Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/KohakuBlueleaf/HyperKohaku
/ types & classes
Types & classes
16 in github.com/KohakuBlueleaf/HyperKohaku
⨍
Functions
78
◇
Types & classes
16
↓ 8 callers
Class
LiLoRALinearLayer
modules/lightlora.py:16
↓ 3 callers
Class
LiLoRAAttnProcessor
modules/lightlora.py:108
↓ 2 callers
Class
Attention
Attention Class without norm and residual (You need to wrap them by your self)
modules/attention.py:83
↓ 2 callers
Class
HyperDream
modules/hypernet.py:173
↓ 2 callers
Class
LiLoRAXformersAttnProcessor
modules/lightlora.py:181
↓ 2 callers
Class
PreOptHyperDream
modules/hypernet.py:238
↓ 1 callers
Class
FeedForward
modules/attention.py:26
↓ 1 callers
Class
GEGLU
modules/attention.py:16
↓ 1 callers
Class
HyperDreamBoothDataset
A dataset to prepare the instance and class images with the prompts for fine-tuning the model. It pre-processes the images and the tokenizes
train_preoptimized_liloras.py:445
↓ 1 callers
Class
HyperDreamBoothDataset
A dataset to prepare the instance and class images with the prompts for fine-tuning the model. It pre-processes the images and the tokenizes
train_hyperdreambooth.py:460
↓ 1 callers
Class
ImgWeightGenerator
modules/hypernet.py:92
↓ 1 callers
Class
TransformerBlock
modules/attention.py:190
↓ 1 callers
Class
WeightDecoder
modules/hypernet.py:31
Class
PromptDataset
A simple dataset to prepare the prompts to generate class images on multiple GPUs.
train_preoptimized_liloras.py:543
Class
PromptDataset
A simple dataset to prepare the prompts to generate class images on multiple GPUs.
train_hyperdreambooth.py:558
Class
TextWeightGenerator
WIP
modules/hypernet.py:166