GPU fleet operations
QUDRA plans inference replicas and batch work together, every minute, on one fleet. The same service runs on fewer GPUs.
Use the account Qamia issued to your team.
Access is logged. Contact Qamia to request or revoke an account.