Recommended Settings (v1 - Z-Image Turbo):
WORKFLOW IS IN THE SHOWCASE IMAGES. You have to use the workflow. If you use the LoRA with a simple workflow you:
Either get shitty overtrained plastic images
Or you barely change the style of turbo model and they might still look shitty
If you use it with my workflow, the seed variety is phenomenal. There is a very very less chance that you get the same-ish composition and lighting twice
Prompt adherence took a little hit (maybe its the workflow to blame). It is okay for me because it still generates the images I want. I loose a bit of adherence to get variety. If you dont like it then you can try other workflow's or reduce the weight of the LoRA or use another LoRA
Some prompts change the image way too much when you compare with Turbo but some prompts not so much. You can start with weight 0.7-0.9 for "Model with LoRA High Weight" node and 0.4-0.5 for "Model with LoRA Low Weight" node. The last stage (stage 3), disable the LoRA so you can bring back the skin texture - the image barely changes here because we use a low denoise. You can use the upscaler that I am using but keep in mind that its slow compared to others
If you want to use a character LoRA along with this LoRA, you have to enable the character LoRA at stages 2 and 3. In stage 2, you can unbypass the character lora loader node. In stage 3, you can use "Model with Character LoRA Only" node
Recommended Settings (v1 - Qwen Image):
Workflow is in the showcase images
My recommendation is to use the "EasyCache" node and let it skip 2 steps. It good with very little quality loss
I never tested it without lightning lora so I don't know how many steps to use or the CFG. You have to experiment yourself
880x1184 or 1328x1776 works best
Recommended Settings (v6 - Flux Dev):
Distilled CFG Scale: 3.5
Sampling method and Schedule type: DEIS with DDIM
Steps: 20
Resolution: 896x1152
Hires fix model: 4x_NMKD-Superscale-SP_178000_G
Steps: 10
Denoise: 0.3
Upscale by: 1.5
Lora Weight: 0.8. You have to experiment based on your prompts
Description
GPT4o Prompt:
I am training a LoRA for the Flux 1D text-to-image model that utilizes the T5XXL transformer in its architecture. To enhance this process, I require your assistance in generating detailed, natural language prompts based on uploaded images. Each prompt should begin with "Amateur photography of" and conclude with "on flickr in 2007, 2005 blog, 2007 blog," all within a single, cohesive paragraph.Do not use words like 'sharp,' 'blur,' 'focus,' 'depth of field,' or 'bokeh' in the prompt. Always provide the prompt without explicitly mentioning focus-related terms. Emphasize the clarity and vividness of the entire scene. Incorporate the use of a camera flash if usedFormat:Subject Description: Provide a comprehensive description of the main subjects in the image, covering aspects such as race, ethnicity, and physical characteristics (e.g., height, build, skin tone, hair color). Include detailed facial features (e.g., smiling with teeth visible, eyes closed, timid expression), specific expressions (e.g., joyful grin, focused gaze), and poses (e.g., side profile, upper body shot, full body shot, hands resting naturally at the sides). Specify their body type (e.g., plus-size, medium build, slim, petite) and their placement within the frame (e.g., positioned on the left, center, or right). If there are additional people in the background, summarize their presence and briefly describe their activities or interactions.Scene Description: Describe the actions and interactions of the main subjects, detailing what they are doing and the context of their activities. Provide a vivid description of the setting, whether urban or rural, indoor or outdoor, and highlight background elements such as buildings, landscapes, or furniture. Include any visible text in the image (e.g., signs, posters) and specify its location within the frame. Mention any objects the subjects interact with and describe the overall atmosphere or mood of the scene.Image Quality Tags: Emphasize uniform clarity and detail across the image. Describe the scene as filled with rich detail where nothing is obscured or lost, suggesting that every aspect is vivid and equally prominent. Highlight the lighting that brings out intricate details across both subjects and the background, creating a crisp, clearly defined image. Incorporate descriptive tags like vivid colors, consistent natural light, detailed textures, overexposure, cluttered background, warm tones, bright natural light, high contrast and harmonious clarity to subtly imply sharpness and focus throughout the scene.The final output should seamlessly integrate these elements into a detailed, coherent prompt that accurately reflects the image content.If you are ready, reply "Ok" and I will start uploading the images.