We are seeing reports of well-established workflows requesting GPUs failing with an error similar to the following.
Task <task_name> failed. The job was stopped before the command finished. PAPI error code 2. Execution failed: generic::unknown: installing drivers: container exited with unexpected exit code 1
This error message is followed by a long string of information regarding errors with the GPU NVIDIA driver installation.
We've reported this issue to Google. They have confirmed that they are seeing the same issue on their end, and are currently investigating. All updates regarding this error will be posted to this thread.
Please sign in to leave a comment.