Sign In

xuesong

2

48

0

0

Updated: Apr 16, 2024

style

Verified:

SafeTensor

Type

LoRA

Stats

48

0

Reviews

Published

Apr 16, 2024

Base Model

SDXL 1.0

Hash

AutoV2
6C99931D96
default creator card background decoration
Holiday 2023: 5 lights
arcgua's Avatar

arcgua

xuesong lora

Training with 50 iterations and 40 epochs, each comprising 3000 steps.

Since the loss is generally below 0.1 and the fitting degree is high, the weights can be reduced to a range of {0.5-0.8}.

I usually train using the anything model, which enhances realism significantly.