openvino
[GPU] Robust INT8 dynamic quantization handling for hybrid linear-attn (GatedDeltaNet) models
#35965
Merged

[GPU] Robust INT8 dynamic quantization handling for hybrid linear-attn (GatedDeltaNet) models #35965

andrew-k-park
andrew-k-park andrew-k-park requested a review 31 days ago
andrew-k-park andrew-k-park requested a review 31 days ago
github-actions github-actions added category: GPU
andrew-k-park andrew-k-park removed category: GPU
andrew-k-park andrew-k-park added backport
andrew-k-park andrew-k-park added category: GPU
andrew-k-park [GPU] WA: force gs=128 dyn-quant for Mamba2 linear_attn.out_proj
a219fe42
andrew-k-park Clarify INT8 dyn-quant vs int4 weight quantization in hybrid linear-a…
7fd59510
isanghao
isanghao commented on 2026-05-19
andrew-k-park Apply comment
23ad183a
geunhwan
geunhwan approved these changes on 2026-05-19
geunhwan geunhwan added this to the 2026.2 milestone 30 days ago
geunhwan geunhwan added Code Freeze
geunhwan geunhwan enabled auto-merge 30 days ago
andrew-k-park andrew-k-park merged 27f1dbfb into releases/2026/2 30 days ago
andrew-k-park andrew-k-park deleted the gpu-dyn-qunat-linear-attn-gs128_backport branch 30 days ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
Milestone