Description:
This is my first attempt at training, developed using high-quality mountain landscape materials on a local RTX 4090. MountainScape LoRA is specifically trained to generate magnificent mountain scenery, accurately capturing the grandeur of mountain ranges along with delicate details. Whether it's snow-capped peaks, rugged cliffs, or misty valleys, this model delivers rich depth and natural lighting effects.

Features:

• High Realism: Simulates authentic mountain scenery with realistic lighting and textures, suitable for photo-grade visuals.

• Rich Details: Excellent performance in depicting mountain textures, vegetation distribution, and rock surfaces.

Recommended Settings:

• Weight: 0.6 - 0.8

• Sampling Method: Euler

• Resolution: 1024x1024 and above; strongly recommended to use 1.5x Hires.fix for generating images at 1920x1080 or higher.

• Compatible Model: flux.1 dev (other Flux fine-tuned models require testing)

• Trigger Word: kmountains

• If you use the Civitai online generator, please set the CFG Scale to 3.5 or higher (important).

Dataset & Training Details:
Trained on a high-quality mountain landscape dataset with multiple optimization rounds to ensure natural aesthetics and artistic quality in the generated images.

Thank you all for your support!