vllm
[ROCm] Normalize fp8 scales through float32
#42247
Open
Overview
Commits
2
Changes
View On GitHub

[ROCm] Normalize fp8 scales through float32 #42247

Bortlesboat wants to merge 2 commits into vllm-project:main from Bortlesboat:codex/vllm-rocm-fp8-scale-normalize-41961
Bortlesboat
Bortlesboat [ROCm] Normalize fp8 scales through float32
7889dc69
mergify mergify added rocm
gemini-code-assist
gemini-code-assist commented on 2026-05-10
Bortlesboat Address fp8 scale review feedback
012c5c4d
Bortlesboat Bortlesboat marked this pull request as ready for review 84 days ago
Bortlesboat Bortlesboat requested a review from mgoin mgoin 84 days ago
Bortlesboat Bortlesboat requested a review from robertgshaw2-redhat robertgshaw2-redhat 84 days ago
Bortlesboat Bortlesboat requested a review from yewentao256 yewentao256 84 days ago
Bortlesboat Bortlesboat requested a review from pavanimajety pavanimajety 84 days ago
Bortlesboat Bortlesboat requested a review from zyongye zyongye 84 days ago
Bortlesboat Bortlesboat requested a review from tlrmchlsmth tlrmchlsmth 84 days ago
claude
claude commented on 2026-05-10
Bortlesboat
mergify mergify added quantization

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
rocm quantization
Milestone
No milestone
Terms Of Use Privacy Policy Refund Policy Impressum