Sci-Fi Diffusion v1.0

Name: Sci-Fi Diffusion v1.0
Rating: 5 (3239 reviews)
Author: Corruptlake

3.6k

17.5k

211

Updated: Oct 4, 2024

style

photorealistic science fiction scifi realistic sci fi

00074-4167971556-landscape of a futuristic sci fi city, sci fi, ultra realistic, high resolution, city.png

00082-3058935260-futuristic sci fi spaceship, star citizen, star atlas, render,.png

00089-2743804849-futuristic sci fi spaceship, star citizen, star atlas, render,.png

00092-1357129916-futuristic sci fi spaceship, in a sci-fi city, star citizen, star atlas, render,.png

00101-3209903627-a futuristic sci fi spaceship, in a sci-fi city, star citizen, star atlas, render, one spaceship, 1spaceship.png

00129-3438468168-A photo of a sci-fi ((1tank)) on a frozen ice planet.png

00142-2461655431-A render of a sci-fi space outpost on a frozen ice planet.png

00151-3115233758-A render of a sci-fi palace on an ice planet, night, galaxy sky.png

00153-4076248628-A render of a sci-fi ((palace)) on an ice planet,.png

00157-26536476-a photo of the bosporus strait night, ((sci-fi)), caspian sci-fi, istanbul sci-fi.png

00171-874457832-a render of alien sci-fi city. ultra detailed, ultra realistic, high resolution, high quality.png

Download (1.99 GB)

Verified: 2 years ago

PickleTensor

This asset is only available as a PickleTensor which is a deprecated and insecure format. We caution against using this asset until it can be converted to the modern SafeTensor format.

Details

Type	Checkpoint Trained
Stats	17,478
Reviews	Overwhelmingly Positive (3,619)
Published	Jan 12, 2023
Base Model	SD 1.5
Training	Steps: 53,898 Epochs: 2
Trigger Words	sci-fi caspian sci-fi star atlas star citizen render
Hash	AutoV2 CD73C52CD0

1 File

default creator card background decoration

Corruptlake

License:

CreativeML Open RAIL-M Addendum

An early version of the upcoming generalist Sci-Fi model based on SD v2.X.

This model has been trained on 26,949 high resolution and quality Sci-Fi themed images for 2 Epochs. Which equals to around 53K steps/iterations. The training resolution was 640, however it works well at higher resolutions. This model is still in developement

Recommended Samplers:

- Euler
- Euler A
- DPM++ SDE Karras