qformer 0.0.5

qformer - Pytorch

Qformer

Implementation of Qformer from BLIP2 in Zeta Lego blocks. The implementation is here straight from Figure 2. In particular the image block and text block.

Install

pip3 install qformer

Usage

import torch
from qformer import QFormer

x = torch.randn(
    1, 32, 512
)  # Create a random tensor of shape (1, 32, 512)

img = torch.randn(
    1, 32, 512
)  # Create another random tensor of shape (1, 32, 512)

qformer = QFormer(
    512, 8, 8, 0.1, 2, 2
)  # Create an instance of the QFormer model

y = qformer(
    x, img
)  # Apply the QFormer model to the input tensors x and img

print(y.shape)  # Print the shape of the output tensor y

License

MIT

Citation

@misc{li2023blip2,
    title={BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models}, 
    author={Junnan Li and Dongxu Li and Silvio Savarese and Steven Hoi},
    year={2023},
    eprint={2301.12597},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}