search

Qsingle / LearnablePromptSAM

Try to use the SAM-ViT as the backbone to create the learnable prompt for semantic segmentation
Apache License 2.0
77 stars 13 forks source link

cuDNN error #12

Open liuyanice opened 1 year ago

liuyanice commented 1 year ago

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

Qsingle commented 1 year ago

Good issue, we use torch 1.13.1+CU117. It seems that the model or data can not be feed into the GPU. I suggest you check the version of CUDA.

liuyanice commented 1 year ago

my issue： torch 1.13.1+CU117 nvidia-smi: CUDA Version:11.7 Driver Version:515.105.01 nvcc: cuda_11.3.r11.3

cuDNN error: CUDNN_STATUS_NOT_INITIALIZED

Qsingle commented 1 year ago

Can you tell me the GPU you use? The version of cudnn. If you use Anaconda or Minianaconda I suggest you to create one virtual environment by the following command. Or you can install the CUDA 11.7 and Cudnn 8.5.0.

conda create -n pytorch python=3.8 pytorch==1.13.1 torchvision==0.14.1 torchaudio==0.13.1 pytorch-cuda=11.7 -c pytorch -c nvidia
conda activate pytorch #activate the enviroment
liuyanice commented 12 months ago

RTX TI3090 single GPU, can you tell me the GPU you use?

liuyanice commented 12 months ago

image

liuyanice commented 12 months ago

image2

liuyanice commented 12 months ago

image3

Qsingle commented 12 months ago

I know the reason for this error. Similar to #6, your label may contain an invalid value like 255.

liuyanice commented 12 months ago

Can you share your datasets, I want to reproduce the results.

liuyanice commented 12 months ago

Beacause I have set num classes is 2 and used the blood vessel datasets. It also displays "cudnn error".

liuyanice commented 12 months ago

Thank you, I have already solved this problem.

liuyanice commented 12 months ago

It is indeed caused by an invalid value