---
tags:
- text-to-image
- deepfloyd-if
- lora
- diffusers
- template:sd-lora
widget:
- text: a happy family
  parameters:
    negative_prompt: blurry, cropped, ugly
  output:
    url: images/step_2000_val_img_78_642C2096.png
- text: a cinematic scene from the film rogue one, showcasing a woman looking off into the distance
  parameters:
    negative_prompt: blurry, cropped, ugly
  output:
    url: images/5e433d725787c7389e0f689ce5e39490.png
- text: two anime characters in a high-energy duel, swords clashing with sparks flying
  parameters:
    negative_prompt: blurry, cropped, ugly
  output:
    url: images/anime_duel.png
- text: powerful lone anime samurai standing tall against a backdrop of a setting sun and ancient temples
  parameters:
    negative_prompt: blurry, cropped, ugly
  output:
    url: images/anime_samurai.png
- text: serene Girl with a Pearl Earring delicately gazing, bathed in ethereal moonlight, with intricate details capturing every strand of hair and every glimmer in her eye, presented in a mesmerizing fusion of oil painting, charcoal sketch, and watercolor, blending the essence of Baroque, Impressionism, and Surrealism.
  parameters:
    negative_prompt: blurry, cropped, ugly
  output:
    url: images/girl_with_pearl_earring.png
- text: a majestic portrait of a snow-capped mountain range, taken with a Canon EOS RP on f/16
  parameters:
    negative_prompt: blurry, cropped, ugly
  output:
    url: images/mountains.png
base_model: DeepFloyd/IF-I-M-v1.0
instance_prompt: cinema, scene, photograph
license: deepfloyd-if-license
pipeline_tag: text-to-image
---
# DeepCinema

<Gallery />

## Model description 

Cinema-style DeepFloyd.

## Trigger words

You should use `cinema` to trigger the image generation.

You should use `scene` to trigger the image generation.

You should use `photograph` to trigger the image generation.


## Download model

Weights for this model are available in Safetensors format.

[Download](/ptx0/deepcinema/tree/main) them in the Files & versions tab.