SnJake_Baikal_Swin_Anime

SnJake Baikal-Swin-Anime x2 is a custom ComfyUI node for upscaling anime/illustration images with a dedicated restoration model. Model in experimental state; V2 is slightly sharper and removes edge noise artifacts.

---

Examples

<img width="4096" height="2048" alt="Example_3" src="https://github.com/user-attachments/assets/12e77d78-acee-4bff-9ccb-e82fc92bf23e" /> <img width="4096" height="2048" alt="Example_2" src="https://github.com/user-attachments/assets/2382d7bf-bdfd-4f40-abd6-834238c825aa" /> <img width="4096" height="2048" alt="Example_1" src="https://github.com/user-attachments/assets/a80bcb47-8568-4365-a6da-2bbd303c6f59" />

---

Installation

The installation consists of two steps: installing the node and making the weights available.

Step 1: Install the Node

1. Open a terminal or command prompt.
2. Navigate to your ComfyUI custom_nodes directory.

   # Example for Windows
   cd D:\ComfyUI\custom_nodes\
   
   # Example for Linux
   cd ~/ComfyUI/custom_nodes/
   

3. Clone this repository:

   git clone https://github.com/SnJake/SnJake_Baikal_Swin_Anime.git
   

4. For standard ComfyUI installations (with venv):
   1. Make sure your ComfyUI virtual environment (venv) is activated.
   2. Navigate into the new node directory and install the requirements:

      cd SnJake_Baikal_Swin_Anime
      pip install -r requirements.txt
      

   For Portable ComfyUI installations:
   1. Navigate back to the root of your portable ComfyUI directory (e.g., D:\ComfyUI_windows_portable).
   2. Run the following command to use the embedded Python to install the requirements. Do not activate any venv.

      python_embeded\python.exe -m pip install -r ComfyUI\custom_nodes\SnJake_Baikal_Swin_Anime\requirements.txt
      

Step 2: Model Weights

On first use the node can automatically download weights from the repository.

• Default weights location: ComfyUI/models/anime_upscale/

If you want to download manually:

1. Download the weights from HF REPO.
2. Place the file(s) into ComfyUI/models/anime_upscale/.

Step 3: Restart

Restart ComfyUI completely. The node will appear under 😎 SnJake/Upscale.

---

Usage

The node menu path is 😎 SnJake/Upscale.

Inputs

• weights_name: Select weights from the dropdown (auto-download if missing).
• image: Source image.
• tile: Tile size for large images. Recommended 256-512. Set 0 to disable tiling.
• overlap: Tile overlap for smooth blending. Recommended 32-64.
• amp: Precision (auto, bf16, fp16, none).
• device: Device (auto, cuda, cpu).

Outputs

• image: Upscaled image.

---

Training Details

V1:

• Dataset: 40,000 images from Danbooru2024: https://huggingface.co/datasets/deepghs/danbooru2024
• Validation: 600 images
• Epochs: 70

V2:

• Slightly sharper output, no edge noise artifacts.
• Epochs: 20
• Dataset: 49,606 images from Danbooru2024: https://huggingface.co/datasets/deepghs/danbooru2024
• Perceptual backbone: Custom SimSiam pre-trained convnextv2_tiny (Experimental)
• Loss schedule: gradual ramp‑in of perceptual/auxiliary losses for stable training.

V2.1:

• Removed Nearest from resample_methods
• Epochs: 30

V2.2:

• Epochs: 40 (For now)

V3 (SwinFIR):

• Epochs: Stage 1 - 20; Stage 2 - 18
• Perceptual backbone: Custom SimSiam pre-trained convnextv2_base (Experimental).
• Dataset: ~50,000 images from Danbooru2024

V3.1 (Current Best - SwinFIR): 🏆

• Epochs: Stage 1 (Charbonnier) - 20; Stage 2 (GAN Fine-tuning) - 16.
• Perceptual backbone: Reverted to the robust ImageNet-pretrained convnextv2_base.fcmae_ft_in22k_in1k for superior high-frequency feature extraction.
• Dataset: ~50,000 images from Danbooru2024

Training code is included in training_code/ for reference.

---

Disclaimer

This project was made purely for curiosity and personal interest. The code was written by GPT-5.2 Codex.

---

License

MIT. See LICENSE.md.