Update README.md
Browse files
README.md
CHANGED
|
@@ -17,7 +17,7 @@ tags:
|
|
| 17 |
|
| 18 |
# SnJake Anime Upscale
|
| 19 |
|
| 20 |
-
A **experimental** lightweight upscaler (x2) for anime/illustration images. Designed for clean, sharp results with minimal artifacts.
|
| 21 |
|
| 22 |

|
| 23 |
|
|
@@ -37,12 +37,17 @@ The model is designed to work with the **SnJake Anime Upscale** ComfyUI node.
|
|
| 37 |
3. Place the file(s) into `ComfyUI/models/anime_upscale/`.
|
| 38 |
4. Select the weights in the node dropdown and run the workflow.
|
| 39 |
|
| 40 |
-
#
|
| 41 |
-
|
| 42 |
- Dataset: 40,000 images from Danbooru2024: https://huggingface.co/datasets/deepghs/danbooru2024
|
| 43 |
- Validation: 600 images
|
| 44 |
- Epochs: 70
|
| 45 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 46 |
Training code is included in `training_code/` for reference.
|
| 47 |
|
| 48 |
## Disclaimer
|
|
|
|
| 17 |
|
| 18 |
# SnJake Anime Upscale
|
| 19 |
|
| 20 |
+
A **experimental** lightweight upscaler (x2) for anime/illustration images. Designed for clean, sharp results with minimal artifacts. V2 is slightly sharper and removes edge noise artifacts.
|
| 21 |
|
| 22 |

|
| 23 |
|
|
|
|
| 37 |
3. Place the file(s) into `ComfyUI/models/anime_upscale/`.
|
| 38 |
4. Select the weights in the node dropdown and run the workflow.
|
| 39 |
|
| 40 |
+
# Training Details
|
| 41 |
+
V1:
|
| 42 |
- Dataset: 40,000 images from Danbooru2024: https://huggingface.co/datasets/deepghs/danbooru2024
|
| 43 |
- Validation: 600 images
|
| 44 |
- Epochs: 70
|
| 45 |
|
| 46 |
+
V2:
|
| 47 |
+
- Slightly sharper output, no edge noise artifacts.
|
| 48 |
+
- Epochs: 20 (For now)
|
| 49 |
+
- Dataset: 49,606 images from Danbooru2024: https://huggingface.co/datasets/deepghs/danbooru2024
|
| 50 |
+
|
| 51 |
Training code is included in `training_code/` for reference.
|
| 52 |
|
| 53 |
## Disclaimer
|