search

jackyjsy / CVPR21Chal-SLR

This repo contains the official code of our work SAM-SLR which won the CVPR 2021 Challenge on Large Scale Signer Independent Isolated Sign Language Recognition.
Creative Commons Zero v1.0 Universal
205 stars 51 forks source link

CUDA capability is not compatible with the current PyTorch installation #20

Closed HajarAlnamshan closed 2 years ago

HajarAlnamshan commented 2 years ago

Hello Initially, I would thank the author for this amazing repository

I following step by step of this instruction [reproduce.md](https://github.com/jackyjsy/CVPR21Chal-SLR/blob/main/reproduce.md)

I'm at > III: Procedues to reproduce our results > A. For RGB track

and I was struggling with this error since yesterday, it happen when I try to run python main.py --config config/sign/test/test_bone.yaml

UserWarning: 
NVIDIA RTX A5000 with CUDA capability sm_86 is not compatible with the current PyTorch installation.
The current PyTorch install supports CUDA capabilities sm_37 sm_50 sm_60 sm_61 sm_70 sm_75 compute_37.
If you want to use the NVIDIA RTX A5000 GPU with
<img width="970" alt="Screen Shot 1443-08-07 at 12 45 11 PM" src="https://user-images.githubusercontent.com/85832533/157636654-90100cea-ca2d-40e7-ae36-7a243278a6be.png">
 PyTorch, please check the instructions at https://pytorch.org/get-started/locally/

Any help? Thank you

LYF199903 commented 1 year ago

You should change the device to 0 in ‘config’ this folder. This is a multi-card parallel program.