1x-Archiver-Rough.json

{
    "name": "Archiver Rough",
    "author": "loganavter",
    "license": "MIT",
    "tags": [
        "denoise",
        "restoration",
        "dehalo",
        "cartoon"
    ],
    "description": "An aggressive restoration model from the Archivist suite for severely degraded footage. It attempts to reconstruct heavily damaged or lost details through hallucination. Note: its capabilities overlap with the RGB model, but it focuses more on structural integrity than color noise.\n\nThis model is optimized for input resolutions between 720p and 1080p. Using it on significantly different resolutions may produce suboptimal results.\n\nAll Archivist models are trained on a custom dataset generated by a physics-based degradation simulator. Recommended Workflow: Use Archivist to fix physical defects, then pass the result through DRUNet (low strength) to stabilize.",
    "date": "2025-12-17",
    "architecture": "esrgan",
    "size": [
        "48nf",
        "16nb"
    ],
    "scale": 1,
    "inputChannels": 3,
    "outputChannels": 3,
    "resources": [
        {
            "platform": "pytorch",
            "type": "pth",
            "size": 1360537,
            "sha256": "82363da3d4516a8a108e7253ad9056b38e998e22870cc1580ba3e20598d8dbd8",
            "urls": [
                "https://github.com/Loganavter/Archivist-Project-Denoiser/releases/download/v1.0/1x-Archivist_Rough.pth"
            ]
        }
    ],
    "trainingIterations": 493000,
    "trainingHRSize": 256,
    "trainingOTF": true,
    "dataset": "Custom dataset (~1800 images) of classic cel animation (1940s-1980s) processed with the Project Degrader physics-based pipeline.",
    "datasetSize": 1800,
    "images": [
        {
            "type": "paired",
            "caption": "Test case: Heavy degradation and noise",
            "LR": "https://imgsli.com/i/0f4817f0-69c6-444b-8718-2b39439ef938.jpg",
            "SR": "https://imgsli.com/i/339dd3c2-40b5-4da6-9402-6c7688fe0f04.jpg"
        }
    ]
}