LearnablePromptSAM

LearnablePromptSAM copied to clipboard

Is it possible to provide pre-training weights?

4

Since I'm not using my own training set at the moment and only have single-digit images for testing, could you please provide me with your pre-trained weights so that I...

zxcvbnmkj

problem

7

After dividing the pixels of the binary image by 255 and setting num_class to 2 for training, I encountered an error when loading the model for segmentation：Missing key(s) in state_dict:...

liuadan

train on dataset iDRiD,iou is very small

1

您好，我在iDRiD数据集，使用代码进行复现，训练过程中，iou很小且一直下降，使用权重进行预测，出来的图是全黑的，完全没有预测正确，请问这是什么问题呢？谢谢！可以提供一下训练这个数据集的相关参数吗？

liyan20180808

Excuse me, which images are to be trained for --image and --mask_path?

2

Hello, author. I am trying to train the AROI dataset using your model. I set the paths for the original images and label images as --image AROIonline/inter_intra/raw and --mask_path AROIonline/inter_intra/groundTruth,...

zys1999

About one-shot and zero shot learning

1

Hi! As reported in the technical report, promptSAM utilized one shot and zero-shot learning. What is the implementation detail? To be more specific, which data are used for training? And...

ghost

Is xFormers a must?

9

Thanks for your great work! I lost a lot at the beginning of the training and the dice metric is approaching 0.At the same time, I use the same dataset...

Heirudy

GPU memory requirement

3

Thanks for the great work! I will be replicating your approach to fine tune SAM on custom few shot medical images. The problem is I have only 8GB GPU memory...

Raspberry-beans

how is num_class decided?

5

I saw few threads in the issues but I am still a bit confused with the class, so I am training the model to do vein segmentation and the mask...

krishnaadithya

cuDNN error

12

 which version of pytorch is used, could you share the package lists?

liuyanice

PromptSAM input Image_size

3

I run learnerable_seg.py and want to use model = PromptSAM("vit_b", "ckpts/sam_vit_b_01ec64.pth").half().cuda() x = torch.randn(1, 3, 518, 518).half().cuda(), But mistakes can occur x = x + self.pos_embed RuntimeError: The size of...

dxw2000