[dream*void*forms]

V4: Mainly for digital artworks such as portraits, illustrations, paintings, etc. Made using a cocktail of my previous loras with significant influence from Nixport lora.

Strength: 1

Steps: 30

CFG: 7.5

V3: Trained on images from Zeen Chin (https://x.com/zeen_chin?lang=en)

• Strength: 1.4 Steps 25

• CFG scale: 3.5

• Sampler: DPM++ 2M SDE Karras

• KSampler: dpmpp_2m_sde_gpu

• Schedule: karras

• Guidance: 3.5

• VAE: vae-ft-mse-840000-ema-pruned.ckpt

• Clip skip: 2

Trained on surreal minimalist compositions featuring solitary figures, geometric architecture, vast atmospheric landscapes, and subtle emotional depth. Emphasizing clean forms, negative space, and soft gradients, this model evokes dreamlike environments where light, scale, and silence converge. Ideal for generating contemplative scenes with ethereal structures, abstract terrains, and cosmic tones.