Downloads · 30 days
0
Mila
IAMJJJJ/Mila
Mila is a text-to-image model from IAMJJJJ. Use it when you need an image from a text prompt. The card lists the license as mit.
Access
Public
Updated Mar 1, 2026
Repo size
1.3 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors170 MB · 100%
At a glance
- Task
- Text-to-Image
- License
- mit
- Access
- Public
- Created
- Feb 24, 2026
- Updated
- Mar 1, 2026
- SHA
- b01eeebc
Base models
Model card
From the Hugging Face model README
Monoro Girl LoRA (v1)
Trigger word: monoro girl
LoRA trained on ~38 high-quality photos of Monoro girl.
Style: realistic, freckles, curly dark hair, detailed face, modest/elegant clothing, high-angle shots, shy expressions.
Base model: ostris/Flex.1-alpha (Z-Image Turbo / Flux.1-based)
Training params:
- Rank: 32
- Learning Rate: 1e-4
- Optimizer: AdamW8Bit
- Batch Size: 2
- Steps: 5000 (best at 1000–2000 steps)
- Resolution: 1024
Version history:
- v1-250: first checkpoint, basic likeness
- v1-500: improved face details
- v1-750: better recognition, good balance
Recommended usage: monoro girl, shy expression, modest clothing, high-angle shot, soft lighting, detailed face, freckles, curly hair
Negative prompt: nude, naked, cleavage, low neckline, blurry, deformed, bad anatomy, extra limbs, low quality
Download .safetensors from files below and use with LoRA weight 0.8–1.0.
Tested in Z-Image Turbo.
- Task
- Text-to-Image
- License
- mit
- Created
- Feb 24, 2026
- Updated
- Mar 1, 2026