deepspeedai / deepspeedai/DeepSpeed

Reshape ZeroStage=0 FP16 Checkpoint

Open
#2,031 5 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

bug
Dominant language
Python
Stars
43.1k
Forks
5k
Avg merge
4d 15h
Merged PRs (30d)
112

Description

What is the best way for reshaping a checkpoint trained with zero stage = 0 & fp16?

I see two options:
a) Continue training with zero stage 1 for 1 step & adapt this PR to work with fp16
b) Adapt the script here to work without the need of zero ckpts; The difficult part will just be reshaping the optimizer states in the mp_rank files

Maybe @tjruwase could give me a quick hint if a) or b) makes more sense before I waste my time? Thanks!

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.

Research direction

Start by comparing the Zero Stage 1 approach in the linked DeepSpeed PR #1953 with the linked Megatron-DeepSpeed reshaping script. The issue does not identify repository files or tests, and it leaves the design choice between adapting either approach unresolved. Done would mean a supported way to reshape FP16 Zero Stage 0 checkpoints, including optimizer states in mp_rank files.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
distributed-systems, machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.