When attempting to launch a Task via Service, I keep getting the following error:
During cluster create, I used a1.medium (arm64 – since I built the image locally on my mac M2 which is arm64)…
I tried using Amazon Linux 2 (arm64), also bottlerocket arm64 as well … no luck (same error)
This error is specifically about GPU resources … On the EC2 instance for this cluster, I can’t seem to see any info about GPU resources for the instance (?)
Why would a simple python flask app require GPU resources?