Very Basic Workflow (txt2img and upscale)

This workflow is for turning text prompts into images and upscaling them. It takes prompt text, a checkpoint, and an upscaler model, and provides preview and saved image outputs.

It loads a checkpoint, encodes and combines prompts, samples and decodes an image, then scales and upscales it before previewing or saving it.

Its named processing nodes include EmptyLatentImage, KSampler, ImageScale, UpscaleModelLoader, and ImageUpscaleWithModel.

Prompt handling uses CLIPTextEncode, SeargePromptText, and easy promptConcat nodes.

PreviewImage and SaveImage provide the named image outputs.

It is listed with SD 1.5 as its base model and was published by Yorgash.

It is described as a simple workflow for people moving from A1111 or online generators to ComfyUI.

It requires node packs named comfyui-easy-use, rgthree-comfy, and SeargeSDXL.

It also lists 4x-AnimeSharp.pth and ostanderani4mix_v10.safetensors as model files.

Set prompt text, checkpoint selection, latent-image settings, sampling, and scaling in the corresponding nodes.

Vorschlag · nicht geprüft

Use the Power Lora Loader (rgthree) node when adding a LoRA-loading step.

Vorschlag · nicht geprüft

Before using the upscale path, verify that 4x-AnimeSharp.pth is available.

Vorschlag · nicht geprüft

Before generating images, verify that ostanderani4mix_v10.safetensors is available.

Vorschlag · nicht geprüft