Update README.md
Browse files
README.md
CHANGED
|
@@ -20,9 +20,6 @@ Since all models are hosted in a single repository, please follow the usage inst
|
|
| 20 |
| ------------------------ | --------------------------------- |
|
| 21 |
| `demetri_15.safetensors` | `#demerti, running and laughing` |
|
| 22 |
|
| 23 |
-
**Demo Video (generated with ComfyUI):**
|
| 24 |
-
<video autoplay loop muted controls width="100%" src="https://huggingface.co/Bisre/character_loras/resolve/main/1_demetri_with_lora.webp"></video>
|
| 25 |
-
<br>
|
| 26 |
|
| 27 |
### 2. Martin
|
| 28 |
| Filename | Trigger Words |
|
|
@@ -38,38 +35,10 @@ Since all models are hosted in a single repository, please follow the usage inst
|
|
| 38 |
|
| 39 |
---
|
| 40 |
|
| 41 |
-
## How to Use with
|
| 42 |
|
| 43 |
Because this repository contains multiple LoRA files, you **must specify the `weight_name`** (the filename) you wish to load.
|
| 44 |
-
|
| 45 |
-
```python
|
| 46 |
-
from diffusers import DiffusionPipeline
|
| 47 |
-
import torch
|
| 48 |
-
|
| 49 |
-
# 1. Load your base model. NOTE: This LoRA was trained on WAN2.1 14B.
|
| 50 |
-
# Replace the model ID with the correct one for your use case.
|
| 51 |
-
base_model_id = "Wane-Ai/wan2.1-14B" # <-- IMPORTANT: Use the correct base model ID
|
| 52 |
-
pipe = DiffusionPipeline.from_pretrained(base_model_id, torch_dtype=torch.float16).to("cuda")
|
| 53 |
-
|
| 54 |
-
# 2. Load the specific LoRA you want from this repository
|
| 55 |
-
repo_id = "Bisre/character_loras"
|
| 56 |
-
lora_filename = "demetri_15.safetensors" # <-- CHANGE THIS to the LoRA you want to use
|
| 57 |
-
|
| 58 |
-
pipe.load_lora_weights(repo_id, weight_name=lora_filename)
|
| 59 |
-
|
| 60 |
-
# 3. Generate an image
|
| 61 |
-
prompt = "photo of #demerti running and laughing, high-speed photography, dynamic action shot, cinematic lighting"
|
| 62 |
-
negative_prompt = "blurry, low quality, deformed, ugly"
|
| 63 |
-
|
| 64 |
-
image = pipe(
|
| 65 |
-
prompt,
|
| 66 |
-
negative_prompt=negative_prompt,
|
| 67 |
-
num_inference_steps=30,
|
| 68 |
-
guidance_scale=7
|
| 69 |
-
).images
|
| 70 |
-
|
| 71 |
-
image.save("demetri_result.png")
|
| 72 |
-
```
|
| 73 |
|
| 74 |
|
| 75 |
## Training Details
|
|
@@ -83,8 +52,7 @@ Hyperparameters: For a detailed look at the training configuration and hyperpara
|
|
| 83 |
(This image is an example from the dataset to show the style and quality.)
|
| 84 |

|
| 85 |
|
| 86 |
-
## Demo
|
| 87 |
|
| 88 |
<video autoplay loop muted controls width="100%" src="https://huggingface.co/Bisre/character_loras/resolve/main/Isaac_laughing.mp4"></video>
|
| 89 |
-
<video autoplay loop muted controls width="100%" src="https://huggingface.co/Bisre/character_loras/resolve/main/1_demetri_with_lora.webp"></video>
|
| 90 |
|
|
|
|
| 20 |
| ------------------------ | --------------------------------- |
|
| 21 |
| `demetri_15.safetensors` | `#demerti, running and laughing` |
|
| 22 |
|
|
|
|
|
|
|
|
|
|
| 23 |
|
| 24 |
### 2. Martin
|
| 25 |
| Filename | Trigger Words |
|
|
|
|
| 35 |
|
| 36 |
---
|
| 37 |
|
| 38 |
+
## How to Use with
|
| 39 |
|
| 40 |
Because this repository contains multiple LoRA files, you **must specify the `weight_name`** (the filename) you wish to load.
|
| 41 |
+
- Download the models you want and load into comfyUI
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 42 |
|
| 43 |
|
| 44 |
## Training Details
|
|
|
|
| 52 |
(This image is an example from the dataset to show the style and quality.)
|
| 53 |

|
| 54 |
|
| 55 |
+
## Demo Video
|
| 56 |
|
| 57 |
<video autoplay loop muted controls width="100%" src="https://huggingface.co/Bisre/character_loras/resolve/main/Isaac_laughing.mp4"></video>
|
|
|
|
| 58 |
|