Skip to content

Add Gemma 4 parallelism#1857

Draft
rltakashige wants to merge 21 commits intomainfrom
leo/add-gemma-4-parallelism
Draft

Add Gemma 4 parallelism#1857
rltakashige wants to merge 21 commits intomainfrom
leo/add-gemma-4-parallelism

Conversation

@rltakashige
Copy link
Copy Markdown
Collaborator

Motivation

Changes

Why It Works

Test Plan

Manual Testing

Automated Testing

@rltakashige rltakashige force-pushed the leo/add-gemma-4-parallelism branch from fc457e9 to f1c9067 Compare April 8, 2026 02:04
@rltakashige rltakashige marked this pull request as draft April 9, 2026 17:15
Base automatically changed from leo/add-gemma-4 to main April 11, 2026 11:29
@Evanev7 Evanev7 linked an issue Apr 11, 2026 that may be closed by this pull request
@Wysie
Copy link
Copy Markdown
Contributor

Wysie commented Apr 13, 2026

I fixed the issue mentioned here: #1873. Attached the patch file (that can be directly applied onto this branch).

gemma4_26b_quant_alignment_v2.patch

@Evanev7 @rltakashige

@RWL-Dittrich
Copy link
Copy Markdown

I just tested the .patch file.

Fixed the issue for me. It's now running with RDMA/Tensor parallelism on two M4 Pro Mac's

@Wysie
Copy link
Copy Markdown
Contributor

Wysie commented Apr 14, 2026

@RWL-Dittrich Thanks for checking!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[BUG] Unable to load gemma-4-26b-a4b-it-6bit with Tensor and RDMA

4 participants