deepspeedai / deepspeedai/DeepSpeedExamples

[BUG] Evaluation sampler is assigned incorrectly in `step_2_reward_model_finetuning`

Open
#701 0 comments 3 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
6.8k
Forks
1.1k
Avg merge
2d 16h
Merged PRs (30d)
1

Description

Hi there,
I notice that the following line seems to be redundant and will disable the distributed eval sampler in line 242.

https://github.com/microsoft/DeepSpeedExamples/blob/32083e51c15a10fd2d90e62cd21e07e9219f8cf3/applications/DeepSpeed-Chat/training/step2_reward_model_finetuning/main.py#L247

Therefore, in multi-gpus setting, all gpu will go through the same data and this wastes a lot of computation.
Thanks.

Contributor guide

No contributing guide indexed for this repository

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.

Research direction

Open applications/DeepSpeed-Chat/training/step2_reward_model_finetuning/main.py around lines 242 and 247 and trace how the evaluation sampler is assigned. Check the multi-GPU evaluation path to confirm whether each GPU receives distinct data. Done means the distributed evaluation sampler remains active and the redundant assignment no longer causes duplicated evaluation work.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
distributed-systems, machine-learning
Issue type
Bug
Difficulty
2/5
Estimated time
1-3 hours
Activity status
Stale
Clarity
Clearly specified
Newbie friendliness
48/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.