Gradio SAM demo offering four prompting modes: point-prompted, box-prompted, mixed point and box prompts, and automatic full image segmentation. Supports: PyTorch, ONNX, and TensorRT runtimes Toggling between l0, l1, l2, xl0, and xl1 models Experimentation with full image segmentation parameters...
trtexec --onnx=assets/export_models/sam/onnx/xl1_decoder.onnx --minShapes=point_coords:1x1x2,point_labels:1x1 --optShapes=point_coords:16x2x2,point_labels:16x2 --maxShapes=point_coords:16x2x2,point_labels:16x2 --fp16 --saveEngine=assets/export_models/sam/tensorrt/xl1_decoder.engine ...