You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When I use it by docker, have the error: 'GPU with CUDA capability 7 0 is not supported' And it is my GPU CUDA info:
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 525.60.13 Driver Version: 525.60.13 CUDA Version: 12.0 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 Tesla V100-PCIE... Off | 00000000:04:00.0 Off | 0 |
| N/A 37C P0 36W / 250W | 0MiB / 32768MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
| 1 Tesla V100-PCIE... Off | 00000000:0C:00.0 Off | 0 |
| N/A 36C P0 35W / 2
The text was updated successfully, but these errors were encountered:
When I use it by docker, have the error: 'GPU with CUDA capability 7 0 is not supported' And it is my GPU CUDA info:
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 525.60.13 Driver Version: 525.60.13 CUDA Version: 12.0 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 Tesla V100-PCIE... Off | 00000000:04:00.0 Off | 0 |
| N/A 37C P0 36W / 250W | 0MiB / 32768MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
| 1 Tesla V100-PCIE... Off | 00000000:0C:00.0 Off | 0 |
| N/A 36C P0 35W / 2
The text was updated successfully, but these errors were encountered: