deepspeedai / deepspeedai/DeepSpeedExamples

One example, multiple config files

Open
#989 0 comments 0 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

I'm wondering if we can take the ZenFlow finetuning example, and extend this example into a test bed of different DeepSpeed technologies.

The ZenFlow finetuning example:
https://github.com/deepspeedai/DeepSpeedExamples/tree/master/training/DeepSpeed-ZenFlow/finetuning

The reason is that DeepSpeed is mostly config driven, which means with the same training script, it only takes a config file change to run DeepSpeed training in different mode.

How about put different config file in the directory, i.e. "z2_config.json", "z3_config.json", "z2_offload_config.json", "zenflow_config.json", etc. into the directory, then change the run.sh to make it take the config as input? This might be a good start point for anyone to quickstart deepspeed with a real example.

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

Start in the DeepSpeed-ZenFlow/finetuning example directory and inspect the existing training script, run.sh, and configuration file. Add the proposed z2_config.json, z3_config.json, z2_offload_config.json, and zenflow_config.json variants, then make run.sh accept a config-file input so each DeepSpeed mode can be launched from the same example.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Feature
Difficulty
2/5
Estimated time
1-3 hours
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
52/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.