baoyue.shen
baoyue.shen
What is the maximum CUDA version supported by the code? My CUDA version is a bit high and cannot install the version mentioned in readme. Has anyone encountered this problem?...
May I ask if anyone encountered this error while running the file 'dist_train. sh'? I haven't resolved it yet. I don't understand why I made an error before assigning values...
Have you encountered this error while running the training file? He can automatically complete, but he always reports errors. Friends who have successfully run the training program, can you tell...
File "/mnt/baoyue-shen-ssd/root/dataset-system/miniforge3/envs/depth_anything_ac/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1786, in _call_impl return forward_call(*args, **kwargs) File "/mnt/baoyue-shen-ssd/root/dataset-system/baoyue.shen/Code/dinov3/dinov3/layers/block.py", line 205, in forward return self._forward_list([x_or_x_list], rope_list=[rope_or_rope_list])[0] File "/mnt/baoyue-shen-ssd/root/dataset-system/baoyue.shen/Code/dinov3/dinov3/layers/block.py", line 193, in _forward_list x_attn = x + self.ls1(self.attn(self.norm1(x), rope=rope))...