mirabarukaso / character_select_stand_alone_app

Character Select SAA This is a Stand Alone App with AI prompt, Semi-auto Tag Complete and ComfyUI/WebUI(A1111) API support. > [!TIP] > If you find a character that isn't show on the list but can be generated correctly, please don't hesitate to let me know. > The default list is based on . > If you are interested in the list, navigate to HF download files and replace > > For browser based SAAC information, check README_SAAC.md > > For Python based CLI tool SAA Agent (BETA) and ClawHub | Item Support | ComfyUI | A1111 | Forge Neo | | --- | --- | --- | --- | | LoRA | Yes | Yes | Yes | | BREAK | No | Yes | Yes | | Refiner | Yes | Yes | Yes | | Image Color Transfer | Yes | No | No | | Regional Condition | Yes | Yes | Yes | | ControlNet/IPA | Yes | Yes | Yes | | ADetailer | Yes | Yes | Yes | | API authentication| No | Yes | Yes | | MiraITU | Yes | No | No | | UNET Model | Yes | No | Yes | | Custom VAE for SDXL | Yes | Yes | Yes | *Try Online Character Select Simple Advanced App* Hugging Face Space Install and run > [!IMPORTANT] > Setup your Image API Interface before you start SAA. > > For ComfyUI, you need ComfyUI_Mira for SAA. Clone this repo into your local folder > [!TIP] > *One-Click package v2.6.2* > The full package embeded_env_for_SAA Update The may not the latest version. If you need to update, please use the GitHub clone version with following command instead. > [!IMPORTANT] > **Updating version from github will not update the database files and .** > > Update to the latest version, then manually delete the and file. Restart the app to automatically download the latest thumbnail database from HF. Chinese Translate and Character Verification Many thanks to the following people for their selfless contributions, who gave up their valuable time to provide Chinese translation and character data verification. They are listed in no particular order. **Silence, 燦夜, 镜流の粉丝, 樱小路朝日, 满开之萤, and two more who wish to remain anonymous.** Special thanks lanner0403 WAI-NSFW-illustrious-character-select 2000+ Verified Character List, please support his plugin. Cell1310 Illustrious XL (v0.1) Recognized Characters List more than 100+ Verified Character List. mobedoor #23 MIssing characters UdinXProgrammer #62 Missing character ------ Highlights Diffusion Models (UNET/CLIP/VAE) Support > [!IMPORTANT] > These models are significantly difference from SDXL in both prompts and sampling configuration. Please make sure you understand their parameters and settings! > Test and Verified Models: Anima / Qwen Image / Z Image / Flux > Supports ComfyUI and Forge neo, NOT support original A1111. **For ComfyUI**, check Comfy-Org@HF requires custom node gguf the small-case one. **For Forge Neo**, check Download Models@Haoming02 Forge also supports the , but the use the same folder. Therefore, manage those models with correct folder yourself. Mira Image Tiled Upscaler > [!NOTE] > MiraITU > A content-aware image super-resolution ComfyUI custom node based on vision models. > Simply drag and drop ANY image into SAA/SAAC to upscale it. > Now supports Checkpoint and Diffuser models. Supports magnification scales from 1.5x to 8x. By adjusting different tile sizes, it can optimize VRAM usage to the maximum, enabling 8GB graphics cards to achieve image upscaling at 1.5x to 3x (or higher) scales, while also unleashing the full potential of high-end graphics cards with 24GB or more VRAM to accomplish 4x to 8x image upscaling. Requires ComfyUI_Mira 0.5.6.0 or above Requires MiraSubPack And Image Tagger for Mira | Settings | Drag and Drop | Flux.2 | | --- | --- | --- | | | | | | Before | 6x After (SDXL) | Before | 3x After (Flux.2) | | --- | --- | --- | --- | | | | | | The generated results will be affected by different models. The model requires a to provide more precise content descriptions, and using an is recommended. Models such as , which allow a to connected, can skip the upscale model and stretch the original image directly. Then, simply use the to configure all tiles consistently. | Upscale Ratio 1024x1360(1536) | Tile Size | VAE 8G | VAE 24G~ | | --- | --- | --- | --- | | x1.5 | 768 | Full | Full | | x2 | 1280 | Full/Tiled | Full | | x3 | 1536 | Tiled | Full | | x4 | 2048 | Tiled | Full | | x6 | 2048/2560 | Tiled | Full | | x8 | 2560/3072 | Tiled | Full/Tiled | Note: Due to limitations of the SDXL model, exceeding 8x magnification will result in overly fragmented visual slices. For higher magnification requirements, it is recommended to use a dedicated ComfyUI workflow instead. Comfyui Workflow MiraITU Image Tagger Supports WD@SmilingWolf, CL@cella110n and Camie@Camais03 models in ONNX format. Download models with tags from HF, manually rename them according to the following rules, then copy them into folder: • cl_tagger_v2.onnx + cl_tagger_v2_tag_mapping.json • wd-eva02-large-tagger-v3.onnx + wd-eva02-large-tagger-v3_selected_tags.csv • wd-v1-4-convnext-tagger.onnx + wd-v1-4-convnext-tagger_selected_tags.csv The Image Tagger running on Node.JS with . *It DOES NOT require any backend support* But, GPU acceleration seems not working The is about 3 times slower than with in CPU mode, and 12 times slower than In other words with my i9-9960x with Titan RTX The good news is, you can run during gegenerate | Device | Avg Tagging Time | Model | Platform | Resolution | Recommend Value | | --- | --- | --- | --- | --- | --- | | onnxruntime | 1.053s | cl_tagger_1_02 | Python | 448 | 0.55/0.60 | | onnxruntime-gpu | 0.297s | cl_tagger_1_02 | Python | 448 | 0.55/0.60 | | onnxruntime-node | 3.185s | cl_tagger_1_02 | Electron(NodeJS@CPU) | 448 | 0.55/0.60 | | onnxruntime-node | 2.917s | wd-eva02-large-tagger-v3 | Electron(NodeJS@CPU) | 448 | 0.35/0.85 | | onnxruntime-node | 2.113s | camie-tagger-v2 | Electron(NodeJS@CPU) | 512 | 0.50/(NOT USE) | ControlNet / IP Adapter *For ComfyUI* Upgrade your ComfyUI_Mira version to requires comfyui_controlnet_aux Put your models in requires comfyui-art-venture and ComfyUI_…