Jenis
LORA
Waktu publikasi
2024-02-03
Model Dasar
SD 1.5
Kata Pemicu
vampire
Salin
Perkenalan versi
推荐使用ChilloutMix模型; DreamShaper模型; Lyriel模型
ReV Animated模型; XXMix_9realistic模型
默认模式：LORA权重：0.8-0.9
采样步骤：28-30 文本强度 ：7-8
分辨率：576X1024 [9 : 16] 全身照片
ps: 频繁出超低深V装，能出图，但不容易超分上广场
分辨率：683X1024 [2 : 3] 提示词加：肖像特写
VAE：vae-ft-mse-840000-ema-pruned
采样方法：DPM++ 2M Karras
【推荐使用高清修复】：4x-UltraSharp
Recommend using the ChilloutMix model; Lyriel model;
DreamShaper model; ReV Animated model
Default mode: LORA Weight: 0.8-0.9
Sampling steps: 28-30 Text intensity: 7-8
Resolution: 576X1024 [9:16] Full body photo
PS: Frequent production of ultra-low depth V-shaped Dress skirt,
able to produce drawings,
But it's difficult to enter the AI square
Resolution: 683X1024 [2: 3]
Hint words added: Portrait close-up
VAE: vae ft mse-840000 ema run
Sampling method: DPM++2M Karras
(Recommended high-definition repair) : 4x UltraSharp
ReV Animated模型; XXMix_9realistic模型
默认模式：LORA权重：0.8-0.9
采样步骤：28-30 文本强度 ：7-8
分辨率：576X1024 [9 : 16] 全身照片
ps: 频繁出超低深V装，能出图，但不容易超分上广场
分辨率：683X1024 [2 : 3] 提示词加：肖像特写
VAE：vae-ft-mse-840000-ema-pruned
采样方法：DPM++ 2M Karras
【推荐使用高清修复】：4x-UltraSharp
Recommend using the ChilloutMix model; Lyriel model;
DreamShaper model; ReV Animated model
Default mode: LORA Weight: 0.8-0.9
Sampling steps: 28-30 Text intensity: 7-8
Resolution: 576X1024 [9:16] Full body photo
PS: Frequent production of ultra-low depth V-shaped Dress skirt,
able to produce drawings,
But it's difficult to enter the AI square
Resolution: 683X1024 [2: 3]
Hint words added: Portrait close-up
VAE: vae ft mse-840000 ema run
Sampling method: DPM++2M Karras
(Recommended high-definition repair) : 4x UltraSharp
