Stuck at (match_worker_ray_wrapper pid=32734) Global seed set to 6660/725 [00:00<?, ?it/s] (match_worker_ray_wrapper pid=32736) Global seed set to 666 (match_worker_ray_wrapper pid=32735) Global seed set to 666 (match_worker_ray_wrapper pid=32765) Global seed set to 666

[brian2lee]

As the title mention, I got stuck when running bash scripts/demo_pipeline.sh $OBJ_NAME. After trying the methods mention in this issue it got through but showed a new error (match_worker_ray_wrapper pid=27622) NVIDIA GeForce RTX 3060 Laptop GPU with CUDA capability sm_86 is not compatible with the current PyTorch installation. in an attempt to solve it I install other version of pytorch usingmamba install -c pytorch -c conda-forge pytorch=1.10.0 torchvision cudatoolkit=11.3 --yes. Which got me back to stuck at Global seed 666.
In conclusion, either I''ll got stuck on Global seed 666 or CUDA capability. Any way to solve this problem?

[lejy026]

Do you solve it ?

[lejy026]

This issue maybe can be solved by add weight/LoFTR_wsize9.ckpt

[anshium]

#5