Trained on the old version of kohya-LoRA-dreambooth on Google Colab.
Trained on a set of 24 images.
45 epochs, 1 batch.
1 repeats
0.0001 LR + 0.0001 unet_LR + 5e-05 text_encoder_LR.
cosine scheduler, 130 lr_warmup_steps.
cosine
128 net_alpha, 128 net_dim.
Training resolution 896x896.
Model: NAI Model (animefull-final-pruned.ckpt), no VAE.
150 max_token_length.
fp16 mixed_precision.
Trained on the old version of kohya-LoRA-dreambooth on Google Colab.
Trained on a set of 24 images.
45 epochs, 1 batch.
1 repeats
0.0001 LR + 0.0001 unet_LR + 5e-05 text_encoder_LR.
cosinescheduler, 130 lr_warmup_steps.128 net_alpha, 128 net_dim.
Training resolution 896x896.
Model: NAI Model (animefull-final-pruned.ckpt), no VAE.
150 max_token_length.
fp16 mixed_precision.