karthikdatta98
karthikdatta98
@DhruvAwasthi Hey, I have one query, did you find a way how to get the mask which is equivalent to the ground truth? are there any config parameters to change?...
@DhruvAwasthi Hey Thanks for the quick reply. Wanted to know as SAM was pre-trained on 1024x1024, and if you used some other image resolution if it hurt the model predictions....
@DhruvAwasthi Thats great to know, are you open to share if your dataset worked well on SAM without finetuning? Did it have a drastic change after fine tuning? I am...
@DhruvAwasthi when you finetuned it with 1200x1200, your groundtruth was 256x256? or you upscaled the predictions to 1200x1200 keeping the gorundtruth also at 1200x1200?
@DhruvAwasthi, can you share your code for inference? like the main function. either here or at [email protected] Thank you so much, it would be a great help, am currently working...
@DhruvAwasthi also, after I did, torch.save, how do I load my model.pth? what was the command you used? thanks.
Does sahi work on non axis oriented yolo versions? Like obb
@NielsRogge @sushmanthreddy Can I know how big should the data set be? would 100 examples be enough? also as the pre-trained model was trained on 1024x1024, do I need to...