Hi,
I'm able to run a certain model on an instance with 32 GiB RAM, but the same model when run with flyte is throwing OOM Error on the same instance type and for the same input parameters. Any thoughts on why this might be happening ?
t
tall-lock-23197
11/16/2022, 8:38 AM
Have you allocated resources to the task? If so, how much?
n
nice-judge-43061
11/16/2022, 8:40 AM
Yes, I have allocated the maximum allocatable resource. In this case 27Gi. This is how I've mentioned it for tasks:
@task(cache_version="1.0", cache=True, requests = Resources(cpu = "10" , mem = "27Gi"), limits=Resources(cpu = "10" , mem = "27Gi"))
t
tall-lock-23197
11/16/2022, 8:52 AM
Can you try setting
ephemeral_storage
to say,
1Gi
? I’m not sure if this will solve the problem, but no harm in giving it a try.