mlcommons / mlcommons/inference

Selection criteria of models for benchmark

Open
#2,200 4 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
1.6k
Forks
650
Avg merge
1d 22h
Merged PRs (30d)
6

Description

Hello,

I have a genuinely curious question. How did you select the models for the inference benchmark? There are many models available, and the MLPerf Datacenter inference benchmark is conducted on the following models:

  1. resnet50-v1.5
  2. retinanet 800x800
  3. bert
  4. dlrm-v2
  5. 3d-unet
  6. gpt-j
  7. stable-diffusion-xl
  8. llama2-70b
  9. llama3.1-405b
  10. mixtral-8x7b
  11. rgat
  12. pointpainting

Is there a particular reason why these models were chosen? Thank you for any pointers or guidance.

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

The issue names no files, tests, or entry points. Start by reviewing the benchmark model list and the MLPerf Datacenter inference models cited in the issue, then look for existing documentation on selection criteria. Done means documenting a clear rationale for the selected models or pointing to the authoritative guidance.

Written by the indexing model from the issue text.

Assessment

Tech stack
machine-learning
Domain
machine-learning, testing-qa
Issue type
Documentation
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.