The blip-image-captioning model can generate captions for an input image.
cargo run --example blip --release -- --image candle-examples/examples/yolo-v8/assets/bike.jpgRunning on CPU, to run on GPU, build this example with `--features cuda`
loaded image Tensor[dims 3, 384, 384; f32]
model built
several cyclists are riding down a road with cars behind them%
