VictorMylle/Kohya-ss-sd-scripts
1
0
Fork 1
mirror of https://github.com/kohya-ss/sd-scripts.git synced 2026-04-06 13:47:06 +00:00
Code Issues Packages Projects Releases Wiki Activity

update multi-res training in SD3.5M

Browse Source
This commit is contained in:
Kohya S
2024-11-02 11:13:04 +09:00
parent 264328d117
commit e0db59695f
1 changed files with 6 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
README.md
View File

@@ -679,12 +679,16 @@ Other options are described below.
5. Multi-resolution Training Support:
- Only for SD3.5M.
- Same as FLUX.1 for data preparation.
- If you train with multiple resolutions, specify `--enable_scaled_pos_embed` to enable the scaled positional embeddings. The default is False. This option is an experimental feature for SD3.5M.
- If you train with multiple resolutions, you can enable the scaled positional embeddings with `--enable_scaled_pos_embed`. The default is False. __This option is an experimental feature.__
Technical details of multi-resolution training for SD3.5M:
The values of the positional embeddings must be the same for each resolution. That is, the same value must be in the same position for 512x512, 768x768, and 1024x1024. To achieve this, the positional embeddings for each resolution are calculated in advance and switched according to the resolution of the training data. This feature is enabled by `--enable_scaled_pos_embed`.
SD3.5M does not use scaled positional embeddings for multi-resolution training, and is trained with a single positional embedding. Therefore, this feature is very experimental.
Generally, in multi-resolution training, the values of the positional embeddings must be the same for each resolution. That is, the same value must be in the same position for 512x512, 768x768, and 1024x1024. To achieve this, the positional embeddings for each resolution are calculated in advance and switched according to the resolution of the training data. This feature is enabled by `--enable_scaled_pos_embed`.
This idea and the code for calculating scaled positional embeddings are contributed by KohakuBlueleaf. Thanks to KohakuBlueleaf!