡片

一个 CLIPA-v2 模型...

模型详情

• 模型类型: 对比式图文模型，零样本图像分类。
• 原始来源: https://github.com/UCSC-VLAA/CLIPA
• 数据集: mlfoundations/datacomp_1b
• 相关论文:
  • CLIPA-v2: 以1万美元预算实现81.1%的零样本ImageNet准确率；额外4千美元解锁81.8%准确率: https://arxiv.org/abs/2306.15658
  • CLIP训练的逆缩放定律: https://arxiv.org/abs/2305.07017

模型使用

使用OpenCLIP

import torch
import torch.nn.functional as F
from urllib.request import urlopen
from PIL import Image
from open_clip import create_model_from_pretrained, get_tokenizer

model, preprocess = create_model_from_pretrained('hf-hub:ViT-bigG-14-CLIPA-336')
tokenizer = get_tokenizer('hf-hub:ViT-bigG-14-CLIPA-336')

image = Image.open(urlopen(
    'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))
image = preprocess(image).unsqueeze(0)

text = tokenizer(["一张图表", "一只狗", "一只猫", "一个贝涅饼"], context_length=model.context_length)

with torch.no_grad(), torch.cuda.amp.autocast():
    image_features = model.encode_image(image)
    text_features = model.encode_text(text)
    image_features = F.normalize(image_features, dim=-1)
    text_features = F.normalize(text_features, dim=-1)

    text_probs = (100.0 * image_features @ text_features.T).softmax(dim=-1)

print("标签概率:", text_probs)  # 输出: [[0., 0., 0., 1.0]]

引用

@article{li2023clipav2,
      title={CLIPA-v2: Scaling CLIP Training with 81.1% Zero-shot ImageNet Accuracy within a $10,000 Budget; An Extra $4,000 Unlocks 81.8% Accuracy}, 
      author={Xianhang Li and Zeyu Wang and Cihang Xie},
      journal={arXiv preprint arXiv:2306.15658},
      year={2023},
}

@inproceedings{li2023clipa,
      title={An Inverse Scaling Law for CLIP Training}, 
      author={Xianhang Li and Zeyu Wang and Cihang Xie},
      booktitle={NeurIPS},
      year={2023},
}