Training Parameters

Network Module LoKr

Base Model FLUX.1 - dev-fp8

Trigger words: raelyn

Repeat 25

Epoch 10

Clip Skip 2

Text Encoder learning rate 0.00001

Unet learning rate 0.0002

LR Scheduler: cosine_with_restarts

Optimizer: AdamW8bit

Network Dim 100000

Network Alpha 1

Factor 8

Gradient Accumulation Steps -

Shuffle caption -

Keep n tokens -

Noise offset: 0.03

Multires noise discount 0.1

Multires noise iterations 10

conv_dim 8

conv_alpha 4

Batch Size -