ostris/nighttime-lora

Nighttime LoRA

Prompt

Prompt

close up of a homeless man, sitting near a dumpster, looking into the aley, dirty face, beard, worn beanie

Negative Prompt

cartoon, cgi, render, illustration, painting, drawing, bad quality, grainy, low resolution

Prompt

a man sitting alone at a bar, beart, worn beanie, side view, big beer, hands in pockets

Negative Prompt

closed eyes, hair, cartoon, cgi, render, illustration, painting, drawing, bad quality, grainy, low resolution

Prompt

photo of a woman looking at christmas lights

Negative Prompt

cartoon, cgi, render, illustration, painting, drawing, bad quality, grainy, low resolution

Prompt

woman reading a book in the park, flashlight , close up

Negative Prompt

cartoon, cgi, render, illustration, painting, drawing, bad quality, grainy, low resolution

Prompt

man holding a lantern, close up, shocked, mustache

Negative Prompt

cartoon, cgi, render, illustration, painting, drawing, bad quality, grainy, low resolution

Prompt

woman sitting in a coffee shop, black robe, red hair, holding a lightsaber

Negative Prompt

cartoon, cgi, render, illustration, painting, drawing, bad quality, grainy, low resolution

Prompt

a woman super excited, holding a jar with fireflies

Negative Prompt

too many fingers, bad anatomy,, cartoon, cgi, render, illustration, painting, drawing, bad quality, grainy, low resolution

Model description

This lora can turn any scene into a night scene, indoors and outdoors. Not just darker, but with flashlights, car lights, etc. Some scenes will need a stronger weight than others. So adjust as needed.

This is an experiment in a training process concept I am working on I call seed locking. Initial results looked good, but it still needs work. The basic idea is to generate training images and regularization images with a specific seed, then alter the training images with the desired effect while maintaining image consistency, and then retrain with those specific seeds to target very specific concepts you want to adjust and keeping everything else the same. It is not perfect because of the sampling methods. I hope to find a way to regularize for the sampler in the loss, hopefully soon.

But for now, enjoy the experiment.

Download model

Weights for this model are available in Safetensors format.

Download them in the Files & versions tab.

Use it with the 🧨 diffusers library

from diffusers import AutoPipelineForText2Image
import torch

pipeline = AutoPipelineForText2Image.from_pretrained('runwayml/stable-diffusion-v1-5', torch_dtype=torch.float16).to('cuda')
pipeline.load_lora_weights('ostris/nighttime-lora', weight_name='nighttime_v1.safetensors')
image = pipeline('a woman super excited, holding a jar with  fireflies  ').images[0]

For more details, including weighting, merging and fusing LoRAs, check the documentation on loading LoRAs in diffusers