Sign In

Lacey Chabert *Circa 2005*

20
342
8
Verified:
SafeTensor
Type
LoRA
Stats
342
Reviews
Published
Sep 19, 2023
Base Model
SD 1.5
Training
Steps: 1,600
Epochs: 2
Usage Tips
Clip Skip: 1
Hash
AutoV2
BB26307375
SDXL Training Contest Participant
Felldude's Avatar
Felldude

Note most checkpoint's have training of Lacey Chabert, but the training appear to have blended multiple years and possibly other celebrity face data.

To use the lora generate at 512x512 or 768x768 (Wide or Portrait can also be used)

DO NOT use the name "Lacey Chabert" during generation this will cause the lora to fight the existing images.

During generation a Lora value of .75-1.0 is recommended

When up scaling a lora value of 0.25-0.5 is recommended

While up scaling using the trained checkpoint data of "Lacey Chabert" can be done however it is recommended to use a low value such as (Lacey Chabert:0.25)

Note for Auto1111 users this is not a lora prompt it is telling the checkpoint to mix in %25 percent of the name"Lacey Chabert" training into the mix.

Since the lora doesn't have the name "Lacey Chabert" it is mixing in checkpoint training with the Lora U-Net