Verified: 8 months ago
SafeTensor
Type | |
Stats | 98 2,277 |
Reviews | (15) |
Published | Jul 18, 2024 |
Base Model | |
Training | Steps: 4,100 Epochs: 20 |
Usage Tips | Clip Skip: 1 Strength: 0.4 |
Hash | AutoV2 A9E1EFD415 |
Created on Civitai
A very rough and over-trained LoRA for tactical gear. Probably best for in-painting.
Used on STOIQO New Reality with their recommended settings
🟡 STOIQO NewReality | SD, XL, Lightning - 🔝 🟡 XL 4.0 | Stable Diffusion Checkpoint | Civitai
"XL Version:
MAIN SAMPLER (Recommended): dpmpp_3m_sde + exponential
OTHER SAMPLERS: dpmpp_sde + Karras
FOR MORE RALISM: CFG: 1-3 | STEPS: 12+
FOR BALANCING (Recommended): CFG: 4 | STEPS: 20+
FOR MORE CREATIVITY: CFG: 5-7 | STEPS: 28+"
Use 0.4 strength as a baseline.