Upload 2856467/3266628/README.md with huggingface_hub
Browse files- 2856467/3266628/README.md +20 -2
2856467/3266628/README.md
CHANGED
|
@@ -11,7 +11,7 @@ tags:
|
|
| 11 |
- woman
|
| 12 |
---
|
| 13 |
|
| 14 |
-
# [MMH3] Mystic XXX - v4.0
|
| 15 |
|
| 16 |
**Model Type**: LORA
|
| 17 |
|
|
@@ -46,6 +46,24 @@ tags:
|
|
| 46 |
|
| 47 |
## Description
|
| 48 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
New ref2va file to download (experimental, no audio trained)
|
| 50 |
============================================================
|
| 51 |
|
|
@@ -67,4 +85,4 @@ Model: [CivitAI Model Page](https://civitai.com/models/2856467?modelVersionId=32
|
|
| 67 |
|
| 68 |
Archive: [CivArchive Page](https://civarchive.com/models/2856467?modelVersionId=3266628)
|
| 69 |
|
| 70 |
-
<!-- Version:
|
|
|
|
| 11 |
- woman
|
| 12 |
---
|
| 13 |
|
| 14 |
+
# [MMH3] Mystic XXX - v4.0 - 3266628
|
| 15 |
|
| 16 |
**Model Type**: LORA
|
| 17 |
|
|
|
|
| 46 |
|
| 47 |
## Description
|
| 48 |
|
| 49 |
+
V5 - REF2VA
|
| 50 |
+
===========
|
| 51 |
+
|
| 52 |
+
This version was refined from the FL2VA using only MLP from 10-49 to reduce interference with other loras and references, it has the same audio as FL2VA and was retrained to gain reference capabilities that FL2VA don't have, for this a dataset with references and goals is used. It was made with Musubi Trainer.
|
| 53 |
+
|
| 54 |
+
V5 - FL2VA
|
| 55 |
+
==========
|
| 56 |
+
|
| 57 |
+
This version was trained with audio using the full layers spectrum, including both MLP and attention layers, with Musubi Trainer.
|
| 58 |
+
|
| 59 |
+
Compared to V4, it delivers more natural skin rendering and lighting, improved movement, more realistic results, and better audio quality. The dataset was increased by approximately 50%, and the training resolution was raised to capture more detail. Anatomy is also slightly improved, with better prompt adherence in some cases.
|
| 60 |
+
|
| 61 |
+
I tested it against V4 across more than 100 videos, and this version produced better results in approximately 70% of them. The validation results also improved, with a particularly significant improvement in T2VA performance.
|
| 62 |
+
|
| 63 |
+
V5 is based on FL2VA and uses all layers, resulting in a larger LoRA with a greater potential for interference with other LoRAs, particularly those targeting overlapping layers. If you encounter compatibility issues, let me know. A version trained without attention and identity layers may be a viable alternative, although removing these layers would likely reduce the overall quality.
|
| 64 |
+
|
| 65 |
+
---
|
| 66 |
+
|
| 67 |
New ref2va file to download (experimental, no audio trained)
|
| 68 |
============================================================
|
| 69 |
|
|
|
|
| 85 |
|
| 86 |
Archive: [CivArchive Page](https://civarchive.com/models/2856467?modelVersionId=3266628)
|
| 87 |
|
| 88 |
+
<!-- Version: 20261020_upload -->
|