Stable Diffusion 2.1

This is an example deploying stable diffusion with truss weights preloaded

Prepare

The weights are ignored in this repository. You need to install them before reunning the model.

To do, the run the following code from this directory

import torch
from diffusers import StableDiffusionPipeline, EulerDiscreteScheduler

_model_id = "stabilityai/stable-diffusion-2-1-base"
_scheduler = EulerDiscreteScheduler.from_pretrained(_model_id, subfolder="scheduler")
_model = StableDiffusionPipeline.from_pretrained(
    _model_id, scheduler=_scheduler, torch_dtype=torch.float16
)
_model.save_pretrained("./data/")

Deploy to Baseten

To deploy the model, run the following from the root of the directory

import baseten, truss. os

sd = truss.load("./examples/stable-diffusion-2-1")

baseten.login(os.environ["BASETEN_API_KEY"])

baseten.deploy(sd, model_name"My SD 2.1", publish=True)