Upload v2.0_wan-t2v-14b/README.md with huggingface_hub
Browse files- v2.0_wan-t2v-14b/README.md +80 -0
v2.0_wan-t2v-14b/README.md
ADDED
|
@@ -0,0 +1,80 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
tags:
|
| 5 |
+
- art
|
| 6 |
+
---
|
| 7 |
+
|
| 8 |
+
# Flat Color - Style
|
| 9 |
+
|
| 10 |
+
**Creator**: [motimalu](https://civitai.com/user/motimalu)
|
| 11 |
+
**Type**: LORA
|
| 12 |
+
**Base Model**: Wan Video 14B t2v
|
| 13 |
+
**Version**: v2.0 [wan-t2v-14b]
|
| 14 |
+
**Trigger Words**: `flat color, no lineart`
|
| 15 |
+
|
| 16 |
+
**Civitai Model ID**: 1132089
|
| 17 |
+
**Civitai Version ID**: 1474944
|
| 18 |
+
|
| 19 |
+
**Stats (at time of fetch for this version)**:
|
| 20 |
+
* Downloads: 3713
|
| 21 |
+
* Rating: 0 (0 ratings)
|
| 22 |
+
* Favorites: N/A
|
| 23 |
+
|
| 24 |
+
---
|
| 25 |
+
|
| 26 |
+
## 📄 Description (Parent Model)
|
| 27 |
+
|
| 28 |
+
Flat Color
|
| 29 |
+
-
|
| 30 |
+
Style
|
| 31 |
+
Trained on images without visible lineart, flat colors, and
|
| 32 |
+
little to no indication of depth.
|
| 33 |
+
ℹ️ LoRA work best when applied to the base models on which they are trained.
|
| 34 |
+
Please read the
|
| 35 |
+
About This Version
|
| 36 |
+
on the appropriate base models
|
| 37 |
+
and workflow/training information.
|
| 38 |
+
This is a small style LoRA I thought would be interesting to try with a v-pred model (noobai v-pred), for the reduced color bleeding and strong blacks in particular.
|
| 39 |
+
The effect is quite nice and easy to evaluate in training, so I've extended the dataset with videos in following versions for text-to-video models like Wan and Hunyuan, and it is what I am generally using to test LoRA training on new models now.
|
| 40 |
+
Recommended prompt structure:
|
| 41 |
+
Positive prompt:
|
| 42 |
+
flat color, no lineart, blending, negative space,
|
| 43 |
+
{{tags}}
|
| 44 |
+
masterpiece, best quality, very aesthetic, newest
|
| 45 |
+
|
| 46 |
+
## Version Notes (v2.0 [wan-t2v-14b])
|
| 47 |
+
|
| 48 |
+
[WAN 14B] LoRA
|
| 49 |
+
Trained with
|
| 50 |
+
diffusion-pipe
|
| 51 |
+
on
|
| 52 |
+
Wan2.1-T2V-14B
|
| 53 |
+
with the same config as the v2.0 [hunyuan] version
|
| 54 |
+
Seems usable for both text-to-video and image-to-video workflows with Wan
|
| 55 |
+
Text to Video previews generated with
|
| 56 |
+
ComfyUI_examples/wan/#text-to-video
|
| 57 |
+
Loading the LoRA with LoraLoaderModelOnly node and using the fp8 14B
|
| 58 |
+
wan2.1_t2v_14B_fp8_e4m3fn.safetensors
|
| 59 |
+
* it is recommended to use the newer scaled fp8 14B
|
| 60 |
+
wan2.1_t2v_14B_fp8_scaled.safetensors
|
| 61 |
+
Image to Video previews generated with
|
| 62 |
+
ComfyUI_examples/wan/#image-to-video
|
| 63 |
+
Using image generations from the IL/Noobai versions of this model card
|
| 64 |
+
|
| 65 |
+
---
|
| 66 |
+
|
| 67 |
+
## Civitai Links
|
| 68 |
+
|
| 69 |
+
* **[🔗 View This Version on Civitai →](https://civitai.com/models/1132089?modelVersionId=1474944)**
|
| 70 |
+
* [View Full Model Page →](https://civitai.com/models/1132089)
|
| 71 |
+
* [View Creator Profile →](https://civitai.com/user/motimalu)
|
| 72 |
+
|
| 73 |
+
---
|
| 74 |
+
|
| 75 |
+
## File Information
|
| 76 |
+
|
| 77 |
+
* **Filename**: `wan_flat_color_v2.safetensors`
|
| 78 |
+
* **Size**: 292.59 MB
|
| 79 |
+
* **Hash (AutoV2)**: `E61910EC19`
|
| 80 |
+
* **Hash (SHA256)**: `E61910EC19FC602D72248C4F3ABD338B614D82394F357D2D67F6E1575B4F4819`
|