A ComfyUI custom node pack for training LoRA / LoKR / LoHA models on the Anima architecture using kohya-ss/sd-scripts. Train Anima LoRAs directly from ComfyUI with zero command-line knowledge required.InstallationDownload and extract the zip into your <code>ComfyUI/custom_nodes/</code> folder, then restart ComfyUI. The nodes will appear under the "Anima Training" category.Features<ul><li>One-click model downloading from HuggingFace (always fetches the latest Anima base model)</li><li>Two workflow modes: Wizard (questionnaire-driven auto-config) and Manual (full control over all parameters)</li><li>Automatic sd-scripts installation and dependency management</li><li>TOML config generation with inline annotations explaining every setting</li><li>Smart VRAM presets (12GB, 16GB, 24GB) with automatic memory optimization</li><li>Live config preview inside ComfyUI before launching training</li><li>Cross-platform support (Windows, Linux, macOS)</li><li>Zero pip dependencies — pure Python stdlib</li></ul>Included Nodes<ul><li>Model Downloader — Downloads Anima models from HuggingFace</li><li>SD-Scripts Manager — Installs/updates kohya-ss/sd-scripts</li><li>Model Selector — Finds and validates model files</li><li>Training Wizard — Questionnaire-driven auto-configuration</li><li>Basic/Advanced/Memory Params — Manual parameter control</li><li>Training Launcher — Generates config and launches training</li><li>Config Preview — Shows annotated TOML config inline</li></ul>Requirements<ul><li>ComfyUI (any recent version)</li><li>Python 3.10+</li><li>kohya-ss/sd-scripts (auto-installed by the SD-Scripts Manager node)</li><li>Anima base model files (auto-downloaded by the Model Downloader node)</li></ul>Dataset PreparationYour training images should follow the sd-scripts convention: place images in a folder named <code>&lt;repeats&gt;_&lt;concept&gt;</code> inside your <code>train_data_dir</code>, with matching <code>.txt</code> caption files for each image.

Files with SHA256: d52873db6de90be65e2e6cbd22e5db46b2ae303c6d60a79008ca841f8f8afcf2