(Default) Qwan Image workflow extended with:<ol><li>LLM Prompt (Ollama DeepSeek R1)</li><li>LoRA</li><li>FaceDetailer (SDXL Turbo - default model)</li><li>Upscaling options:</li></ol><ul><li>(pixel) SD1.5 tile upscale with ControlNet</li><li>(pixel) SDXL tile upscale with ControlNet</li><li>(pixel) SDXL Lightning tile upscale with ControlNet</li><li>(pixel) Flux-Dev (Q8_0)</li><li>(latent) WAN2.2-T2V-LowNoise (Q4_K_M)</li></ul>Additional notes from testing LoRA's, settings, etc.Default settings require 24GB of VRAM.

Files with SHA256: 9e03706a69cc875983469048fff2fded4d47744363875154b3235478c956bd90