Fine-Tunes
Learn about Fine-Tuning a Stable Diffusion Model with your pictures. Fine-Tune a model, retrieve, delete, and list Fine-Tunes.
The Fine-Tune Model
The Fine-Tune Model contains all the information you need about Fine-Tuning a Stable Diffusion Model. It also contains the prompts that you run to your Fine-Tune Model in Bodia.
Properties
- Name
id
- Type
- string
- Description
Unique identifier of the Fine-Tune.
- Name
title
- Type
- string
- Description
The Title of the Fine-Tune.
- Name
name
- Type
- string
- Description
This is the name of the identifier object. It must be included in every prompt.
- Name
token
- Type
- string
- Description
Token, together with the name property which references the identifier object. If it is not included in the Prompt, it is automatically included just before the object name. Recommended Options:
zwx
,ukj
orsks
- Name
images_url
- Type
- array
- Description
The images used to Fine-Tune the model.
- Name
model
- Type
- string
- Description
Options:
sd1.5
,sd2.1
oropenjourney2
. The base Stable Diffusion Model you want to Fine-Tune.
- Name
prompts
- Type
- array
- Description
You can Fine-Tune a Stable Diffusion Model with pre-made prompts. You can see the prompt properties in the Prompt Properties section below.
- Name
styles
- Type
- array
- Description
You can also Fine Tune a Stable Diffusion Model using Styles, created by you or the public ones. You can see the collections and styles properties in the Styles Properties and Collections Properties sections.
- Name
steps
- Type
- number
- Description
Number of training steps (minimum: 500; maximum: 5000).
- Name
metadata
- Type
- JSON
- Description
You can add metadata info about your Fine-Tune, like
user_id
, as well as other information you want to include to integrate it with your system.
- Name
error_message
- Type
- string
- Description
An error message just in case any errors occur.
Create Fine-Tune
This endpoint allows you to Fine-Tune a new Model and adds prompts to it.
Required attributes
- Name
title
- Type
- string
- Description
The Title of the Fine-Tune.
- Name
name
- Type
- string
- Description
This is the name of the identifier object. It must be included in every prompt.
- Name
token
- Type
- string
- Description
Token that together with the name property references the identifier object. If it is not included in the Prompt, it is automatically included just before the object name. Recommended Options:
zwx
,ukj
orsks
- Name
images_url
- Type
- array
- Description
The images to Fine-Tune the model.
Optional attributes
- Name
model
- Type
- string
- Description
Options:
sd1.5
,sd2.1
oropenjourney2
. The base Stable Diffusion Model you want to Fine-Tune.
- Name
prompts
- Type
- array
- Description
You can Fine-Tune a Stable Diffusion Model with pre-made prompts. You can see the prompts properties in the Prompts Properties section below.
- Name
styles
- Type
- array
- Description
You can also Fine Tune a Stable Diffusion Model using styles or collections of styles, created by you or the public ones. You can see the collections and styles properties in the Collections Properties and Styles Properties sections.
- Name
steps
- Type
- number
- Description
Number of training steps (minimum: 500; maximum: 5000).
- Name
metadata
- Type
- JSON
- Description
You can add metadata info about your Fine-Tune, like
user_id
, and other information you want to include to integrate it with your system.
Request
curl --request POST \
--url https://api.bodia.ai/v1/finetunes \
--header 'Authorization: Bearer {API_KEY}' \
--data '{
"title": "Fine-Tune Title",
"name": "car",
"images_url": [
"https://via.placeholder.com/512x512.jpg",
"https://via.placeholder.com/512x512.jpg",
"https://via.placeholder.com/512x512.jpg",
"https://via.placeholder.com/512x512.jpg"
],
metadata: {
"user_id": "example_id"
},
"prompts": [{
"prompt": "A car on the route 99 road",
"negative_prompt": "stones",
"cfg": 7.5,
"steps": 35,
"seed": 2323,
"high_resolution": true,
"aspect_ratio": "square",
"metadata": {
"info": "example info I want to include"
}
},{
"prompt": "A mediterranean portrait of a car driving in the mediterranean sea"
}],
"styles":[{
"id": "YF1xORgvKIbHy-LeF1vvWYDp",
"aspect_ratio": "portrait",
"high_resolution": true
}]
}'
Response
{
"id": "WAz8eIbvDR60rouK",
"title": "Training 1",
"images_url": [
"https://via.placeholder.com/512x512.jpg",
"https://via.placeholder.com/512x512.jpg",
"https://via.placeholder.com/512x512.jpg",
"https://via.placeholder.com/512x512.jpg"
],
"metadata": {
"user_id": "example_id"
},
"steps": 2000,
"created_at": "2023-02-08T22:14:03.800Z",
"name": "car",
"token": "zwx",
"model": "sd1.5"
}
List Fine-Tunes
GET request to the Fine-Tunes endpoint to retrieve a list of all your fine tunes.
Request
curl -G https://api.bodia.ai/v1/finetunes \
-H "Authorization: Bearer {API_KEY}"
Response
[
{
"id": "WAz8eIbvDR60rouK",
"title": "Car Fine-Tune",
"name": "car",
"token": "zwx",
"error_message": null,
"metadata": null,
"images_url": [
"https://via.placeholder.com/512x512.jpg",
"https://via.placeholder.com/512x512.jpg",
"https://via.placeholder.com/512x512.jpg",
"https://via.placeholder.com/512x512.jpg"
],
"steps": 2000,
"created_at": "2023-01-31T15:16:39.959Z",
"model": "sd1.5"
}
]
Retrieve a Fine-Tune
This endpoint allows you to retrieve a Fine-Tune by providing their ID. Refer to Fine-Tune Model Properties to see which properties are included with Fine-Tune objects.
Request
curl -G https://api.bodia.ai/v1/finetunes/{TUNE_ID} \
-H "Authorization: Bearer {API_KEY}"
Response
{
"id": "WAz8eIbvDR60rouK",
"title": "Car Fine-Tune",
"name": "car",
"token": "zwx",
"images_url": [
"https://via.placeholder.com/512x512.jpg",
"https://via.placeholder.com/512x512.jpg",
"https://via.placeholder.com/512x512.jpg",
"https://via.placeholder.com/512x512.jpg"
],
"error_message": null,
"metadata": null,
"steps": 2000,
"created_at": "2023-01-31T15:16:39.959Z",
"model": "sd1.5"
}
Delete a Fine-Tune and its Prompts
This endpoint allows you to delete Fine-Tunes in Bodia.
Note: This will also delete the Prompts of that Fine-Tune and the model.
Request
curl -X DELETE https://api.bodia.ai/v1/finetunes/{TUNE_ID} \
-H "Authorization: Bearer {API_KEY}"
Response
{
"delete": "it has been deleted successfully"
}