DeepSpeed
Fix FPDT K/V gradient buffers under grouped-query attention
#8665
Merged

Fix FPDT K/V gradient buffers under grouped-query attention #8665

jinyouzhi
jinyouzhi jinyouzhi requested a review from tohtana tohtana 39 days ago
jinyouzhi jinyouzhi requested a review from tjruwase tjruwase 39 days ago
jinyouzhi jinyouzhi requested a review from loadams loadams 39 days ago
jinyouzhi Fix FPDT K/V gradient buffers under grouped-query attention
d7c98989
pengdurice pengdurice added will-review
hwchen2017
hwchen2017 approved these changes on 2026-09-26
tohtana Merge branch 'master' into fix-fpdt-gqa-backward-kv-grad
49d5bc9d
tohtana
tohtana approved these changes on 2026-09-29
tohtana tohtana enabled auto-merge 35 days ago
tohtana tohtana merged ece19cfe into master 35 days ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone