tools / README_t2i_xl.py
patrickvonplaten's picture
all
6c27fdd
raw
history blame
2.5 kB
import sys
model_name = sys.argv[1]
model_card = f"""---
language:
- en
license: openrail++
tags:
- stable-diffusion
- stable-diffusion-diffusers
- stable-diffusion-xl
- text-to-image
- art
- artistic
- diffusers
- anime
---
# {model_name.split("/")[-1].replace("-", " ").capitalize()}
`{model_name}` is a Stable Diffusion model that has been fine-tuned on [stabilityai/stable-diffusion-xl-base-1.0](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0).
Please consider supporting me:
- on [Patreon](https://www.patreon.com/Lykon275)
- or [buy me a coffee](https://snipfeed.co/lykon)
## Diffusers
For more general information on how to run text-to-image models with 🧨 Diffusers, see [the docs](https://huggingface.co/docs/diffusers/using-diffusers/conditional_image_generation).
1. Installation
```
pip install diffusers transformers accelerate
```
2. Run
```py
from diffusers import AutoPipelineForText2Image, DEISMultistepScheduler
import torch
pipe = AutoPipelineForText2Image.from_pretrained('{model_name}', torch_dtype=torch.float16, variant="fp16")
pipe.scheduler = DEISMultistepScheduler.from_config(pipe.scheduler.config)
pipe = pipe.to("cuda")
prompt = "portrait photo of muscular bearded guy in a worn mech suit, light bokeh, intricate, steel metal, elegant, sharp focus, soft lighting, vibrant colors"
generator = torch.manual_seed(0)
image = pipe(prompt, num_inference_steps=25).images[0]
image.save("./image.png")
```
![](./image.png)
"""
from huggingface_hub import HfApi
api = HfApi()
read_me_path = "./README.md"
with open(read_me_path, "w") as f:
f.write(model_card)
api.upload_file(
path_or_fileobj=read_me_path,
path_in_repo=read_me_path,
repo_id=model_name,
repo_type="model",
)
from diffusers import AutoPipelineForText2Image, DEISMultistepScheduler
import torch
pipe = AutoPipelineForText2Image.from_pretrained(model_name, torch_dtype=torch.float16)
pipe.scheduler = DEISMultistepScheduler.from_config(pipe.scheduler.config)
pipe = pipe.to("cuda")
prompt = "portrait photo of muscular bearded guy in a worn mech suit, light bokeh, intricate, steel metal, elegant, sharp focus, soft lighting, vibrant colors"
generator = torch.manual_seed(0)
image = pipe(prompt, num_inference_steps=25).images[0]
image_path = "./image.png"
image.save(image_path)
api.upload_file(
path_or_fileobj=image_path,
path_in_repo=image_path,
repo_id=model_name,
repo_type="model",
)
pipe.push_to_hub(model_name, variant="fp16")