DistributedSampler(num_replicas=0) raises a rank error instead of rejecting num_replicas

Open Beginner friendly
#196,090 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Assessment

Difficulty
2/5
Estimated time
1-3 hours
Newbie friendliness
72/100
Issue type
Bug
Clarity
Mostly clear
Activity status
Active
Tech stack
python
Domain
distributed-systems

Research direction

Start at the DistributedSampler constructor in torch.utils.data.distributed and reproduce the reported call with num_replicas=0. Review the constructor checks and verify that invalid num_replicas values are rejected directly rather than reported as an invalid rank. Done means the reported input produces a ValueError identifying num_replicas, without reaching division by zero.

Written by the indexing model from the issue text.

Description

bot-triaged module: dataloader module: error checking triaged

I confirmed this on current pytorch main. num_replicas <= 0 is blamed on rank. Drafted with AI assistance; I reviewed the constructor checks.

🐛 Describe the bug

DistributedSampler(..., num_replicas=0, rank=0) raises a ValueError about rank, not num_replicas.

import torch
from torch.utils.data.distributed import DistributedSampler

DistributedSampler(range(10), num_replicas=0, rank=0)
# ValueError: Invalid rank 0, rank should be in the interval [0, -1]

The constructor only checks rank >= num_replicas or rank < 0. When num_replicas is 0 or negative, that interval is meaningless and a valid rank=0 is reported as invalid. Later len(dataset) / num_replicas would divide by zero if the rank check were reordered.

Versions

pytorch main.

cc @andrewkho @divyanshk @SsnL @VitalyFedyunin @dzhulgakov @scotts @malfet

Dominant language
Python
Stars
103k
Forks
29.9k
No merged PRs in 30d

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from pytorch/pytorch

All issues in pytorch/pytorch

Similar issues

More Python issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.