Closed AxuJIuyc closed 1 year ago
Hi @AxuJIuyc, thanks for your interest in our work.
The error is probably related to an issue with your GPU. Could you provide the following information?
Also, could you look at this issue and check if reducing the maximum number of threads might fix the problem?
If nothing works, please share the complete trackback for the error.
PyTorch==1.10.1, CUDA==11.1 GPU - Nvidia Tesla V100
The required file (VolumetricUpSamplingTrilinear.cu), as in your link, is not in my pytorch directory for some reason. I'll try to reinstall or checking out different versions.
I created a new clean environment, carefully installed all the libraries with the same versions - the oneformer was launched! And earlier I received a lot of different errors in the style of "directory/file not found". The versions didn't match, maybe. ¯_(ツ)_/¯
Hello! I am trying to run a demo for a single image. I use "oneformer_dinat_large_IN21k_384_bs16_160k.yaml" as config file and "250_16_dinat_l_oneformer_ade20k_160k.pth" as model weights. When I run the demo.py, I see the following lines: "error in ms_deformable_im2col_cuda: too many resources requested for launch" and the code ends up saving {task}.jpg without significant information. The error string occurs when executing the timeline code in demo/defaults.py: line 81 "predictions = self.model([inputs])[0]"