Publish Mila multi-model LoRAs and samples

README.md

@@ -0,0 +1,106 @@
+---
+tags:
+- lora
+- character
+- mila
+- ltx-2.3
+- z-image-turbo
+- krea-2
+- text-to-image
+- text-to-video
+---
+
+# Mila — Multi-Model Character LoRAs
+
+**Mila** is a fictional character identity trained across multiple generative model families.
+
+This repository contains three separate LoRA weights for the same character:
+
+| Model | Base Model | Trigger | Weight |
+|---|---|---|---|
+| LTX-2.3 | `Lightricks/LTX-2.3` | `mila` | `models/ltx-2.3/mila_ltx23_lora.safetensors` |
+| Z-Image Turbo | `Tongyi-MAI/Z-Image-Turbo` | `mila_v1` | `models/z-image-turbo/mila_zimage_turbo_lora.safetensors` |
+| Krea 2 | Krea 2 Turbo / Raw | `mila` | `models/krea-2/mila_krea2_lora.safetensors` |
+
+## Repository structure
+
+```text
+Mila/
+├─ README.md
+├─ models/
+│  ├─ ltx-2.3/
+│  │  └─ mila_ltx23_lora.safetensors
+│  ├─ z-image-turbo/
+│  │  └─ mila_zimage_turbo_lora.safetensors
+│  └─ krea-2/
+│     └─ mila_krea2_lora.safetensors
+└─ samples/
+   ├─ ltx-2.3/
+   ├─ z-image-turbo/
+   └─ krea-2/
+```
+
+## LTX-2.3
+
+**Base model:** `Lightricks/LTX-2.3`  
+**Trigger:** `mila`  
+**Recommended LoRA strength:** `1.0`  
+**Training:** 2000 steps
+
+Example prompt:
+
+```text
+mila, blonde ponytail, soft facial features, natural skin texture,
+cream outfit, luxury coastal balcony, warm sunlight,
+realistic lifestyle video
+```
+
+## Z-Image Turbo
+
+**Base model:** `Tongyi-MAI/Z-Image-Turbo`  
+**Trigger:** `mila_v1`  
+**Training:** 2000 steps
+
+Example prompt:
+
+```text
+mila_v1, blonde ponytail, soft facial features, natural skin texture,
+cream outfit, luxury coastal street, warm sunlight,
+realistic photography
+```
+
+## Krea 2
+
+**Base model:** Krea 2  
+**Trigger:** `mila`
+
+The exact Krea 2 variant should be stated here before publishing:
+
+- `Krea 2 Turbo`, or
+- `Krea 2 Raw`
+
+Example prompt:
+
+```text
+mila, realistic blonde woman, soft facial features,
+natural skin texture, premium fashion photography,
+warm cinematic daylight
+```
+
+## Samples
+
+Example generations are organized by model family:
+
+- `samples/ltx-2.3/`
+- `samples/z-image-turbo/`
+- `samples/krea-2/`
+
+## Important
+
+These LoRAs are **not interchangeable** between base models. Load the weight that matches the corresponding base model.
+
+## Licensing
+
+Each adapter is derived from a different base model family. Usage and redistribution must comply with the license and usage terms of the corresponding base model.
+
+Please review the original model licenses before commercial use, redistribution, or deployment.

models/krea-2-turbo/mila_krea2_lora.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e521d59eee2ea4cf1f45c1044bced86d71e7d5edbdcab0dc2467bf33d9cf0017
+size 228587768

models/ltx-2.3/mila_ltx23_lora.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6819dc381ad01691831e34ba753a636c4cb0bf090f96c85144d9655cd1286433
+size 337360752

models/z-image-turbo/mila_zimage_turbo_lora.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e72a54b12be28fc62988b4761ab114b3a29857e105d33006ab05ed76f9d1ed85
+size 170128296