FastSAM issues

using text promp to get masks for all same classes not one mask

使用text prompt模式时我希望能够分割出图像中所有指定类别的物体，比如我的提示词是“dog“时，我希望能够分割出图像中所有的dog，而fastsam只分割出其中一条狗，即使我更换提示词为”all dogs“时也不起作用。 When using the text prompt mode, I hope to be able to segment all objects of the specified category in the image. For example, when my prompt...

limmax

Added clip import in prompt.py

# Issue solved : - clip not found during text prompt inference. # Changes made : - Added import statement for clip in fastsam/prompt.py

Arjun-A-I

User control for mask alpha

Added parameter for mask plotting functions to set the mask alpha dynamically. Defaults to 0.6 which was originally set statically

AmoghDhaliwal

Why isn't the model I trained working???

2

I use the training code `from ultralytics import YOLO if __name__ == '__main__': model = YOLO(model="FastSAM-x.pt") model.train(data="aaa.yaml", epochs=200, batch=32, imgsz=1024, overlap_mask=False, save=True, save_period=5, device=[0,1], project='fastsam', name='test', val=False)` Is it normal...

xpbag

为什么我一轮训练完之后没有输出MAP而是直接进入了下一轮

![image](https://github.com/user-attachments/assets/037907a3-9a86-465d-be6d-1f4e7bdccbc5) 我并没有白色的这个框框里面显示的，而是直接进入了下一轮训练，这个是我的问题还是代码这里就没输出呀

Sjjnb666

is prompt_process.everything_prompt() exectly yolov8?

is prompt_process.everything_prompt() exectly yolov8? i don't find any about SAM module when inference with prompt_process.everything_prompt() branch.

cqray1990

用gpu cuda跑大一点的图片

1

``` everything_results = model(input_image, device="cuda:0", retina_masks=args.retina, imgsz=args.imgsz, conf=args.conf, iou=args.iou) print("model loaded " , everything_results) ``` 用gpu cuda跑大一点的图片 everything_results 直接 None了，小图片不会有问题，大家有遇到这个情况吗？我用gpu就是为了切大图片呀。日志： model loaded None prompt_process Executing took 6.141...

alexhmyang