quiz.json

{
  "lesson": "78-zero-parameter-sharding",
  "title": "ZeRO Optimizer State Sharding",
  "questions": [
    {
      "stage": "pre",
      "question": "Why does ZeRO shard optimiser state first instead of parameters?",
      "options": [
        "Optimiser is more important",
        "Optimiser state is the largest memory term and is only touched during the step, not forward or backward",
        "Parameters are easier",
        "DeepSpeed mandates it"
      ],
      "correct": 1,
      "explanation": "For Adam in mixed precision the fp32 master plus two moments is 12P bytes per rank vs 4P for fp16 params + grads. Sharding the biggest term first wins the most memory per unit complexity."
    },
    {
      "stage": "pre",
      "question": "What is the per-step wire pattern of ZeRO stage 1?",
      "options": [
        "One allreduce",
        "One reduce_scatter on gradients plus one allgather on updated parameters",
        "One broadcast",
        "Nothing"
      ],
      "correct": 1,
      "explanation": "Reduce_scatter delivers each rank only its gradient shard; allgather distributes the updated parameter shards back. Total bytes equal allreduce, memory is divided by N."
    },
    {
      "stage": "check",
      "question": "Why is the per-step bandwidth the same as DDP?",
      "options": [
        "Magic",
        "Allreduce equals reduce_scatter plus allgather in bandwidth; ZeRO splits that into two halves on different data",
        "ZeRO uses compression",
        "It is not the same"
      ],
      "correct": 1,
      "explanation": "Ring allreduce is implemented as reduce_scatter then allgather. ZeRO does the same two operations but the allgather rotates the updated parameters, not the summed gradient."
    },
    {
      "stage": "check",
      "question": "For a 7B model with Adam in mixed precision on 8 ranks, what is the memory drop vs vanilla DDP?",
      "options": [
        "10%",
        "Around 65% (vanilla 16P vs ZeRO-1 4P + 12P/N = 5.5P)",
        "99%",
        "0%"
      ],
      "correct": 1,
      "explanation": "Vanilla: 2P + 2P + 4P + 4P + 4P = 16P. ZeRO-1: 2P + 2P + (4P+4P+4P)/8 = 5.5P. Drop is (16-5.5)/16 = 65.6%."
    },
    {
      "stage": "check",
      "question": "What does ZeRO-2 add over ZeRO-1?",
      "options": [
        "Nothing",
        "Also drops the non-shard gradients after reduce_scatter, freeing per-rank gradient memory; bandwidth stays the same",
        "Switches backend to NCCL",
        "Removes the optimiser"
      ],
      "correct": 1,
      "explanation": "ZeRO-2 zeroes the non-shard gradient portion after reduce_scatter; same bandwidth as ZeRO-1 but lower gradient memory."
    },
    {
      "stage": "post",
      "question": "Why does ZeRO require the optimiser state checkpoint to record which rank owns which shard?",
      "options": [
        "Cosmetic",
        "Without per-rank ownership the saved state is unreadable at restart; resuming on the same world size needs to put the right shard back on the right rank",
        "It does not",
        "Compression"
      ],
      "correct": 1,
      "explanation": "Lesson 80 builds the sharded checkpoint manifest precisely so a ZeRO run can resume on the same topology."
    }
  ]
}