Slurm fun Awọn iṣupọ Ikẹkọ AI Idanwo

Awọn idahun ati awọn alaye han bi o ṣe n mu ibeere naa.

1. Aṣẹ wo ni awọn olumulo lo nigbagbogbo lati fi iṣẹ ipele kan silẹ si Slurm?
   • sbatch
   • kubectl waye
   • docker run
   • git titari
2. Bawo ni iṣẹ Slurm kan ṣe n beere awọn GPUs?
   • Nipa ṣiṣatunkọ awọn igbasilẹ DNS
   • Nipasẹ itọnisọna --gres=gpu
   • Nipasẹ fọọmu wẹẹbu nikan
   • GPU ko ṣee beere ni Slurm
3. Apakan Slurm wo ni o ṣe awọn ipinnu ṣiṣe eto aarin?
   • slurmd
   • srun
   • slurmctld
   • sbatch
4. Why can srun help launch distributed training tasks inside a Slurm allocation?
   • It automatically restores every model checkpoint
   • It replaces slurmctld as the scheduler
   • It converts a single GPU into multiple physical GPUs
   • It launches coordinated tasks across allocated nodes; the training setup may still need explicit DDP configuration
5. Kini idi ti ṣiṣe eto ifẹhinti ni Slurm?
   • Lati ṣiṣe awọn iṣẹ kukuru ni awọn ela laišišẹ lai ṣe idaduro iṣẹ ti o ga julọ
   • Lati pa awọn iṣẹ atijọ rẹ
   • Lati overclock GPUs
   • Lati encrypt awọn akọọlẹ iṣẹ
6. What must an AI training script do so a requeued Slurm job can continue after a node failure?
   • Assume Slurm stores model weights automatically
   • Save checkpoints and reload them when the job restarts
   • Disable resource accounting
   • Submit a new job without restoring state
7. Which Slurm daemon normally runs on each compute node to launch tasks and report node status?
   • slurmctld
   • sbatchd
   • slurmd
   • gresctl
8. What can Slurm resource accounting help cluster administrators track?
   • The model’s validation accuracy automatically
   • The semantic content of training data
   • Whether every checkpoint is scientifically valid
   • Job resource use such as allocated CPUs, GPUs and time