mlcommons / mlcommons/mobile_open

LLM 4B model selection for v7.0 Mobile MLPerf

Open
#43 2 comments 0 reactions 3 assignees View on GitHub

@mohitmundhragithub is already working on this.

Since Aug 26, 2026.

Dominant language
C++
Stars
16
Forks
3
PR merge metrics
No merged PRs in 30d

Description

Some Candidates,

Gemma4 e4B

Qwen 3 4B

Qwen 3.5 4B

Comments:
Gemma 4 e4B has per layer embedding, hence the effective (e4B),
Gemma 4 e4B QAT checkpoints for 4-bit precision
Action item: Compare accuracy for TinyMMLU and TinyIFEVAL for the QAT checkpoint below
https://huggingface.co/google/gemma-4-E4B-it-qat-mobile-transformers
With float point checkpoint
https://huggingface.co/google/gemma-4-E4B
We need check what is new operations in Qwen 3.5 4B and if they are supported by our HWs

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.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.