VisualNote Detailer DoRA v3

VisualNote Detailer DoRA v3

DORA
Original


Updated:

DoRA: Compared to LoRA,

Better Stability,

Reliability and

Style Adherence

Version Detail

Trained by Tensor
Illustrious
3
VisualNote Test DoRA v3 Key training settings changes: - LR Scheduler: cosine (from constant/_with_warmup) - steps: 2583 - LoRA to DoRA Parameter Settings Network Module DoRA Use Base Model WAI-illustrious - v16 Trigger words VisualNote style Image Processing Parameters Repeat 1 Epoch 3 Save Every N Epochs 1 Training Parameters Seed - Clip Skip 2 Text Encoder learning rate 0.00001 Unet/DiT learning rate 0.0001 LR Scheduler cosine Optimizer AdamW8bit Network Dim 64 Network Alpha 32 Gradient Accumulation Steps - Label Parameters Shuffle caption true Keep n tokens 1 Advanced Parameters Noise offset 0.03 Multires noise discount 0.3 Multires noise iterations 10 conv_dim 128 conv_alpha 64 Batch Size 1 Sample Image Settings Prompt 1girl, solo, amazing composition, masterpiece, best quality, highres, absurdres, amazing quality, very aesthetic, amazing illustration, blurry background, VisualNote style 1girl solo breasts short hair gloves navel holding jewelry medium breasts underwear green eyes standing panties full body weapon red hair earrings boots socks belt sword holding weapon armor black panties profile underboob holding sword helmet shoulder armor gauntlets pauldrons shield bikini armor Sampler euler_ancestral loss: 0.0757

Project Permissions

    Use Permissions

  • Use in TENSOR Online

  • As a online training base model on TENSOR

  • Use without crediting me

  • Share merges of this model

  • Use different permissions on merges

    Commercial Use

  • Sell generated contents

  • Use on generation services

  • Sell this model or merges

Related Posts