README.md

metadata

license: openrail++
tags:
  - text-to-image
  - stable-diffusion
  - inferentia2
  - neuron

Neuronx model for stabilityai/stable-diffusion-xl-base-1.0

This repository contains are AWS Inferentia2 and neuronx compatible checkpoint for stabilityai/stable-diffusion-xl-base-1.0. You can find detailed information about the base model on its Model Card.

Usage on Amazon SageMaker

coming soon

Usage with optimum-neuron

from optimum.neuron import NeuronStableDiffusionXLPipeline

# Load pipeline from Hugging Face repository
pipeline = NeuronStableDiffusionXLPipeline.from_pretrained(model_dir, device_ids=[0, 1])

# prompt for generating images
prompt = "a photo of an astronaut riding a horse on mars"
# generate image
image = stable_diffusion(prompt).images[0]
image.save("generated_image.jpg")

This repository contains branches specific to versions of neuronx. When using with optimum-neuron, use the repo revision specific to the version of neuronx you are using, to load the right serialized checkpoints.

Compilation Arguments

compilation arguments

{
  "auto_cast": "all",
  "auto_cast_type": "bf16"
}

input_shapes

{
  "height": 1024,
  "width": 1024,
  "num_images_per_prompt": 1,
  "batch_size": 1
}