SDXL ControlNet
An image generation pipeline built on Stable Diffusion XL that applies a provided control image during text-to-image inference.
Deploy SDXL ControlNet behind an API endpoint in seconds.
Deploy modelExample usage
The model accepts a few main inputs:
prompt: This is text describing the image you want to generate. The output images tend to get better as you add more descriptive words to the prompt.image: Is an image that must be provided by the user as a base64 string. This input image gets used by the ControlNet to control the output from Stable Diffusion XL.
The output JSON object contains a key called result which represents the generated image as a base64 string.
Input
1import requests
2import os
3import base64
4from PIL import Image
5from io import BytesIO
6
7# Replace the empty string with your model id below
8model_id = ""
9baseten_api_key = os.environ["BASETEN_API_KEY"]
10BASE64_PREAMBLE = "data:image/png;base64,"
11
12# Function used to convert a base64 string to a PIL image
13def b64_to_pil(b64_str):
14 return Image.open(BytesIO(base64.b64decode(b64_str.replace(BASE64_PREAMBLE, ""))))
15
16# Function used to convert a PIL image to base64 string
17def pil_to_b64(pil_img):
18 buffered = BytesIO()
19 pil_img.save(buffered, format="PNG")
20 img_str = base64.b64encode(buffered.getvalue()).decode("utf-8")
21 return img_str
22
23data = {
24 "prompt": "An illustration of a forest, at sunset, fall colors, 4k, hd",
25 "image": pil_to_b64(Image.open("/path/to/input_image.png")),
26}
27
28# Call model endpoint
29res = requests.post(
30 f"https://model-{model_id}.api.baseten.co/production/predict",
31 headers={"Authorization": f"Api-Key {baseten_api_key}"},
32 json=data
33)
34
35# Get output image
36res = res.json()
37output = res.get("result")
38
39# Convert the base64 model output to an image
40img = b64_to_pil(output)
41img.save("output_image.png")
42os.system("open output_image.png")JSON output
1{
2 "result": "/9j/4AAQSkZJRgABAQAAAQABAAD/2wBDAAgGBg..."
3}Preview
Here is another example using a different prompt and image.
Input
1import requests
2import os
3import base64
4from PIL import Image
5from io import BytesIO
6
7# Replace the empty string with your model id below
8model_id = ""
9baseten_api_key = os.environ["BASETEN_API_KEY"]
10BASE64_PREAMBLE = "data:image/png;base64,"
11
12# Function used to convert a base64 string to a PIL image
13def b64_to_pil(b64_str):
14 return Image.open(BytesIO(base64.b64decode(b64_str.replace(BASE64_PREAMBLE, ""))))
15
16# Function used to convert a PIL image to base64 string
17def pil_to_b64(pil_img):
18 buffered = BytesIO()
19 pil_img.save(buffered, format="PNG")
20 img_str = base64.b64encode(buffered.getvalue()).decode("utf-8")
21 return img_str
22
23data = {
24 "prompt": "aerial view, a futuristic research complex in a bright foggy jungle, hard lighting",
25 "image": pil_to_b64(Image.open("/path/to/input_image.png")),
26}
27
28# Call model endpoint
29res = requests.post(
30 f"https://model-{model_id}.api.baseten.co/production/predict",
31 headers={"Authorization": f"Api-Key {baseten_api_key}"},
32 json=data
33)
34
35# Get output image
36res = res.json()
37output = res.get("result")
38
39# Convert the base64 model output to an image
40img = b64_to_pil(output)
41img.save("output_image.png")
42os.system("open output_image.png")JSON output
1{
2 "result": "/9j/4AAQSkZJRgABAQAAAQABAAD/2wBDAAgGBg..."
3}Preview
