Closed
Description
LocalAI version:
Dockerimage quay.io/go-skynet/local-ai:v2.0.0-cublas-cuda11-ffmpeg
Describe the bug
I'm trying to implement an Image2Image transformation using the Stable Diffusion model with Diffusers, as described in the guide https://localai.io/model-compatibility/diffusers/. I am using the latest Docker image v2.0.0. Normal Text-to-Image generation is working. However, I keep encountering an error, and I'm not sure what I'm doing wrong. Here's the error message I receive:
{
"error":{
"code":500,
"message":"rpc error: code = Unknown desc = Exception calling application: 'image'",
"type":""
}
}I have set up my YAML configuration as follows:
backend: diffusers
diffusers:
cuda: true
enable_parameters: negative_prompt,num_inference_steps,image
pipeline_type: StableDiffusionImg2ImgPipeline
f16: true
name: stablediffusion-edit
parameters:
model: nitrosocke/Ghibli-Diffusion
step: 25And here's the Python script I'm using to encode the image in Base64 and create the JSON request:
import base64
import json
# Convert image to Base64
with open("test.jpg", "rb") as image_file:
encoded_string = base64.b64encode(image_file.read()).decode()
# Create JSON body
data = {
"image": encoded_string,
"prompt": "a sky background",
"size": "512x512",
"model": "stablediffusion-edit"
}
# Save JSON to file
with open('request.json', 'w') as outfile:
json.dump(data, outfile)
# Execute cURL command
!curl -H "Content-Type: application/json" -d @request.json http://localhost:8080/v1/images/generationsAlso tried it with Openai:
from openai import OpenAI
client = OpenAI(base_url="http://localhost:8080/v1", api_key="sk-xxx")
client.images.edit(
image=open("original.png", "rb"),
mask=open("mask.png", "rb"),
prompt="A cute baby sea otter wearing a beret",
n=2,
size="256x256"
)Here I am getting the error:
NotFoundError: Error code: 404 -
{
"error":{
"code":404,
"message":"Cannot POST /v1/images/edits",
"type":""
}
}