V1 (Anima Colab):
60-image Lora capable of generating images in the style of PC-98 games. Model created in Google Colab.
prompts: PC-98 Style, pc-98 \(style\), 1990s \(style\), fake screenshot, retro artstyle, faux retro artstyle, pixel art, dithering, parody, style parody, oekaki, limited palette
negative prompts: artist name, censored, copyright name, copyright notice, cursor, email address, (enema:1.5), gameplay mechanics, jpeg artifacts, jaggy lines, logo, lowres, mosaic censoring, multiple views, phone number, signature, text, user interface, visual novel
Base model: anima-base-v1.0
Resolution: 1024
Repeats: 2
Flip Augmentation: Off
Shuffle Tags: On
Activation Tags: 1
Steps: 1200
Epochs: 20
Unet LR: 2e-5
Text Encoder LR: 5e-5
LR Scheduler: cosine_with_restarts
LR Scheduler Number: 2
LR Warmup Ratio: 0.1
Min-SNR Gamma: 5
IP Noise Gamma: 0.05
Network Dim: 32
Network Alpha: 16
Train Batch Size: 2
Optimizer: AdamW8Bit
V1 (NoobAI):
50-image Lora capable of generating images in the style of PC-98 games.
prompts: PC-98 Style, pc-98 \(style\), 1990s \(style\), fake screenshot, retro artstyle, faux retro artstyle, pixel art, dithering, parody, style parody, oekaki, limited palette
negative prompts: artist name, censored, copyright name, copyright notice, cursor, email address, gameplay mechanics, jpeg artifacts, jaggy lines, logo, lowres, mosaic censoring, multiple views, phone number, signature, text, user interface, visual novel
Base model: NoobAI-XL EPS 1.1
Repeats: 6
Steps: 1000
epochs: 20
Unet LR: 0.0001
Text Encoder LR: 0.00001
LR Scheduler: constant
LR Scheduler Cycles: 1
Min-SNR Gamma: 5
Network Dim: 32
Network Alpha: 16