From f4e236f96e1ed6b5d039e2b94ecc4d879c069e0a Mon Sep 17 00:00:00 2001 From: Vladimir Mandic Date: Thu, 11 Sep 2025 15:36:53 -0400 Subject: [PATCH] group offloading Signed-off-by: Vladimir Mandic --- CHANGELOG.md | 14 +++++++++----- extensions-builtin/sdnext-modernui | 2 +- 2 files changed, 10 insertions(+), 6 deletions(-) diff --git a/CHANGELOG.md b/CHANGELOG.md index eca799291..54077ea5f 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -53,11 +53,15 @@ Also, there are quite a few offloading improvements and many quality-of-life cha in *settings -> user interface -> grid image size* - gallery now includes reference model images - **Offloading** - - enable offload during pre-forward by default - - improve offloading of models with multiple dits - - improve offloading of models with impliciy vae processing - - improve offloading of models with controlnet - - more aggressive offloading of controlnets with lowvram flag + - **balanced** + - enable offload during pre-forward by default + - improve offloading of models with multiple dits + - improve offloading of models with impliciy vae processing + - improve offloading of models with controlnet + - more aggressive offloading of controlnets with lowvram flag + - **group** + - new offloading method, using *type=leaf* works on a similar level as sequential offloading + and can present siginificant savings on low-vram gpus, but comes at the higher performace cost - **Quantization** - option to specify models types not to quantize: *settings -> quantization* allows for having quantization enabled, but skipping specific model types that do not need it diff --git a/extensions-builtin/sdnext-modernui b/extensions-builtin/sdnext-modernui index ad1765251..c72f08828 160000 --- a/extensions-builtin/sdnext-modernui +++ b/extensions-builtin/sdnext-modernui @@ -1 +1 @@ -Subproject commit ad176525127ff358f9d0b9e04e21302b5a23de85 +Subproject commit c72f08828e1cf10b53a8128e2e7c0b6de1f168d5