README.md

---
license: creativeml-openrail-m
base_model: runwayml/stable-diffusion-v1-5
datasets:
- lyogavin/pint_char_anim_interr_static_img_bat1-4
tags:
- stable-diffusion
- stable-diffusion-diffusers
- text-to-image
- diffusers
inference: true
---
    
# Text-to-image finetuning - lyogavin/pint_char_img_bat1-4_v15

This pipeline was finetuned from **runwayml/stable-diffusion-v1-5** on the **lyogavin/pint_char_anim_interr_static_img_bat1-4** dataset. Below are some example images generated with the finetuned pipeline using the following prompts: ['a pixel art of a fire breathing man with a sword, melted cyborg, featured on dribbble, symmetry!! yellow ranger, biker, fantasy ttrpg villain, engulfed in flames, laser wip, in style of primal apes, blackened space, 8bits videogame', 'a pixel art of a fire breathing man with a sword, melted cyborg, featured on dribbble, symmetry!! yellow ranger, biker, fantasy ttrpg villain, engulfed in flames, laser wip, in style of primal apes, blackened space, 8bits videogame', 'a pixel art of a fire breathing man with a sword, melted cyborg, featured on dribbble, symmetry!! yellow ranger, biker, fantasy ttrpg villain, engulfed in flames, laser wip, in style of primal apes, blackened space, 8bits videogame', 'a close up of a cartoon dragon with a purple and blue head, game concept art sprite sheet, rhino beetle, by Shitao, made of lava, laser wip, bulbasaur, metal border, very clear image, tank, 64x64, from overlord, inside stylized border', 'a close up of a cartoon dragon with a purple and blue head, game concept art sprite sheet, rhino beetle, by Shitao, made of lava, laser wip, bulbasaur, metal border, very clear image, tank, 64x64, from overlord, inside stylized border', 'a close up of a cartoon dragon with a purple and blue head, game concept art sprite sheet, rhino beetle, by Shitao, made of lava, laser wip, bulbasaur, metal border, very clear image, tank, 64x64, from overlord, inside stylized border']: 

![val_imgs_grid](./val_imgs_grid.png)


## Pipeline usage

You can use the pipeline like so:

```python
from diffusers import DiffusionPipeline
import torch

pipeline = DiffusionPipeline.from_pretrained("lyogavin/pint_char_img_bat1-4_v15", torch_dtype=torch.float16)
prompt = "a pixel art of a fire breathing man with a sword, melted cyborg, featured on dribbble, symmetry!! yellow ranger, biker, fantasy ttrpg villain, engulfed in flames, laser wip, in style of primal apes, blackened space, 8bits videogame"
image = pipeline(prompt).images[0]
image.save("my_image.png")
```

## Training info

These are the key hyperparameters used during training:

* Epochs: 97
* Learning rate: 1e-05
* Batch size: 4
* Gradient accumulation steps: 1
* Image resolution: 512
* Mixed-precision: fp16


More information on all the CLI arguments and the environment are available on your [`wandb` run page](https://wandb.ai/lyo-gavin/text2image-fine-tune/runs/jdnr25vg).