分享如何解决 RuntimeError: Found NVIDIA GeForce GTX 1060 which is too old to be supported by the triton GPU compiler, which is used as the backend. Triton only supports devices of CUDA Capability >= 7.0, but your device is of CUDA capability 6.1

[94zuocom]

打开core.py
import torch._dynamo
torch._dynamo.config.suppress_errors = True

[mw66]

I added torch._dynamo.config.suppress_errors = True, but the message still showing.

However, this is just WARNING:

[2024-06-18 23:38:49,504] torch._dynamo.convert_frame: [WARNING]     raise RuntimeError(
[2024-06-18 23:38:49,504] torch._dynamo.convert_frame: [WARNING] torch._dynamo.exc.BackendCompilerFailed: backend='inductor' raised:
[2024-06-18 23:38:49,504] torch._dynamo.convert_frame: [WARNING] RuntimeError: Found NVIDIA GeForce GTX TITAN X which is too old to be supported by the triton GPU compiler, which is used as the backend. Triton only supports devices of CUDA Capability >= 7.0, but your device is of CUDA capability 

The training seems working despite such messages. So I assume it's safe to just ignore these messages?

[github-actions]

This issue was closed because it has been inactive for 15 days since being marked as stale.