Andy718811

Below is my code. when I test my model with detect.py, the precision is always up tp 0.98, but when I test it with the following code, taking the image...

How to train vit_tiny (Light HQ-SAM for real-time need): ViT-Tiny HQ-SAM model?

4

I tried to train the vit_tiny with tis argument, "python -m torch.distributed.launch --nproc_per_node=1 train.py --checkpoint ./pretrained_checkpoint/sam_hq_vit_tiny.pth --model-type vit_b --output work_dirs/hq_sam_tiny_l", but faced this problem, it seems tain.py can't be used...

Here is my code, the result is not very well. ``` import json from PIL import Image, ImageDraw import matplotlib.pyplot as plt import fitz # pip install pymupdf import json...